CHEMICAL ANALYSIS. A MANUAL OF CHEMICAL ANALYSIS, QUALITATIVE AND QUANTITATIVE. OF ST,UJ)ENT8. HENEY M. NOAD, Pn.D.,F.R.S v F.C.S., AUTHOR OF 'A MANUAL OF ULKCTKICITY,' LECTUEER ON CHEMISTRY AT ST. GEORGE'S HOSPITAL. LONDON : LOVELL REEVE & CO., 5, HENRIETTA STREET, COVENT GARDEN. 1864. JOHN EDWARD TAKLOH, PKIMTER, 1.ITTLK QUEEN' STRKKT, LINCOLN'S INN FIELDS. CONTENTS. CHAPTER I. GENERAL REMARKS ON CHEMICAL ANALYSIS, 1-3. QUALITATIVE ANALYSIS OPERATIONS TO BE PERFORMED, AND APPA- RATUS REQUIRED : Solution, 4. Precipitation, 8-9. Filtration, decantation, washing, 10-12. SOURCES OF HEAT, 13. Griffin's Gas-burner, 14. Normandy's Gas-burner, 15. Berzelius's Spirit-lamp, 16. FURNACES : Faraday's, Aikin's, Black's, 17. CRUCIBLES : Hessian, Cornish, Black-lead, Berlin, Platinum, and Silver, 18. THE BLOWPIPE : Black's, Plattner's, 19. STRUCTURE OF FLAME : the oxidizing flame, the reducing flame, 20. Blowpipe supports for fusions and reductions, 21. Blowpipe spoons, forceps, basins, and mortars, 22. Evaporation, 23. Distillation, precautions necessary ; Liebig's condenser ; Woulfe's appa- ratus ; tube retort, and receiver ; sublimation, 24. CHAPTER II. ON REAGENTS. 1, Blue litmus paper ; 2, Red litmus paper ; 3, Turmeric paper ; 4, Geor- gina paper ; 5, Solution of Indigo ; 6, Starch paste ; 7, Lead paper * VI CONTENTS. 8, Sulphuric acid ; 9, Nitric acid ; 10, Hydrochloric acid ; 11, Nitro-hy- drochloric acid ; 12, Acetic acid ; 13, Oxalic acid ; 14, Tartaric acid ; 15, Hydrate of Potassa ; 16, Carbonate of Potassa ; 17, Carbonate of Soda ; 18, Ammonia; 19, Sesquicarbonate of Ammonia ; 20, Chloride of Am- monium ; 21, Sulphide of Ammonium ; 22, Sulphide of Potassium ; 23, Oxalate of Ammonia ; 24, Hydrosulphuric acid ; 25, Ferrocya- nide of Potassium ; 26, Ferridcyariide of Potassium ; 27, Chromate of Potassa ; 28, Sulphate of Potassa ; 29, Bisulphate of Potassa ; 30, Cream of Tartar ; 31, Cyanide of Potassium ; 32, Antimoniate of Potassa ; 33, Caustic Baryta ; 34, Chloride of Barium ; 35, Nitrate of Baryta ; 36, Phosphate of Soda ; 37, Phosphate of Soda and Ammonia ; 38, Nitrate of Potassa ; 39, Biborate of Soda ; 40, Nitrate of Silver ; 41, Ammonio- nitrate of Silver ; 42, Sulphate of Copper ; 43, Iodide of Potassium ; 44, Acetate of Lead ; 45, Basic Acetate of Lead ; 46, Protosulphate of iron ; 47, Lime-water ; 48, Sulphate of Lime ; 9, Chloride of Calcium ; 50, Protochloride of Tin ; 51, Bichloride of Platinum ; 52, Sesquichloride of Iron ; 53, Sulphurous acid ; 54, Hydrofluosilicic acid ; 55, Chlorine- water ; 56, Chloride of Mercury ; 57, Subnitrate of Mercury ; 58, Mo- lybdate of Ammonia ; 59, Protonitrate of Cobalt ; 60, Distilled water ; 61, Alcohol; 62, Ether, 25. CHAPTER III. . ON THE COMPORTMENT OF THE PRINCIPAL METALLIC OXIDES WITH REAGENTS. Arrangement of the metallic oxides into groups, and analytical classifica- tion of them, 26. General characters of, and method of detecting POTASSA, 27. General characters of, and method of detecting SODA, 28. General characters of, and method of detecting LITHIA, 29. General characters of, and method of detecting AMMONIA, 30. General characters of BARYTA, and comportment with reagents, 31. General characters of STRONTIA, and comportment with reagents, 32. General characters of LIME, and comportment with reagents, 33. General characters of MAGNESIA, and comportment with reagents, 34. General remarks on the Oxides of the SECOND GROUP, 34. General characters of SESQUIOXIDE OF ALUMINUM, and comportment with reagents, 35. General characters of OXIDE OF GLUCIXUAI, and comportment with re- agents, 36. CONTENTS. Vll General characters of OXIDE OF YTTRIUM, and comportment with re- agents, 37. General characters of OXIDE OP THORINUM, and comportment with re- agents, 38. General characters of OXIDE or ZIRCONIUM, and comportment with re- agents, 39. General characters of SESQUIOXIDE or CHROMIUM, and comportment with reagents, 40. General characters of TITANIC ACID, and comportment with reagents, 41. General remarks on the Oxides of the THIRD GROUP, 41. General characters of OXIDE OF ZINC, and comportment with reagents, 42. General characters of OXIDE OF NICKEL, and comportment with re- agents, 43. General characters of OXIDE OF COBALT, and comportment with re- agents, 44. General characters of OXIDE OF MANGANESE, and comportment with re- agents, 45. General characters of PROTOXIDE OF IRON, and comportment with re- agents, 46. General characters of SESQUIOXIDE OF IRON, and comportment with re- agents, 47. General characters of SESQUIOXIDE OF URANIUM, and comportment with reagents, 48. General remarks on the Oxides of the FOURTH GROUP, 48. General characters of PROTOXIDE OF LEAD ; and comportment with re- agents, 49. Poisoning by Lead ; detection of the metal in organic mixtures, 50. Action of Water on Lead, 51. Recognition and quantitative estimation of Lead in Water, 52. General characters of OXIDE OF SILVER, and comportment with re- agents, 53. General characters of SUBOXIDE OF MERCURY, and comportment with reagents, 54. General characters of OXIDE OF MERCURY, and comportment with re- agents, 55. Poisoning by Mercury ; detection of the metal in organic mixtures, 56. General characters of TEROXIDE OF BISMUTH, and comportment with re- agents, 57. General characters of OXIDE OF CADMIUM, and comportment with re- agents, 58. Vlll CONTENTS. General characters of OXIDE OP COPPER, and comportment with re- agents, 59. Poisoning by Salts of Copper, and detection of the metal in organic mix- tures, 60. G-eneral characters of OXIDE OF PALLADIUM, and comportment with re- agents, 61. General characters of SESQTTIOXIDE OF RHODIUM, and comportment with reagents, 62. General characters of the Oxides of OSMIUM ; General remarks on the Oxides of the FIFTH. GROUP (Section A), 63. General characters of TEROXIDE OF ANTIMONY, and comportment with reagents, 64. Detection of Antimony in organic mixtures, 65. General characters of ARSENIOUS ACID, and comportment with reagents, 66. Detection of Arsenic in organic mixtures, 67. General characters of ARSENIC ACID, and comportment with reagents, 68. General characters of PROTOXIDE OF TIN, and comportment with reagents, 69. General characters of PEROXIDE OF TIN, and comportment with reagents, 70. General characters of BINOXIDE OF PLATINUM, and comportment with reagents, 71. General characters of BINOXIDE OF IRIDIUM, and comportment with re- agents, 72. General characters of TEROXIDE OF GOLD and comportment with re- agents, 73. General characters of SELENIOUS ACID, and comportment with reagents, 74. General characters of TELLUROUS ACID, and comportment with reagents, 75. General characters of BINOXIDE OF TUNGSTEN, and of TUNGSTIC ACID, 76. General characters of the OXIDES OF YANADIUM, 77. Binoxide of Vanadium, 78. Vanadic Acid, 78. General characters of the OXIDES OF MOLYBDENUM, 80. Molybdic Oxide, 81. Molybdic Acid, 82. General remarks ou the Oxides of the FIFTH GROUP (SECTION B), 82. CONTENTS. lx CHAPTER IV. ON THE COMPORTMENT OF THE PRINCIPAL INORGANIC AND ORGANIC ACIDS WITH REAGENTS. Classification of the Acids, 83. General characters of CARBONIC ACID, and comportment with reagents, 84. General characters of SULPHUROUS ACID, and comportment with reagents, 85. General characters of HYPO-SULPHUROUS ACID, and comportment with re- agents, 86. General characters of HTPOSULPHUEIC ACID, 87. General characters of SULPHURIC ACID, and comportment Tnth reagents, 88. * General characters of SELENIC ACID, 89. General characters of PHOSPHORIC ACID, comportment with reagents: Characters of the different forms of TRIBASIC alkaline phosphates, 90. General characters of PHOSPHOROUS and HYPOPHOSPHOROUS ACIDS, 91. General characters of BoRACic ACID, and comportment with reagents, 92. General characters of SILICIC ACID, 93. General characters of HYDROCHLORIC ACID, and comportment with re- agents, 94. General characters of CHROMIC ACID, and comportment with reagents, 95. General characters of HYDROCHLORIC ACID, and comportment with re- agents, 96. General characters of HYDROBROMIC ACID, and comportment with re- agents, 97. General characters of BROMIC ACID, and comportment with reagents, 98. General characters of HYDRIODIC ACID, and comportment with reagents, 99. General characters of IODIC ACID, and comportment with reagents, 100. General characters of HYDROCYANIC ACID, and comportment with re- agents, 101. Detection of Hydrocyanic Acid in organic mixtures, 102. Comportment of the SULPHOCYANIDES with reagents, 103. Comportment of the FERROCYANIDES with reagents, 104. Comportment of the FERRIDCYANIDES with reagents, 105. X CONTENTS. General characters of HYDKOSULPHUEIC ACID, and comportment with reagents, 106. General characters of NITEIC ACID, and comportment with reagents, 107-8. General characters of NITEOUS ACID, and comportment with reagents, 109. General characters of CHLOEIC ACID, and comportment with reagents, HO. General characters of OXALIC ACID, and comportment with reagents, HI. General characters of TAETAEIC ACID, and comportment with reagents, 112. General characters of CITEIC ACID, and comportment with reagents, 113. General characters of MALIC ACID, and comportment with reagents, 114. General characters of SUCCINIC ACID, and comportment with reagents, 115. General characters of BENZOIC ACID, and comportment with reagents, 116. General characters of TANNIC ACID, and comportment with reagents, 117. General characters of GALLIC ACID, and comportment with reagents, 118. General characters of ACETIC ACID, 119. General characters of FOEMIC ACID, 120. General characters of UEIC ACID, 121. General characters of MECONIC ACID, 122. CHAPTER V. ON THE COMPORTMENT OF THE PRINCIPAL POISONOUS ALKA- LOIDS WITH REAGENTS. General characters of NICOTINE, and reactions with reagents, 123. General characters of CONIA, 124. General characters of MOEPHIA, and reactions with reagents, 125. General characters of NAECOTINA, 126. General characters of STEYCHNIA, and reactions with reagents, 127. General characters of BETICIA, and reactions with reagents, 128. General characters of ACOKITINA, 129. General characters of ATEOPIA, 130. Detection of the poisonous alkaloids in organic mixtures, 131. CONTESTS. XI CHAPTER VI. ON SYSTEMATIC QUALITATIVE ANALYSIS. General observations, 134. Preliminary examination of a solid compound, 135-6. Preliminary examination of a liquid, 137-9. Examination for bases: principles upon which the method depends, 140-1. Detailed scheme of analysis, 142. Preliminary examination for acids, 143. General rules for the preparation of the solution to be tested for acids ; detailed scheme of analysis, 144. CHAPTER VII. ON QUANTITATIVE OPERATIONS. WEIGHING : the balance ; Eobinson's, Kater's, Oertling's : rules to be observed in weighing ; weights, li-5, 146. SPECIFIC GRAVITY : determination of the specific gravity of a solid ; a, heavier than water ; b, lighter than water ; of a powder ; of a liquid ; Hydrometers : Beaume's, Neiibauer's ; determination of the specific gra- vity of a gas ; of a vapour ; a, method of Gay-Lussac ; b, method of Dumas, 147. MEASURING : the burette and pipette, 148. DESICCATION: the water- bath and water-oven; the air-bath; drying in vac no, 149. CHAPTER VIII. ON THE QUANTITATIVE ESTIMATION OF SUBSTANCES, AND THEIR SEPARATION FROM EACH OTHER. Table of names, symbols, and equivalents of the elements, 150. Quantitative estimation of the bases ; quantitative determination of PO- TASSIUM, 151. Quantitative determination of SODIUM, 152. Separation of SODA from POTASSA, 153. Alkalimetry : methods of Gay-Lussac and Fresenius, 154. Quantitative estimation of AMMONIA ; analysis of ammoniacal salts ; volu- metric estimation of Ammonia and Nitrogen ; separation of Ammonia from Potussa and Soda, 155. Xll CONTESTS. Quantitative estimation of BAEYTA, and separation from the alkalies, 156. Quantitative estimation of STEONTIA, and separation from Baryta and the alkalies, 157. Quantitative estimation of LIME ; separation from Baryta, Strontia, and the alkalies, 158. Quantitative estimation of MAGNESIA ; separation from Baryta, Strontia, Lime, and the alkalies, 159. Quantitative estimation of ALUMINA ; separation from the alkaline earths and alkalies, 160. Quantitative estimation of GLTTCINA ; separation from Alumina, the alka- line earths, and alkalies, 161. Quantitative estimation of YTTEIA, THOEINA, and ZIECONIA, 162-4. Quantitative estimation of SESQTJIOXIDE OF CHEOMIUM ; separation from Alumina, the alkaline earths, and alkalies; analysis of chrome ores, 165. Quantitative estimation of OXIDE OF ZINC ; separation from Oxide of Chro- mium, and from the Oxides of the Third, Second, and First Groups, 166. Quantitative estimation OF OXIDE OF NICKEL ; separation from Oxides of Zinc and Chromium, and from the Oxides of the Third and Second Groups ; analysis of Kupfernickel, 167. Quantitative estimation of OXIDE OF COBALT ; separation from Oxides of Nickel, Zinc, and Chromium ; from Alumina and the alkaline earths ; analysis of Cobalt ores ; extraction of Cobalt and Nickel from the ore, 168. Quantitative estimation of OXIDE OF IEON ; separation from Oxides of Nickel, Cobalt, Manganese, and Zinc, and from the Oxides of the Third and Second Groups ; dry assay of Iron ores ; humid assay of Iron ores ; analysis of pig-iron, 169. Quantitative estimation of OXIDE OF UEANIUM ; separation from other me- tallic oxides ; analysis of Pitchblende. 170. Quantitative estimation of OXIDE OF MANGANESE; separation from the Oxides of the Fourth, Third, and Second Groups ; determination of the value of commercial Binoxide of Manganese, 171. Quantitative estimation of LEAD ; separation of Oxide of Lead from the Oxides of the Fourth, Third, and Second Groups ; analysis of galena ; separation of Sulphate of Lead from Sulphate of Baryta, 172. General characters of OXIDE OF THALLIUM ; comportment of its salts with reagents ; properties of metallic Thallium, 173. Quantitative estimation of SILVEE ; estimation of silver in alloys ; wet method ; dry method ; cupellation, 174. Quantitative estimation of MEECUEY ; separation of Oxide of Mercury from Oxide of Silver, 175. CONTENTS. Xlll Quantitative estimation of OXIDE OF BISMUTH ; separation from Oxides of Lead, 176. Quantitative estimation of OXIDE OF CADMIUM ; separation from Oxides of Lead, Silver, Mercury, Bismuth, and Zinc, 177. Quantitative estimation of COPPER ; estimation of Copper in ores ; volu- metric methods of estimating Copper ; separation of Oxide of Copper from Oxides of Lead, Silver, Mercury, Bismuth, and Cadmium; analysis of Brass ; analysis of commercial Copper, 178. Quantitative estimation of PALLADIUM ; separation from Copper, 179. Quantitative estimation of KHODIUM ; separation from Copper, 180. Extraction of OSMIUM, KUTHENIUM, and IEIDIUM from Platinum ores, 181. Quantitative estimation of ANTIMONY ; assay of Antimony ores, 182. Quantitative estimation of AESENIC ; separation from Antimony, 183. Quantitative estimation of TIN ; analysis of alloys of Tin and Copper ; separation of Tin from Antimony ; from Antimony and Lead ; from An- timony and Arsenic, 184. 'Quantitative estimation of PLATINUM ; separation from Antimony, Ar- senic, and Tin ; analysis of Platinum ores, 185. Quantitative estimation of GOLD ; separation from Platinum ; from Anti- mony, Arsenic, and Tin ; analysis of alloys of Gold ; purification of Gold by cementation, 186. Quantitative determination of SELENIUM ; analysis of Seleniate of Lead, 187. Quantitative estimation of TELLURIUM ; separation from Selenium and Sulphur, 188. Quantitative estimation of TUNGSTEN ; analysis of Wolfram, 189. Quantitative determination of MOLYBDENUM ; analysis of Molybdate of Lead, 190. Quantitative estimation of .TITANIC ACID ; estimation in silicates ; sepa- ration from the metals of the Fifth Group ; from Oxide of Iron ; de- termination of Titanium in pig Iron, 191. Quantitative estimation of SULPHURIC ACID ; analysis of compounds of Sulphur with Oxygen ; estimation of Hydrosulphuric acid ; separation of Sulphur from the alkaline earths and alkalies ; analysis of compounds of Sulphur with bases ; of Iron and Copper Pyrites ; of Sulphide of Lead ; estimation of Sulphur in nitrogenous, and in volatile organic compounds, 192. Quantitative estimation of PHOSPHORIC ACID ; volumetric processes ; ana- lysis of bone earth, 193. Quantitative estimation of BORACIC ACID ; determination of, in silicates ; separation of, from Phosphoric acid, 194. XIV CONTENTS. Quantitative estimation of SILICIC ACID ; analysis of natural silicates, 195. Quantitative estimation of CARBONIC ACID ; analysis of carbonates ; sepa- ration from other acids ; analysis of gunpowder, 196. Quantitative estimation of OXALIC ACID, 197. Quantitative estimation of HYDROCHLORIC ACID ; analysis of chlorides ; estimation of free Chlorine ; chlorimetry, 198. Quantitative estimation of IODINE ; separation from Chlorine ; determina- tion in Kelp, 199. Quantitative estimation of BROMINE; separation from Chlorine and Iodine, 200. Quantitative estimation of HYDROCYANIC ACID ; analysis of cyanides ; separation of Cyanogen from Chlorine, Bromine, and Iodine, 201. Quantitative estimation of HYDROFLUORIC ACID ; analysis of fluorides, 202. Quantitative estimation of NITRIC ACID ; analysis of nitrates ; valuation of Nitre, 203. Quantitative estimation of CHLORIC ACID ; analysis of chlorates, 204. CHAPTER IX. ON THE ELEMENTARY ANALYSIS OF ORGANIC BODIES. Preliminary testing of the organic substances ; analysis of non-nitrogenous solids, 206(1). Analysis of non-nitrogenous volatile compounds, 206 (2). Analysis of organic substances containing Nitrogen, 206 (3). Determination of the rational formula, and atomic weight of an organic substance, 207. Gas as a fuel in organic analysis, 207. CHAPTER X. ON THE ANALYSIS OF MINERAL WATERS. Qualitative analysis, 208. Quantitative analysis, 209. Determination of the degree of hardness of natural water, 209 (xviii.). Arrangement of the results of analysis, 209 (xix.). Analysis of the saline water of Purton, 209. Determination of the volume of gases, 209 (xx.). CONTENTS. XV CHAPTEK XI. ON THE ANALYSIS OF SOILS. I. Mechanical analysis, 210. Selection of sample, 210. II. Chemical analysis, 210. The organic constituents of the soil, 210. Determination of the organic acids, 210 (vii., viii.). Determination of the total amount of Carbon, 210 (ix.). Determination of the Nitrogen, 210 (5.). Analysis of LIMESTONES, 211. Analysis of CLAYS, 212. Analysis of MANURES, 213. Analysis of GUANO, 213 (1). Analysis of SUPERPHOSPHATE OF LIME, 213 (2). Analysis of BONES, 214. CHAPTER XII. ON THE ANALYSIS OF THE ASHES OF VEGETABLE AND ANIMAL SUBSTANCES. Analysis of the ashes of VEGETABLES, 215. Method of Will and Fresenius, 215 (1). Method of Rose, 215 (2). Method of Staffel, 215 (3). Method of Way and Ogston, 215 (4). Method of Strecker, 215 (5). Analysis of ashes of BLOOD, 216 (1). Analysis of ashes of ANIMAL SUBSTANCES generally, 216 (2). APPENDIX. TABLE I. For the Calculation of Analyses. TABLE II. French measures of weight, with their English equivalents. TABLE III. French measures of length, with their English equivalents. TABLE IV. French measures of surface, with their English equivalents. TABLE V. French measures of capacity, with their English equivalents. XVI CONTENTS. TABLE VI. For the comparison of the Centigrade and Fahrenheit ther- mometers. TABLE VII. Showing the quantity of oil of vitriol of specific gravity T8485, and of anhydrous acid, in 100 parts of dilute sulphuric acid of different densities, according to Dr. Ure. TABLE VIII. Showing the quantity of real or anhydrous nitric acid in 100 parts of liquid acid, at different densities, according to Dr. Ure. TABLE IX. Showing the quantity of absolute alcohol in spirits of dif- ferent specific gravities, according to Lowitz. TABLE X. Showing the quantity of absolute alcohol by weight in mix- tures of alcohol and water of -different specific gravities, according to Mr. Drinkwater. TABLE XI. Showing the tension of aquecms vapour at different tem- peratures. ERRATA IN PART I. Page 54. Lino 4, for " (NaO,NH 4 O,PO 5 ) " read (NaO,NH,O, HO,P0 6 + 8aq.)." 64. 16 lines from the bottom, for " 4(HgI,KO) " read 67. 3 lines from the bottom, add to the second part of the equation + 5(NaO,SO 3 ). 75. 9 lines from the top, for " glaucina " read " glucina ;" 19, ditto, for " it " read " its." 82. 15 lines from the top, for " 2FeCl " read " 2Fe 2 01 3 ." 109. 14 lines from the bottom 1 , for " oxide " read " suboxide." 111. 3 lines from the top, for " sesquichloride " read " terchlo- ride ;" 12, ditto, for " soluble in excess " read " soluble in alkaline sulphides." 112. 16 lines from the top, for " is tartaric acid" read "in tartaric acid." 141. 14 lines from bottom, to " starch paste " add " and iodic acid." 151. 11 lines from the top, for " Hydrofluosilicic " read "Hy- drofluoric" ,, 164. 5 lines from the bottom, for " KCfy " read " K,Cfy." 165. 10 lines from the bottom, for " (3FeO,SO 3 ) " read "3(FeO,SO 3 )." 170. 19 lines from the bottom, for " C1 5 O " read " C1O 5 ." ERRATA IN PART II. Page 256. After " strontium," line 26, insert : Element. Equivalent. Principal compounds with oxygen. S. Sulphur. 16 SO 2 , SO 3 Page 260. Line 22, for " new " read " two." 317. Line 18, read thus : 3. Ores containing silica and various bases, but little ' or no lime. 4. Ores containing one or more bases, such as lime, magnesia, alumina, etc. etc., but little or no silica. 336. Line 39, for 4*35 read 43'5. 506. Line 39, for three " read " four." 527. Line 15, for "formed" read "found." 533. Line 25, for " the dry mass is then fused in a Hessian crucible" read "the dry mass is exposed for several hours to a tolerably strong heat on the sand-bath, but not fused." MANUAL OP CHEMICAL ANALYSIS. PART L : ; .- QUALITATIVE. CHAPTEE I. GENEEAL EEMAEES ON CHEMICAL ANALYSIS. 1. CHEMISTRY is essentially an experimental science. Its ob- ject is the investigation of the material constituents of the globe, and the study of their different properties and relations. The student is, therefore, constantly engaged in observing the phenomena presented on submitting the different substances that come under his notice to the action of various agents, and in the accumulation of facts derived from experimental in- quiry. It is, indeed, this circumstance which gives to Chemis- try its principal charm ; the tyro no sooner begins to read than he begins to make experiments, and being thus enabled, to a certain extent, to verify for himself the facts brought before him, he acquires an interest in his pursuit which attaches him daily more and more to it. 2. In order to pursue Analytical Chemistry with any chance of success, the student must possess certain qualifications ; these may be stated to be habits of strict order and scrupu- lous neatness ; a dexterity of manipulation which practice alone can give ; a firm conviction that the laws of nature are unchangeable, and that variations in experimental results must consequently imply either a non-fultilment of certain necessary conditions, or some error in manipulation; and a rigorous PABT I. B 2 QUALITATIVE ANALYSIS. honesty, not only in recording results, but in experimenting, and in interpreting the pheDomena which present themselves in the course of an investigation. 3. The determination of the constituent parts of a compound body is termed its Analysis ; and this may be either of two kinds, according to the object which the operator has in view. If he merely seek a knowledge of the general nature of the substance, he is satisfied, when by the application of certain tests, and by the performance of certain operations, he has obtained evidence of the presence of those elements of which tha compound , is." made up, and the analysis he performs is called Qualitative ; but if he desire to appeal to the balance, and to ascertain not only the nature, but the actual amount of the. elements present, he must shape his analysis in such a manner as to separate the constituents of the compound com- pletely from each other, and obtain them either pure or in some other well-known form of combination : he then performs a Quantitative analysis. For example, if on the addition of a few drops of solution of ferrocyanide of potassium to a clear liquid a blue precipitate be obtained, the operator is at once satisfied of the presence of iron ; and if he is working qualita- tively only, this experiment is with regard to that particular element conclusive, it requires no additional confirmation ; but if he desire to estimate the exact amount of iron present, he must conduct his analysis in such a manner as to separate every particle of the metal in the state of sesquioxide, in which state it is weighed, and the amount of metal deduced by calcu- lation. Now, before the chemist can proceed to estimate the proportions of the constituents of a compound body, he must know exactly how many elements are present, and what those elements are. The Qualitative analysis always, therefore, pre- cedes the Quantitative ; and it will be necessary to separate the two courses of study, and to treat each individually. QUALITATIVE ANALYSIS OPEEATIONS TO BE PEEFOEMED AND APPAEATUS EEQUIEED. SOLUTION. 4. To prepare a substance for the exertion of chemical action, and to suit it for the reception of the various tests and reagents to which we may wish to submit it, recourse is always, if possible, had to solution. Now solution is of two kinds, OPERATIONS AND APPARATUS. 3 being effected either by fluids which exert no chemical action on the body dissolved, as when sugar is dissolved in water, or camphor in spirits of wine ; or by fluids which do exert a chemical action on the substance, as when chalk is dissolved in vinegar, or iron in dilute sulphuric acid. Whenever it is prac- ticable, the substance under examination is dissolved in water ; but, if that menstruum should fail, recourse is had to other fluids, the selection of which requires care and judgment. 5. A compound may be made up of constituents, some of which are soluble in water, and some insoluble. In order to ascertain this point, a small quantity of the substance, brought into a fine state of division by pulverization, is introduced into a test tube covered with distilled water and submitted to heat ; the liquid is filtered, while hot, into a small platinum or porce- lain capsule, and evaporated carefully and slowly ; if a residue remain, it of course indicates that a substance soluble in water existed in the compound. Should there be reason to suspect the presence of a volatile body soluble in water, heat should not be applied in the first instance, but the substance under ex- amination repeatedly agitated with cold distilled water, and the clear solution allowed to evaporate spontaneously. Bodies, how- ever, capable of assuming the gaseous state by the heat of boil- ing water, may in most cases be recognized by their odour, or by the application of certain tests which will hereafter be described. 6. Alcohol, ether, pyroxylic spirit, and certain oils, are occasionally successfully employed to dissolve substances on which water has no action ; sometimes also a mixture of alcohol and water is found very useful. For example, if the object be to separate sulphate of lime (gypsum) from chloride of sodium (common salt), this cannot be effected by water alone, sulphate of lime being partially soluble in that fluid ; but by employing a mixture of one part of alcohol and two or three parts of water the two substances are easily separated, sulphate of lime being insoluble in diluted spirit. Acids, again, sometimes act as mere solvents, and may be conveniently employed as such ; thus oxalate of lime dissolves in hydrochloric acid without de- composition, as does also phosphate of lime, and both of these salts may be obtained again from their acid solution, unaltered by supersaturation with ammonia. 7. The apparatus required for the purpose of solution is in general very simple. For qualitative experiments there is nothing so convenient as the Test Tube. It should be made B 2 4 QUALITATIVE ANALYSIS. of hard German glass, and the bottom blown round and tolerably thin. The student should possess a good stock of them : they should be of various sizes, from three to six inches long, and of such a diameter that they may readily be closed with the thumb, for the purpose of agitation. They should be kept ready for use in a rack, to which it is advisable that there should be attached a series of pegs on which the tubes may be inverted, to drain after being washed. If the tube-rack stand be provided with a couple of drawers, they will be found to add to its convenience : they serve to hold the blowpipe, platinum foil and wire, charcoal supports, fluxes, etc. etc. A spring- holder is a useful accompaniment to the test-tube apparatus ; it serves to grasp the tube while its contents are under the influence of heat. When larger quantities of liquid have to be operated upon, flasks of various sizes may be substituted for test tubes. They should be of thin glass, with flattened bottoms and bordered mouths, ground smooth for the reception of corks, and to admit of being closed air-tight by ground- glass plates. The common Florence-oil flask is a useful chemical vessel ; the edge of the mouth should be rounded in the blowpipe flame, whereby it is rendered less likely to crack by the insertion of a cork. It is, generally speaking, of very uniform thickness throughout, and will bear the direct applica- tion of heat without cracking. As a general rule, however, the naked flame of the oil, gas, or spirit lamp should not be allowed to come into actual contact with the bottoms of large glass vessels, but the heat communicated through the medium of sand ; for this purpose a set of thin shallow iron dishes should be provided, from three to six inches in diameter, which should be about half filled with fine siliceous sand. They may be supported on the rings of a retort stand, and they then form very convenient media for conveying heat to flasks, beakers, etc. To support Florence flasks, and other round-bottomed vessels, rings of thin sheet-iron or copper, covered with list, will be found very serviceable : they should be of different diameters and heights ; they are likewise of great use for sup- porting evaporating-basins, and the laboratory should, therefore, be plentifully supplied with them. PRECIPITATION. 8. The term precipitation does not necessarily imply the subsidence of a solid substance to the bottom of a vessel, inas- OPERATIONS AND APPARATUS. 5 much as the body rendered insoluble may either float on the surface of the liquid, or remain suspended in it. When the chemist speaks of a body being precipitated, he means that it has, by the action of certain agents, passed from a soluble to an insoluble state. The body separated is called the precipitate, and the substance occasioning the separation is called the pre- cipitant. The vessels employed for effecting precipitations may be the same as those already described for solutions, to which may be added the beaker, a thin-bottomed tumbler, which may be procured at any chemical glass warehouse, of all sizes, generally in nests of eight or ten from half a gill to a pint and upwards. For small quantities the test tube is always employed ; when the quantity is large, the beaker is, on the whole, the most convenient, from the admirable manner in which it bears sudden elevations of temperature. Precipitates make their appearance in very different forms, and under very different circumstances. The student must, therefore, be careful not to draw too hasty conclusions with regard to the action which bodies exert on each other ; and if, on bringing two substances into contact, he observe no immediate effect, he must not, on that account, conclude that no action has taken or will take place. A few examples will serve to illustrate this. 1st. The precipitate may not make its appearance until after the lapse of a considerable time. Thus, in testing for magnesia, by means of phosphate of soda and ammonia, if the earth exist in the solution only in very minute quantity, no precipitate will perhaps make its appearance for several hours ; days may even elapse ere the minute stellar crystals of the double phosphate^of magnesia and ammonia can be distinctly recognized. 2nd. Decomposition may have taken place, and yet, from the solubility of the educt in the fluid menstruum, no precipitate makes its appearance. Thus, on adding oxalic acid to a solution of lime in hydrochloric acid containing free acid, no precipitate appears, oxalate of lime being soluble in hydrochloric acid; but, on neutralizing the liquid by ammonia, oxalate of lime is immediately precipitated. Again, on adding to an aqueous solution of malate of soda a few drops of chloride of calcium, the fluid still remains clear, malate of lime being soluble in water, but on the addition of alcohol it instantly separates as a white powder, 3rd. Occasionally (though this is rare) the precipitate makes its appearance only on the appli- cation of heat, the newly-formed compound being more soluble O QUALITATIVE ANALYSIS. in cold than in hot water. Thus, on adding lime water to a cold solution of citric acid, or a citrate, no precipitate takes place till the solution is heated, when basic citrate of lime appears as a white powder, which redissolves as the solution cools : again, on addiog a cold dilute solution of caustic potassa to newly precipitated tartrate of lime, a clear solution is ob- tained, but on heating, a gelatinous precipitate separates, which is again dissolved as the solution cools. 4th. The precipitated substance is frequently redissolved by an excess of the precipi- tant : thus, hydrated oxide of chromium, hydrated oxide of zinc, and alumina, are each thrown down from solutions of their salts by potassa, but the precipitates are all completely and in- stantly redissolved by the addition of an excess of the alkali : again, on pouring iodide of potassium into a solution of cliloride of mercury, a beautiful scarlet precipitate immediately ensues ; but, if excess of the iodide be added, the precipitate is 'dissolved, and a clear and colourless solution obtained. 9. The forms in which precipitates make their appearance have been distinguished by various names, such as crystalline, examples of which are furnished by the acetate of silver, a salt formed on adding solution of a neutral acetate to solution of nitrate of silver ; by the double chloride of platinum and potas- sium, and by the bitartrate of potassa ; pulverulent, as the oxalate and sulphate of lime ; flocculent, as alumina ; curdy, as chloride of silver ; gelatinous, as when oxide of zinc is precipi- tated from its solution by potassa or ammonia, or as when chloride of calcium is added to an aqueous solution of hydro- fluoric acid, or of the soluble fluorides, in which case the pre- cipitate is so transparent that it may easily be supposed that no change whatever has occurred. Sometimes, again, a reaction is evinced by the mere turbidity of the liquid, as when nitrate of silver is added to a solution containing minute traces of chlorine, or when a solution of a protosalt of tin is diluted with water, in which case the neutral salt is decomposed into a soluble acid and an insoluble basic salt ; or as when solution of protochloride of tin is added to a very dilute solution of gold, when a purple-red tint is produced without the formation of any precipitate. The subsidence of a precipitate is frequently greatly facilitated by agitation. In cases where an elastic substance is evolved by agitation, or where there is excess of ammonia present, the flask employed should not be too thin, or it will be liable to burst from the pressure of the vapour : OPERATIONS AND APPARATUS. 7 the addition of acids and salts also occasionally greatly facilitates precipitation ; chloride of silver and sulphate of baryta, for example, separate much more readily by the addition of nitric acid; and oxide of tin, formed by the action of nitric acid on the metal, has its precipitation hastened by the addition of common salt. Oxide of iron again is sometimes eliminated in so finely comminuted a form that it remains suspended in the liquid for days ; an addition of a known quantity of nitrate of lead, and a corresponding quantity of sulphuric acid, may often be very advantageously made to the turbid liquid ; a known quantity of sulphate of lead is thereby introduced, which in its precipitation seldom fails to carry with it the whole of the suspended oxide of iron ; lastly, Prussian Hue precipitates from a solution containing free hydrochloric acid much more readily and completely than from a neutral liquid. In qualita- tive testing, where the object is generally merely to ascertain the colour and general chemical characters of a precipitate, a few drops of the precipitant usually suffice, and the operation may almost always be performed in a test tube : but in quanti- tative experiments, where every particle of the precipitate must be obtained, the precipitant is added drop by drop until no further effect is produced, and the operation is most con- veniently conducted in a beaker. FILTRATION, DECANTATION, AND WASHING. 10. The separation of a precipitate from the fluid in which it has been produced may be effected either by filtration or decantation. For the former purpose white unsized paper is used, the perfect purity of which in qualitative operations is not very often a matter of consequence : it should be suf- ficiently strong to bear the weight of a considerable quantity of fluid, and yet so porous as to admit the free and ready passage of liquids. There is no difficulty in procuring paper of this kind, though it is not easy to obtain unobjectionable paper for quantitative operations ; for, as it frequently happens that the filter has to be ignited with the precipitate, it is of course necessary to take into account the quantity and quality of the ashes left by the paper. An excellent article is some- times to be met with under the name of Swedish paper, a filter four inches in diameter of which does not leave more than a hundredth of a grain of ash. It is difficult, however, to pro- cure it genuine, and it is therefore advisable to free paper QUALITATIVE ANALYSIS. intended for minute and delicate experiments as much as possible from inorganic matter by treating it with hydrochloric acid, and subsequently washing with distilled water, so as to remove every trace of the acid : after drying, it may be con- sidered fit for use. The inorganic impurities in filtering paper are iron, lime, and sometimes magnesia. The paper is cut into a circular form of the required size and folded twice in opposite directions, so as to bring the four quadrants together; one quadrant is then opened from the other three so as to produce a conical cavity, as shown in Fig. 1. The paper thus prepared is placed inside a glass funnel; care should be taken that it does not extend be- yond, or even quite reach, the edge of the funnel. Previous to pouring the fluid to be filtered on the paper, the latter should be moistened with distilled water, which quickens the process and diminishes the chance of Fig- ! solid matter passing through the pores of the paper. In qualitative experiments where it is desirable to expedite the operation as much as possible, and in cases where a large quantity of a bulky precipitate has to be separated, the filtrate being the valuable part, as for example, in the separation of the alkaline earths by baryta water in the process for estimating alkalies, the filter may be plaited so as to prevent its close adhesion to the glass. Bibbed funnels are sometimes employed for the same purpose, but they do not answer the purpose nearly so well as plaited paper. It is difficult to de- scribe the method of making these plaited filters : an inspection of Eig. 2, which shows the appearance the paper should present, will probably, however, convey all the requisite information. In cases where the precipitate has to be carefully col- lected, plain filters should invariably be used, as it is difficult to remove the solid matter effectually, from the external and re-entering angles of the plaited paper. The funnel, during the operation of filtering, is placed on the ring of the Fig. 2. OPERATIONS AND APPARATUS. 9 retort stand, or in one of the holes of a filtering-stand, and the fluid, if in a beaker, is carefully poured down a glass rod in the manner shown in Pig. 3. The edge of the beaker is greased, by which the adhesion of particles of the liquid is prevented, and all loss from trickling down the outside of the glass obviated. The clear fluid as it drops from the funnel should not be allowed to fall directly into the receiving vessel, but caused to impinge on its side, as shown in the figure ; by attending to this, we avoid all chance of loss by the splashing of the liquid. These minutiae of course principally apply to quantitative experiments. The precipitate, being col- lected on the filter, has to be washed. This operation is best performed by means of the wash bottle seen in Fig. 3, the Fig. 3. method of using which useful piece of apparatus is sufficiently obvious ; it is filled with distilled water, and on forcing a stream of air from the lips through the short tube, a jet of water is propelled from the bent tube by the pressure of the air, and may be directed either in drops or in a pretty power- 10 QUALITATIVE ANALYSIS. fill' stream wherever it is required. It is useful to have four of these bottles at hand : one stout one, for cold water for general purposes ; one made from a thin flat-bottomed flask, for hot water ; and two smaller ones, for alcohol and ether. The wash- ing of the precipitate is to be continued until a few drops of the nitrate leave no residue when evaporated to dryness on a strip of platinum foil ; occasionally, however, this indication would be fallacious in consequence of the partial solubility of the precipitate in water ; in these cases, special testing of the nitrate must from time to time be had recourse to. In washing a precipitate on a filter, the solid matter should be washed from the sides of the paper and collected, as much as possible, in a thick stratum at the apex of the cone ; the water will not in this case pass through so quickly, but under more favourable circumstances for exerting its solvent power on the substances to be removed. During the operation of filtering, the funnel should be covered with a glass plate, to protect its contents from the dust and dirt of the laboratory. When the precipi- tate subsides rapidly and collects into a small space at the bottom of the vessel, the supernatant fluid may frequently be advantageously removed by a siphon, which may be made from a piece of glass tube about 2 feet long and O3 of an inch in- ternal diameter. The extremities should be somewhat con- tracted. When this instrument is about to be used, it is filled with water from the wash bottle, and, keeping the longer leg closed by a finger, the shorter leg is introduced into the fluid, taking care not to disturb the precipitate, and then, on remov- ing the finger, the fluid is permitted to run out into a vessel placed to receive it. With care, and by gradually inclining the vessel, nearly the whole of the clear fluid may be removed from a precipitate, the washing of which may either be continued by decantation, or, which in most cases is preferable, it may be thrown on a filter and washed as usual. In cases where small quantities of fluid have to be removed from a precipitate, the pipette is frequently found very serviceable. This is a narrow glass tube, with a bulb about an inch in diameter blown in it, and drawn out below to a moderately fine aperture. To use it, the aperture is immersed in the liquid, and the mouth being applied to the other end the liquid is raised by withdraw- ing the air ; the finger being then placed on the upper end of the tube so as to close it, the instrument may be removed, and its contents transferred to another vessel. When the object is OPERATIONS AND APPARATUS. 11 to remove finely divided matter from concentrated acids, filters of powdered glass are employed ; coarse fragments of glass are first put into the neck of the funnel, these are successively covered with other portions more and more comminuted, the top being finished by a layer of small fragments ; on this the acid is poured, and, on passing through, it becomes clear. It is sometimes of importance to prevent the access of air as much as possible during the filtration of certain liquids, as, for instance, in the separation of caustic soda or potash from carbonate of lime. An apparatus for this purpose has been described by Mr. Donovan : it consists of a vessel somewhat resembling in shape an hour-glass provided with a tube of com- munication between each bulbj in the narrow part of this vessel a small pellet of cotton or asbestos is somewhat loosely placed, the. liquid being introduced through the neck of the upper vessel, which is closed by a stopper ; now, for every drop of liquid that passes through the filter into the lower vessel, it is obvious that an equal volume of air must find its way through the tube of communication with the upper vessel. The filtration of the liquid thus proceeds in an uninterrupted manner without any communication with the external air. 11. The precipitate, being collected on the filter and well washed, has in quantitative experiments to be dried and weighed ; its further treatment for this purpose depends on whether or not it may be ignited. If its nature be such that it will not bear a high temperature without decomposition, which is the case with all organic substances, the filter must have been previously dried at the same temperature to which it is subsequently to be submitted with the precipitate, and in that state carefully weighed. As filtering-paper is a very hygroscopic substance, the filter must be weighed in a closed vessel (a platinum crucible, with an accurately fitting cover, may be used for the purpose) ; it must be introduced while hot, and it is advisable to allow the whole to cool over a vessel of sulphuric acid underneath a bell jar. The weight of the filter and crucible being noted, the precipitate is collected, and, the rinsing-water having been allowed to run off as much as possible, it is removed from the funnel, and placed upon two or three folds of bibulous paper, or, which is better, on a warm tile or brick, by which a great deal of the adher- ing water is removed ; it is subsequently dried in the crucible until it ceases to lose weight, and by simply subtracting the 12 . QUALITATIVE ANALYSIS. weight of the filter and crucible from that of the filter, crucible, and precipitate, we have the exact weight of the latter. 12. When the precipitate may be ignited, it is not neces- sary to determine the weight of the filter, but the weight of the ash which a filter of an equal size leaves on ignition must have been ascertained by a previous experiment. The pre- cipitate is allowed to get as dry as possible in the funnel, for which purpose it is well covered with a piece of bibulous paper, and placed for some hours in a warm situation; the contents of the filter are then transferred as completely as possible into the crucible in which it is to be ignited, and the filter itself burned, either by doubling it up, and, having applied a light to it, allowing it gradually to consume, holding it by a forceps over the crucible placed on a glazed sheet of paper ; or, in the manner recommended by Fresenius, by cut- ting it in small pieces and incinerating it on the cover of the crucible, heated to redness over a gas or spirit lamp. The heat must be continued till every trace of blackness is removed from the ash, and this, in some cases, as for instance when the subject of the experiment is the phosphate of magnesia and ammonia, takes a considerable time. Care must be taken to conduct the operation of burning the filter in a spot entirely protected from draught. The ash of the filter is mixed with the precipitate in the crucible, and the whole is exposed to a heat gradually increasing until it gets red-hot ; it is then covered, allowed to cool, and weighed: from the gross weight, that of the crucible and filter ash have to be deducted to arrive at the weight of the precipitate itself. The operator must be careful not to heat the crucible too strongly at first, for, if the pre- cipitate should not be perfectly dry, particles are apt to fly out. SOUECES OF HEAT. 13. Where gas can be obtained, there is nothing so con- venient or so cheap. When coal gas is allowed to mix with a sufficient quantity of atmospheric air to effect a complete com- bustion of its carbon, it burns with a flame which, though having but little luminosity, possesses great heat, and throws off no solid charcoal on cold bodies held over it. It is thus admirably adapted for the ignition of crucibles. Eig. 4 shows one of the simplest arrangement of the coal-gas jet for this purpose. The mixture of gas and air takes place in the copper cylinder, and it is inflamed above the wire-gauze, fastened over OPERATIONS AND APPARATUS. 13 the top. The crucible to be heated is supported on a triangle of platinum wires laid across the ring of a retort stand. 14. The author has Mr. Griffin's permission to avail himself of the following descriptions and drawings of a simple form of gas-burner, which, aided by suitable fittings, can be used as a convenient source of heat for most operations of the chemical laboratory and lecture table.* It will boil a quantity of liquid exceeding two gallons at once ; it will raise a 4f inch fire-clay crucible to full redness ; it will fuse anhydrous carbonate of soda in greater quantity than is required for the analysis of a siliceous mineral ; and it will melt small quantities of sterling silver. This amount of power is sufficient for most chemical operations that are not metal- lurgic. Tlie Gas Burner. Fig. 5 re- presents the gas-burner of this apparatus. The gas is supplied by the horizontal tube, whence it passes through a set of small holes into the box a, in which it mixes with atmospheric air that enters freely by the holes shown in the sketch. The gaseous mix- ture passes up the vertical tube 5, and is inflamed at the top, where it burns with a single, tall, blue flame, which gives no smoke, very little light, but much heat. In this condition, the apparatus differs from " Bunsen's gas-burner" only in size, c re- presents a thin brass cap, which fits the air- box a, but moves easily round it. d is a flat, cast-iron box, with many holes round the margin, and a few small ones on the top. This box fits loosely on the upper Fig. 5. part of the tube b, and when it is placed upon it, and the gas Fig. 4. First described in the c Chemical News ' of Nov. 2nd, 1861. 14 aUALITATIVE ANALYSIS. Fig. 6. is lighted, the flame produced consists of a series of radiating jets, forming a horizontal circular flame of about seven inches in diameter. Pig. 6 a shows a ring of horizontal flames thus produced, and b shows the single vertical flame.* The ring of flame is suited to the pur- poses of boiling and evapora- tion ; the single flame to igni- tion and fusion. The height of the apparatus represented by Fig. 5 is 12 inches ; the bore of the tube b is 1 inch ; and the diameter of the fire-box d is 4 inches. Bunsen's gas-burner, what- ever its size, is subject to two defects : sometimes the flame burns white and smoky, and sometimes it blows down, the gaseous mixture explodes, and the gas then burns with a smoky flame in the air-box , Fig. 5. The remedies for these defects are as follows : If the flame is white only when the gas is turned on very full, the remedy is to lessen the supply of gas ; but if the flame continues to burn white at the top when the gas is gradually turned off, and the mass of flame slowly sinks, then the holes which deliver the gas from the supply pipe into the air-box are too large, and are placed too directly under the centre of the vertical tube 5, Fig. 5, and these defects must be corrected in the instrument. Finally, when the flame blows down, it is because the supply of atmospheric air is too large in propor- tion to the supply of gas, and their relative proportions must be altered. To effect this alteration, the cap c is to be turned round on the air-box a, so as partially to close the holes, and thus lessen the supply of air. If, when the gas is alight, the flame needs to be lowered, first the supply of air is to be lessened, and then the supply of gas. If the flame is to be enlarged, first the supply of gas must be increased, and then the supply of air. In short, to prevent the flame blowing * Fig. 6 I represents a small variety of this gas-burner, in which the head is not removable, but the efflux of the mixed gases is regulated by a sliding valve. OPERATIONS AND APPARATUS. 15 Fig. 7. down, the gas must always be first placed in excess, and then have the proper quantity of air adjusted to suit it by means of the regulator c. Arrangement for Soiling and Evaporation. Fig. 7 represents this gas furnace arranged for boiling and evaporation, a is the gas-burner, Fig. 5 ; b, an iron stool with three legs ; c, a furnace body, or iron jacket, lined with plumbago or fire-clay. Fig. 8 shows the jacket and lining in section, and marks the position of the fire-box, d, of the gas-burner. This furnace is 14 inches high, and 9 inches in diameter. The three brackets fixed on the upper part of the jacket serve to support the vessel that contains the liquid that is to be boiled or evaporated. A porcelain basin of 16 or 18 inches in diameter can be thus supported. It is important to allow between the jacket c and the evaporating-basin plenty of space for the escape of the heated air, which ascends from the interior of the furnace. When the evaporating-basin is of small diameter, it may be supported on iron triangles, placed in the furnace c. Fig. 8 shows that, around the vertical tube of the gas-burner, a, there is in the bottom of the furnace, Fig. 27. and for the reduction of lead and tin oxides contained in calcined and uncalcined minerals, etc., is given by Plattner. A fine proof-clay is kneaded into a stiff paste with water, and having rubbed the surfaces of the boxwood press, Fig. 28, with oil, a slip of paper three inches in length and one- fourth of an inch in breadth is placed on the middle of the concavity of the press, which is seven-eighths of an inch wide and five-six- teenths of an inch deep, and upon this a small clay-ball about half an inch in diameter ; the upper surface of the press is then stamped horizontally on the clay mass as far as is re- quired. This being done, the superfluous clay will have exuded, and the handle or upper part of the apparatus can be removed easily by careful turning ; with a small knife the clay which is driven out may be cut away, and it can then be seen whether the basin is sufficiently thin and uniform ; if so, the slip of paper is gen- tly pulled and the dish extracted. After a few hours' drying, the paper detaches itself from the little clay dish, which is then heated to redness in a platinum crucible. These basins should not exceed one thirty-second of an inch in thick- ness, and the proper consistence of the clay is soon ascertained: if the edges of two of these little vessels be ground with a file, one may serve as a cover to the other. The steel mortar, Fig. 29, is an apparatus of great use to the blowpipe and mineral analyst. It consists of three separate parts: the lower portion is a shallow dish of steel into which a massive hollow hemispherical PABT I. 34 QUALITATIVE ANALYSIS. cylinder, also of steel, is accurately fitted by grinding; the upper portion is a solid cylinder of the same metal, which exactly fills up the hollow cylinder. When a mineral has to be crushed, it is introduced into the bed of the mortar, the solid cylinder is then replaced and struck forcibly several times with a mallet, by which it is reduced to a coarse powder, and may afterwards be brought to an impalpable state by grinding in the agate mortar. EVAPORATION. 23. Evaporation is an operation to which the chemist is con- stantly resorting ; he has recourse to it for concentrating liquids previous to the application of certain tests ; for separa- ting volatile fluids from fixed substances ; for inducing crystal- lization ; or for obtaining in a solid form substances held in solution by water or other liquids. When the object is to retain the fluid evaporated, as well as the residue, the pro- cess is called distillation. The ordinary operation of evapora- tion is conducted in basins of earthenware, silver, or platinum : watch-glasses also occasionally form very useful vessels. The earthenware basins should be as thin as is consistent with strength, and should resist the action of acids and alkalies in solution ; the silver and platinum vessels should be provided with a projecting slip of metal to serve as a handle, whereby they may be held by a pair of pincers. In ordinary cases, the object sought by evaporation is attained by exposing the fluid to heat ; sometimes, however, it is effected by leaving it for a certain time in contact with the atmosphere at common tem- peratures, or in confined air kept dry by hygroscopic substances. In quantitative experiments the evaporating substance should never be allowed to enter into actual ebullition, as a loss would almost unavoidably be sustained by bubbling. It is frequently advisable, therefore, to apply heat through the medium of the water bath, which may be a copper basin about six inches in diameter and three inches deep, provided with a series of rings of different diameters to suit dishes of different sizes. While a liquid is evaporating, it is requisite carefully to protect it from dirt and dust : this is best done in the manner recom- mended by Fresenius, viz. by providing two small, thin, wooden hoops, one of which must be made to fit loosely in the other; a sheet of blotting-paper is spread over the smaller, and the larger pushed over it : an excellent cover is thus formed, which, OPERATIONS AND APPARATUS. 35 while it effectually guards the liquid, does not come into con- tact with it, and does not in the least impede or retard the process of evaporation. Sometimes it is necessary to evaporate fluids from solid substances contained in crucibles : in such cases the crucible should be supported in an inclined position, and the heat applied a little above the level of the liquid : an equal distribution of heat is thereby secured, and the loss of a portion of the sediment by spirting avoided. The same method may be adopted in the evaporation of solutions of those salts which have a tendency to effloresce and creep up the sides of the evaporating vessel : in the latter case the object may be also attained by slightly greasing the upper part of the sides of the dish. Very frequently in the course of an analysis large quantities of nitrates have to be evaporated down to perfect dryness : in such cases the fluids are first concentrated by evaporation over the naked fire, and afterwards transferred into smaller vessels, to be finished in the water or sand bath. The process of transferring liquid from one basin to another, so as to avoid the smallest loss, requires care and a steady hand ; the edge of the large dish should be slightly smeared with tallow, and the liquid poured down an inclined rod, as recommended above when treating of filtration. The evaporation of a mineral water from which carbonic acid or other gases are expelled by heat, should always be commenced in a flask loosely covered with a piece of bibulous paper. As the evaporation of a fluid in which a considerable sedi- ment is formed draws to a conclusion, the thick mixture re- quires to be treated with great care ; for, the circulation of the heat being now interrupted, the temperature frequently in- creases at the bottom of the vessel until above the boiling-point of the solution, and then the sudden evolution of small quan- tities of steam occasions a projection of portions of the sub- stance from the basin. Sometimes, indeed, when the solid residue dries hard, it forms a cake on the surface, underneath which steam accumulates, and suddenly explodes with a force not merely sufficient to disperse the contents of the vessel, but even to break it in pieces. In these cases, the substance should be continually stirred with a glass rod, by which the loss by spirting is prevented. DISTILLATION. 24. This operation is performed when it is desired to collect D 2 36 QUALITATIVE ANALYSIS. the evaporating substance. It must obviously, therefore, be conducted in an apparatus in which the vapour as it rises may be refrigerated, and again reduced to a liquid or solid state, and collected in a separate vessel. When the elastic fluid or vapour assumes on cooling the solid state, the process is called sublimation. The vessel in which almost every laboratory distillation is effected is the glass retort ; those made of hard green glass are the most serviceable, as they withstand the action of caustic alkalies ; it is difficult, however, to get them stoppered. Ee- torts made of hard German glass are now much used in this country ; their high price, though lately much reduced, is the only objection to be raised against them. Fig. 30. The general process of distillation is represented in Fig. 30. The bend of the retort is protected from the cooling influence of the air by a thick paper or cardboard cone, with a broad notch to admit the neck ; this interferes with a ready change of air at the top of the retort, and saves much heat which would otherwise escape at that part, from the condensation of the vapour within. The end of the neck of the retort is connected with a long tube, either by means of a cork or occasionally by OPERATIONS AND APPARATUS. 37 cement, or loosely, the junction being made tight by strips of moistened bladder or sheet caoutchouc, and the end of this tube enters the mouth of the vessel into which the distilled products are to be received. The receiver is plunged into a basin of cold water in which lumps of ice are floating. In order to liquefy the vapour rising from the retort as soon as possible, several folds of bibulous paper are wrapped round the neck, just below the bend ; and lower down, just above the point where the neck enters the adapter, a ring of tow is tied, the ends of which are caused to hang down four or five inches. Cold water is supplied to this paper from a large funnel con- taining a filter placed immediately above it in such a position that the drops of water shall fall on and spread over the paper, which will soon become saturated, and the water will run down to the tow, from which it will descend and be caught in a vessel placed beneath, and, provided it does not flow in too rapid a stream, none will enter the flask. In the distillation of many liquids, particularly such as con- tain alcohol, the vapour is frequently evolved with such difficulty as not only to endanger the sudden expulsion of part of the substance, but the safety even of the whole apparatus. A tranquil and regular evolution of vapour may always be ob- tained by introducing into the retort certain angular solids on which the liquid exerts no chemical action, as fragments of glass, or, still better, slips of platinum foil ; but care must be taken not to introduce these promoters of evaporation while the fluid is hot, or the burst of vapour might probably be so instantaneous as to do more harm than the previous irregular boiling. The same precaution is applicable to the introduction of solids, of whatever nature, into liquids while boiling or near their boiling-points. Sulphuric acid, which is the most difficult and dangerous of all substances to distil, may be drawn over quietly and regularly by previously dropping into the retort a few pieces of platinum foil or wire. Various other contrivances have been resorted to for the purpose of condensing the vapour in the process of distillation. In the common still, this is well known to be accomplished by causing the vapours to pass through a long spiral tube, called a worm, fixed in a tub, and surrounded by cold water. In cases where the products are not easily condensed, Lielig's con- denser is a very useful and convenient arrangement, and is applicable in all cases where an open apparatus is admissible. 38 QUALITATIVE ANALYSIS. The condenser is a hollow metal cylinder, through the centre of which a glass tube passes, being fixed water-tight by means of perforated corks covered with cement. This tube is con- nected either by a cork or by tubing with the beak of a retort; a constant stream of water is caused to flow from the reservoir placed above, down the funnel, thus entering the cylinder at its lower part. The water, warmed by the condensation of the Fig. 31. vapour, flows out through the vertical tube underneath the upper part of the cylinder : the whole apparatus is thus kept constantly cool, and the distillation proceeds in a uniform and steady manner. The operation is represented in Pig. 31. It is sometimes required to conduct the gaseous products of distillation into vessels containing water or other liquids, as in the preparation of hydrochloric and sulphurous acids, solution of ammonia, etc. ; for such occasions the apparatus of Liebig would obviously be unfitted, and the arrangement represented in Fig. 32 is best adapted to the purpose. It consists of a OPERATIONS AND APPARATUS. 39 series of vessels placed side by side, and connected by tubes in such a manner that the tube originating in one vessel descends nearly to the bottom of the vessel following it. A simple in- spection of the figure will show the direction in which the current is supposed to be passing. After having acted on the water or solution in the first bottle, it passes through the bent tube into the second, and thence into the third. The junctions of the tubes with the bottles are made in various ways : some- times they pass through corks as in the figure, at other times Fig. 32. they are made tight by glazier's putty, or linseed paste, or plaster of Paris. To prevent stiffness and rigidity, the tubes leading from bottle to bottle may be made of separate pieces united together tight by tubes or collars of caoutchouc. This arrangement has received the name of * Woulfe's apparatus,' but it was first devised by Glauber. The upright tubes in the centre of each bottle are safety tubes, and are intended to admit air, when from any cause the pressure within is so far diminished as to be considerably less than that of the atmo- sphere. It is sometimes required to collect for examination the un- condensable elastic fluids evolved during distillation, besides the liquid results ; the object in such cases is readily attained 40 aUALlTATIVE ANALYSIS. by bending the delivering-tube of the third bottle into a curve, and bringing it under the shelf of the hydro-pneumatic or the mercuric-pneumatic trough ; when the operation is conducted on a small scale, the receiver may be made out of a piece of tube, bent in the manner represented in Pig. 33 ; the distillation is conducted precisely in the same manner as with an ordinary retort and receiver. The apparatus required for sublimation may be tubes, flasks, retorts, capsules, or cru- cibles. When no great heat is required, as with camphor^ naphthaline , io- dine , etc., the alembic may be used with advantage. For substances requiring a higher temperature, ** 33 ' such as calomel, cinnabar, etc., Florence flasks, bedded and heated in sand and provided with bent tubes, tightly luted, for conveying the volatile pro- ducts into other flasks for condensation, are exceedingly useful. Many substances may be conveniently and simply purified by placing them in a dish or pan, the mouth of which is covered with coarse filtering-paper perforated with holes, and over the whole a cap or cone of stiff paper, secured round the rim of the pan by twine or paste ; heat is then applied to the dish, the volatile substance rises and condenses upon the inner surface of the cap, its mechanical impurities being retained by the filtering-paper. 41 CHAPTER II. ON KEAGENTS. 25. THOSE substances which are employed by the chemist to give him information as to the nature of the subject of his examinations, have received the general name of Reagents ; though the manner in which they act, and the phenomena to which they give rise, are exceedingly varied. These bodies are of the highest importance to the analyst ; indeed the judicious use of them, and the correct interpretation of the appearances presented by their action, constitute the skill of the analytical chemist. We shall here describe the preparation and uses of the most important of these substances, without, however, making any attempt to classify them. The reagents to which we shall direct attention are the fol- lowing : 1. Blue litmus-paper. 2. Bed litmus-paper. 3. Turmeric paper. 4. Georgina paper. 5. Solution of indigo. 6. Starch paste. 7. Lead paper. 8. Sulphuric acid. 9. Nitric acid. 10. Hydrochloric acid. 11. JNitro-muriatic acid (aqua regia). 12. Acetic acid. 13. Oxalic acid. 14. Tartaric acid. 15. Hydrate of Potassa. 16. Carbonate of potassa. 17. Carbonate of soda. 18. Ammonia. 19. Sesquicarbonate of am- monia. 20. Chloride of ammonium. 21. Sulphide of ammonium (hydrosulphuret of am- monia) . 22. Sulphide of potassium. 23- Oxalate of ammonia. 24. Hydrosulphuric acid (sul- phuretted hydrogen) . 25. Ferrocyanide of potas- sium. 26. Ferridcyanide of potassium 42 QUALITATIVE ANALYSIS. 44. Acetate of lead. 45. Basic acetate of lead. 46. Protosulphate of iron. 47. Lime water. 48. Sulphate of lime. 49. Chloride of calcium. 50. Protochloride of tin. 51. Bichloride of platinum. 52. Sesquichloride of iron. 53. Sulphurous acid. 54. Hydrofluosilicic acid. 55. Chlorine water. 56. Chloride of mercury. 57. Subnitrate of mercury. 58. Molybdate of ammonia. 59. Protonitrate of cobalt. 60. Distilled water. 61. Alcohol. 62. Ether. 27. Chromate of potassa. 28. Sulphate of potassa. 29. Bisulphate of potassa. 30. Cream of tartar. 31. Cyanide of potassium. 32. Antimoniate of potassa. 33. Caustic baryta (solution) . 34. Chloride of barium. 35. Nitrate of baryta. 36. Phosphate of soda. 37. Phosphate of soda and ammonia (microcosmic salt). 38. Nitrate of potassa. 39. Biborate of soda (borax). 40. Nitrate of silver. 4 1 . Ammonio-nitrate of silver. 42. Sulphate of copper. 43. Iodide of potassium. (1.) Litmus Paper. This is an exceedingly delicate test of the presence of an acid ; it is most conveniently prepared by dipping thin unsized paper into an infusion of the colouring principle in hot water until it acquires a full blue colour. The paper is dried by exposure to the air, and kept carefully pro- tected from the light, which injures and finally destroys the colour. The blue colour of this paper is instantly changed red by contact with a fluid having an acid reaction. (2.) Red Litmus Paper. This is a valuable test of the pre- sence of an alkali. To prepare it, a few drops of hydrochloric acid are mixed with a large quantity of water, and the blue paper immersed in it until it becomes slightly reddened : it is then removed and dried for use. The blue colour is restored by contact with an alkali. (3.) Turmeric Paper. This is prepared in the same man- ner as litmus paper ; it should have a fine yellow colour ; it in- dicates the presence of an alkali by changing to a red brown. (4.) Qeorgina Paper. This, when properly prepared, is an excellent test of both acids and alkalies ; by the former it is coloured red, and by the latter green. It is prepared by dipping paper into the coloured infusion of the petals of the Georgina purpurea, and should have a fine violet colour. (5.) Solution of Indigo. Commercial indigo is digested in ON REAGENTS. 43 concentrated sulphuric acid, and the solution diluted with water till it is just distinctly blue. It is an excellent test for nitric acid, which, aided by heat, discharges the colour. (6.) Starch Paste. Common arrowroot starch is rubbed with cold water, and boiling water then added until a thin paste is formed. It is an invaluable test for free iodine ; when brought into contact with which, an intense blue compound is formed. (7.) Lead Paper. Paper is saturated with a strong solu- tion of basic acetate of lead, cut into strips and dried. It forms an extremely delicate test of the presence of sulphuretted hydrogen, which instantly communicates to it a deep brown- black colour. (8.) Sulphuric Acid, H 0, S O 3 . The commercial oil of vitriol always contains sulphate of lead, sometimes also nitric acid, arsenic, and tin. The first of these impurities is removed by diluting the acid with water ; a turbidity indicates sulphate of lead, which is insoluble in the diluted acid. Nitric acid is in- dicated by the blue colour of solution of indigo being dis- charged when boiled with the acid. Arsenic is indicated by passing a stream of sulphuretted hydrogen through the clear diluted acid ; a yellow precipitate is formed : if the precipitate be brown, it indicates tin. Prom all these impurities it may be freed by distillation ; the first portions being rejected, and not more than three-fourths of the acid in the retort drawn over ; for almost every qualitative operation the commercial acid may be employed. Sulphuric acid is of the most exten- sive use to the chemist, from its strong affinity for bases. It liberates most other acids from their combinations ; and, from its powerful affinity for water, it effects remarkable changes in many substances in which the elements of that fluid exist. It is a powerful oxidizing agent, and in a diluted state it serves as a test for barium, strontium, and lead. (9.) Nitric Acid, HO,N0 5 . This acid, the aqua-fortis of commerce, is prepared by distilling equal weights of oil of vitriol and nitre. It frequently contains sulphuric and hydro- chloric acids, from which it may be freed, by adding nitrate of silver as long as a precipitation takes place, and then redistil- ling. This operation may, however, be avoided, if in the ori- ginal preparation of the acid the first portions, about one- tenth or one-eighth of the whole, be collected in a separate receiver; these portions will contain all the impurities, and 44 QUALITATIVE ANALYSIS. the remainder will be quite pure. Nitric acid is used as a solvent for metals, sulphides, etc., and as a powerful oxidizing agent. (10.) Hydrochloric Acid, HC1. The muriatic acid of com- merce is not sufficiently pure for analytical purposes. It con- tains sulphuric acid and iron, sometimes also sulphurous acid, chlorine, and arsenic. It is best prepared by the following pro- cess of Gregory : Six parts by weight of pure common salt are introduced into a flask, and gently heated with a cool mixture of ten parts by weight of oil of vitriol and four parts of water. The gas is conducted into a flask containing a quantity of dis- tilled water, equal in weight to the salt, and surrounded with water in which lumps of ice are floating. The tube delivering the gas must dip about one-eighth of an inch into the water in the bottle ; the process takes about two hours, and the acid ob- tained is quite pure and colourless. Hydrochloric acid is very extensively used as a solvent, and for the detection of silver, mercury, and lead. (11.) Nitro-muriatic Acid, or AquaJZegia. This acid is pre- pared by adding nitric acid to twice or thrice its volume of strong hydrochloric acid ; both acids undergo decomposition, hyponitric acid, chlorine, and water being formed. When the liquid is saturated with chlorine, this mutual decomposition ceases ; but it recommences on the removal of the chlorine either by heat or by its combination with some other substance. Aqua regia is consequently the most powerful of solvents : its principal use in analytical chemistry is for dissolving gold and platinum, and for decomposing certain metallic sulphides. (12.) Acetic Acid, HO,C 4 H 3 3 . The acetic acid of com- merce frequently contains traces of sulphuric acid, but it may be obtained sufficiently pure for most analytical operations. If required quite free from all impurities, it is most conveni- ently prepared by distilling a mixture of ten parts of neutral acetate of lead with three of sulphate of soda, in a retort, with a cooled mixture of two and a half parts of sulphuric acid, and an equal weight of water : the distillation is continued to dry- ness. The acid thus obtained leaves no residue on evaporation, Acetic acid is employed as a solvent, and for acidifying li- quids in the place of the mineral acids. (13.) Oxalic Acid, HO,C 2 O 3 . The commercial acid is pu- rified by two or three recrystallizations. It should leave no re- sidue on ignition. It is employed as a precipitant of certain ON REAGENTS. 45 substances, particularly calcium, for the detection of which it is a very valuable reagent. All the oxalates are soluble in the stronger acids. (14.) Tartaric Acid, 2HO,C 4 H 8 10 . The commercial acid is sufficiently pure : well-defined crystals should be selected. It should be kept in powder, as its solution decomposes by keeping. It is employed to prevent the precipitation of cer- tain metallic oxides by alkalies, and as a test for potassium. (14.) Hydrate of Potassa, KO,HO. The best method of preparing this valuable reagent is to dissolve two parts of pure carbonate of potassa in twenty parts of boiling water in an iron pot, and to add in small portions at a time, to the boiling liquid, cream of lime (made by slaking one part of quicklime with boiling water ); after boiling a few minutes, the vessel is covered and allowed to stand for twenty-four hours ; the clear liquid is then decanted. To obtain the potassa in the solid state, the liquid is evaporated to an oily consistence in a silver basin, poured out on a silver dish, and allowed to cool ; it is then broken into fragments and preserved in well-stoppered bottles. Schubert recommends the following simple method of preparing a solution of pure potash : Add a hot solution of hydrate of baryta to a solution of sulphate of potassa until the liquid gives no further precipitate either with baryta or with sulphate of potassa. The uses of potassa in analytical chemistry are very numer- ous ; as a precipitant ; as a solvent ; as a means of separating certain oxides from others ; and as a test for ammonia, which, aided by heat, it expels from all its salts. (L6.) Carbonate of Potassa, KO,CO 2 . This salt is best prepared by calcining pure cream of tartar : the incinerated mass is boiled in distilled water, filtered, and the clear liquid evaporated to dryness in a clean iron vessel, with constant stirring towards the end of the process ; the dried mass must be kept in a well- stoppered bottle, and one part dissolved in five or six of distilled water for use. The carbonate of potassa of commerce usually contains alkaline sulphates and chlorides, alumina and silica. Carbonate of potassa is extensively em- ployed as a precipitant, and for the decomposition of many in- soluble salts, particularly organic, with metallic bases. (17.) Carbonate of Soda, NaO,C0 2 . This salt is obtained pure by heating the best bicarbonate of soda of commerce for some time to low redness : its uses are the same as those of 46 QUALITATIVE ANALYSIS. carbonate of potassa. It is an indispensable reagent in blow- pipe operations ; as a flux ; as a solvent ; and as a decompos- ing agent. Its solution should not be discoloured by sulphide of ammonium, when neutralized with hydrochloric acid it should give no turbidity with chloride of barium, and when evaporated to dryness with hydrochloric acid the residue should dissolve completely in water. (18.) Ammonia, NH 3 . Sal-ammoniac is mixed with an equal weight of slaked lime, a little water added, and the mixture heated in a stoppered retort. The disengaged gas is first al- lowed to pass through a small quantity of water in a wash-bot- tle, and from thence into another bottle nearly filled with dis- tilled water immersed in a vessel containing ice-cold water ; this bottle, for better security against sudden absorption, may be furnished with a safety tube. The water will absorb 670 times its bulk of the gas, and become possessed of all its che- mical properties in a very high degree. It should be kept in small well-stoppered bottles, and not in one large one, as every time it is exposed to the air it absorbs a certain quantity of carbonic acid, its freedom from which is proved by its not rendering lime-water turbid. Ammonia is in constant use for neutralizing acids, its peculiar fitness for which consists in its not introducing any fixed matter ; for precipitating insolu- ble bases ; and for separating them from each other. The am- monia of commerce is generally pure. It should give no tur- bidity with chloride of calcium, or with oxalate of ammonia, nor with nitrate of silver, when neutralized with nitric acid, and when evaporated should leave no residue. (19.) Sesquicarbonate of Ammonia, 2NH 4 0,3CO 2 . The sesquicarbonate of ammonia of commerce is dissolved in four parts of distilled water, and one of liquor of ammonia added. The solution when evaporated should leave no residue. This reagent is employed as a precipitant, and is very useful as a substitute for carbonate of potass, in cases where the intro- duction of a fixed base would be inconvenient. It is of special use in the separation of calcium, barium, and strontium from magnesium, the latter not being precipitated in the pre- sence of ammoniacal salts. (20.) Chloride of Ammonium, NH 4 C1. Sal-ammoniac of commerce is purified by two or three recrystallizations. Its solution in water should be neutral, and hydrosulphuret of ammonia should not discolour it : it should volatilize entirely ON REAGENTS. 47 when heated on platinum foil. The salt should be dissolved for use in eight parts of distilled water. It is of great use in analysis as a precipitant of various substances soluble in po- tassa, but insoluble in ammonia, and for keeping in solution certain oxides or salts when others are precipitated by ammonia or other reagents. (21.) Sulphide of Ammonium, NH 4 S (HydrosulpJiuret of Am- monia) . This reagent is prepared by transmitting sulphuretted hydrogen gas through solution of ammonia, till the liquid gives no precipitate with sulphate of magnesia. It must be kept in well-stoppered bottles Iree from lead. "When first prepared, it contains excess of sulphuretted hydrogen, is nearly colourless, and does not give a precipitate of sulphur when mixed with an acid ; but by exposure to the air it gradually absorbs oxygen, and assumes a yellow tint from the presence of excess of sulphur, of which element it now yields a precipitate on the addition of an acid. It is necessary to bear in mind these facts. Sulphide of ammonium is of great use for subdividing into two groups those metals which are precipitated as sulphides by sulphu- retted hydrogen, from their acid solutions ; some of these sulphides being soluble, others insoluble in sulphide of ammo- nium. It also subdivides into groups those metals that are not precipitated by sulphuretted hydrogen from their acid solutions ; some of these metals being precipitated by sulphide of ammonium, while others remain in solution : it likewise precipitates certain oxides as hydrates by the action of its am- monia alone, and certain salts that are dissolved only in free acids. (22.) Sulphide of Potassium, KS 5 . This reagent is easily prepared by transmitting a stream of sulphuretted hydrogen through a solution of caustic potassa as long as the gas is absorbed, and then mixing the saturated solution with an equal volume of the same caustic potassa. (23.) Oxalate of Ammonia, NH 4 0,C 2 3 . This reagent is prepared by slightly supersaturating a solution of pure oxalic acid with carbonate of ammonia, and crystallizing : one part of the salt is dissolved in twenty or twenty-four parts of water for use ; it is employed for the detection and precipitation of lime, and is more convenient than oxalic acid, as its solution does not decompose by keeping. (24.) Hydrosulphuric Acid, HS (Sulphuretted Hydrogen). Fragments of protosulphide of iron are covered with water in a 48 QUALITATIVE ANALYSIS. gas evolution apparatus connected with a wash-bottle : sulphuric acid is poured into the apparatus through a tube funnel, and the evolved gas is received into a bottle containing air-free dis- tilled water. The gas should not be allowed to pass too rapidly, as in that case much escapes absorption and passes into the at- mosphere, diffusing a most unpleasant odour, to guard against which the operations should be performed in a place that is pro- perly ventilated. The solution should be kept in well-stopped bottles. Hydrosulphuric acid is a reliable reagent for separating metals into groups, and also as a means of reduction. Fig. 34 represents the arrangement generally employed for passing a current of hydrosulphuric acid gas through a solution. a is a wide-mouthed bottle, containing water and small lumps of protosulphide of iron. It is closed with a good sound cork or tight-fit- ting caoutchouc cap, through which are inserted, perfectly air-tight, two tubes, one sur- mounted by a funnel, the other bent at right angles and joined by a connector of caoutchouc with a third tube also bent at right an- gles, and which passes to the bottom of the bottle, b, con- taining water. Through the cork with which this second bottle is closed, or through one of the apertures in its caoutchouc cap, a stout tube, c, passes, this tube is bent -pi 34 twice at right angles ; the delivering-tube, d, slightly curved at one end, is connected with the tube c by means of a good sound cork. AVhen this apparatus is about to be used for transmitting a stream of sulphuretted hydrogen gas through a liquid, the solution to be operated upon is placed in a tall jar, and the tube d being immersed in it, it is united by its cork with the tube c; sulphuric acid is now poured into d through the funnel, the gas is immediately evolved, and in passing through ON REAGENTS, 49 the water in b, which should be quite cold, it becomes washed, and thus passes in a state of purity into the solution. Some- times, when the evolution of gas is too rapid, portions of sul- phide of iron and of sulphuric acid are carried forward with it ; both are however retained in the wash-bottle Z>. When the solu- tion has become thoroughly saturated with the gas, the deliver- ing-tube d is removed, and, after being rinsed out, or, if necessary, cleaned with a small feather, it is fit for another operation. The constant use of this important reagent in the la- boratory renders it very de- sirable to have a supply of it always at command. Many forms of apparatus have been devised for this purpose, but that by Kipp, represented in Fig. 35, seems on the whole the best. It is thus described by Mr. Griffin, (' Chemical Recreations,' p. 617) : It is formed of glass, about six times the size of the figure ; that is to say, the largest globe is about six inches in diame- ter, and the whole apparatus is about eighteen inches high. The funnel-shaped neck of the uppermost globe is ground to fit the neck c air-tight, but it passes loosely through the neck , where a loose collar of caoutchouc is put about it to lgt prevent the falling of small lumps of sulphide of iron from the middle globe down to the lowermost globe. The middle globe is about half filled with lumps of sulphide of iron, in the largest pieces that will pass through the neck d, which is then to be closed with a sound cork carrying a glass stopcock, and that carrying a caoutchouc tube terminating with a glass gas-delivering tube. The charge of sulphide of iron being put in, the decomposing PAET I. E 50 QUALITATIVE ANALYSIS. acid, consisting of one part of oil of vitriol mixed with six or eight parts of water, is to be poured into the uppermost globe, from which it passes down into the lowermost globe, and thence up into the middle globe, the stopcock being opened to let out the atmospheric air. Hydrosulphuric acid gas is immediately gene- rated, and as much of it is allowed to escape as serves to sweep the atmospheric air completely out of the apparatus. The stop- cock is then closed, and the apparatus is ready for use. The quantity of gas delivered depends upon the arrangement of the stopcock. It can be delivered in single bubbles slowly, or in a rapid current. The delivery pipe is made short or long, accord- ing to the depth of liquor into which it is to pass. It must of Fig. 36. course be changed for every experiment. The siphon at the top of the apparatus must contain so much water as to allow atmospheric air to pass either way into or out of the apparatus, according as the stopcock is open or closed. If any sulphide of iron falls into the globe b, it makes a constant disengagement of gas, which, if not observed, may drive so much acid up into the uppermost globe as to cause an overflow to take place. It is proper, therefore, that this apparatus should always stand in a stonew r are pan, to catch any acid that may overflow. When the apparatus requires cleaning or the sulphide of iron needs ON REAGENTS. 51 washing, the acid can be poured off through the stoppered neck in the lowest globe. In Fig. 36 is represented Mohr's apparatus for a constant supply of a saturated solution of sulphuretted hydrogen.* It not only serves to prepare a solution of the gas without con- tact with atmospheric air, but to preserve it in that condition, and yet to afford a supply of the solution either in drops or in quantity, readily and conveniently. The Woulff s bottle, , is tilled with air-free distilled water ; b c is an apparatus for preparing hydrosulplmric acid gas ; the jar b contains dilute sulphuric acid ; the flask c contains lumps of sulphide of iron. The bottom of the flask is cut out, and replaced by a perforated plate of lead, upon which the sulphide of iron rests at about three-quarters of an inch from the cut bottom-edge of the flask. To set this apparatus in action, the vessels a, b, and c, having received their respective charges, the cork which fixes the siphon d in the bottle a is loosened, and the apparatus being then placed together, the gas is produced, and gradually drives out the atmospheric air from the vessels c and a. The siphon cork is then fixed, and sulphuretted hydrogen gas is produced in c and absorbed* by the water in , until the latter is saturated. The acid then descends from the flask c into the jar 5, and the operation stops. The siphon d is terminated outside by a glass delivery-tube attached by a caoutchouc con- nector and a pinchcock. Whenever the sulphuretted hydrogen water is required, the pinchcock is opened, a little liquor run off to waste to clean the end of the delivery tube, and then as much liquor is run out as the experiment requires. To supply the vacuum thus produced in the bottle a, the atmospheric air pressing upon the liquor in the jar b, drives it up into the flask c, and causes the production of as much hydros ulphuric acid gas as is required to make up the quantity run off by the siphon d. When the liquor of the bottle d is exhausted, a fresh supply of cold air-free distilled water is put into the bottle by the siphon neck, and the process of saturation goes on afresh. The corks of this apparatus should be coated with a mixture of fat and wax, to make them air-tight. (25.) Ferrocyanide of Potassium, K 3 EeCy 3 ; or K 2 Cfy. The commercial yellow prussiate of potassa is sufficiently pure for analytical purposes ; one part is dissolved for use in ten or * Mohr's ' Commentar zur Preussischen Pharmacopoeia,' and Griffin's ' Chemical Recreations.' E 2 52 QUALITATIVE ANALYSIS. twelve parts of water. It is of especial use for the detection of sesquioxide of iron and oxide of copper. (26.) Ferridcyanide of Potassium, K 3 Fe 2 Cy 6 ; or K 3 Cfdy. This reagent is prepared by transmitting a stream of chlorine gas through a solution of the above salt until it ceases to pro- duce a blue precipitate, with a solution of sesquichloride of iron. Its crystals have a magnificent red colour ; to procure them, the solution is concentrated by evaporation, and rendered feebly alkaline by carbonate of potassa. This reagent serves to detect protoxide of iron by the formation of a characteristic blue precipitate. (27.) Chromate of Potassa, KO,Cr0 3 . Bichromate of potash of commerce is dissolved in water, and carbonate of potash added till the solution reacts slightly alkaline ; from the con- centrated liquid yellow crystals may be obtained. It is em- ployed principally as a test for lead, with which it forms a pig- ment known as chrome yellow. (28.) Sulphate of Potassa, KO,SO S . The salt of commerce is purified by two or three crystallizations, and dissolved for use in ten or twelve parts of water ; it is used for the detection and separation of strontium and barium. (29.) Bisulphate of Potassa, K O, S O 3 ; H 0, S 3 . This is the fusible salt remaining when nitrate of potash is decomposed by two equivalents of oil of vitriol in the process for making nitric acid ; it is extensively employed in blowpipe operations ; in solution it indicates lithium, boracic acid, nitric acid, hydro- fluoric acid, bromine, and iodine, and separates oxides of ba- rium and strontium from other earths and metallic oxides. (300 Bitartrate of Potassa, KO,HO,C 8 H 4 O 10 ; or KO, HO,T. The cream of tartar of commerce is sufficiently pure ; it is useful in certain cases for separating metals from each other. (31.) Cyanide of Potassium, KCy. Eight parts of roasted ferrocyanide of potassium are fused at a bright red-heat in a covered crucible with three parts of dry carbonate of potassa ; the fused mass is poured carefully into a warm dish, and when cold, broken into fragments, and kept in a well-closed bottle : it must not be kept in solution, but dissolved as required in four or five parts of water. In analysis, its most important application is as a means of separating cobalt from nickel. As a blowpipe reagent mixed with an equal weight of carbonate of soda, it is exceedingly valuable from its powerful reducing ON REAGENTS. 53 action ; and from its easy fusibility it is of special application in the reduction of arsenic. (32.) Bimetantimoniate of Potassa, 2KO,SbO 5 . A mixture of one part of crude antimony with four parts of powdered nitre is thrown, a little at a time, into a crucible at a dull red- heat ; the mass is kept in a pasty state, with occasional stirring, for about half an hour, after which it is cooled, well washed, and heated to bright redness for half an hour, with two-thirds of its weight of pure carbonate of potassa. The cooled mass is digested with about fifty parts of warm water, and filtered for use when cold. It should not contain excess of alkali. Its use is as a test for soda, with which, provided no other base be present, it forms a very sparingly soluble crystalline precipitate. (33.) Caustic Baryta, BaO,HQ. Sulphide of barium is boiled with excess of oxide of copper or oxide of lead, and filtered when the liquid gives a white precipitate, with acetate of lead : it is then diluted with water, and preserved in well- closed bottles. Its most important use is as a precipitant of magnesia. (34.) Chloride of Barium, Bad. To prepare this useful re- agent, six parts of heavy spar (sulphate of baryta) are exposed to an intense red-heat, with a mixture of one part of powdered charcoal and one and a half of fluor spar; the resulting sulphide of barium is boiled with slight excess of hydrochloric acid, filtered, and crystallized two or three times. The solution of these crystals must be neutral to test-papers, not affected by sulphuretted hydrogen or sulphide of ammonium : it must, moreover, leave no residue when mixed with excess of sul- phuric acid, filtered and evaporated. Its most important use is as a means of detecting and estimating sulphuric acid. From the property which baryta possesses of forming soluble salts with some acids, and insoluble salts with others, it is like- wise a valuable reagent for distinguishing one group of acids from another. (35.) Nitrate of Baryta, BaO,NO 5 . Native carbonate of baryta is digested with dilute nitric acid, the solution filtered and crystallized two or three times. Its uses and applications are the same as those of chloride of barium. (36.) Plwsphate of Soda, 2JS T aO,HO,PO 5 . The commercial salt is crystallized and dissolved for use in ten or twelve parts of water. It serves as a test for alkaline earths in general, but 54 QUALITATIVE ANALYSIS. especially for the detection and estimation of magnesia, which it precipitates with the addition of ammonia as the basic phos- phate of ammonia and magnesia. (37.) Phosphate of Soda and Ammonia, NaO,]N"H 4 O,P0 5 . This salt (microcosmic salt) is prepared by boiling one hundred parts of crystallized phosphate of soda with sixteen of sal- ammoniac. Chloride of sodium separates, and the liquid, when filtered and evaporated, yields the double salt in fine crystals. When this salt is heated on charcoal or platinum wire, it loses water and ammonia, metaphosphate of soda being formed, which has the power of fusing a great number of chemical compounds. Hence its great use as a blowpipe reagent. (38.) Nitrate of Potassa, KO,JN"0 5 . The nitre of commerce is purified by repeated crystallizations ; its solution should give no precipitate with nitrate of silver, or chloride of barium : it is extensively employed as an oxidizing agent. (39.) Biborate of Soda, K"aO,2BO s The borax of com- merce is purified by recrystallization. It should be exposed to a gentle heat in a platinum crucible till it no longer swells up; it is then powdered and kept for use. When heated on the ring of platinum wire, it should give a clear trans- parent glass. This glass possesses the property of dissolv- ing most metallic oxides, the smallest portions of which com- municate to it a colour; hence its important use as a blow- pipe reagent. (40.) Nitrate of Silver, AgO,N0 5 . Standard silver is dis- solved in nitric acid, evaporated to dryness, and heated till all the copper present is converted into black oxide, which may be known by dissolving a portion of the fused salt in water and adding ammonia, which should not make the solution blue. The fused mass is dissolved in water, filtered and crystallized ; the crystals are dissolved for use in fifteen or twenty parts of dis- tilled water. It may be known to be pure by the filtrate from the precipitate, which it forms with excess of hydrochloric acid, leaving no residue when evaporated on a watch-glass. It is employed for arranging acids into groups, and is of special ap- plication in testing for, and estimating hydrochloric acid. (41.) Ammonio-Nitrate of Silver. Ammonia is dropped into solution of nitrate of silver till the precipitate which first forms is nearly redissolved. It is employed for the detection of arsenic. (42.) Sulphate of Copper, CuO, S0 3 . Blue vitriol is purified ON REAGENTS. 55 by two or three crystallizations. A solution of one part of this salt mixed with two and a quarter parts of protosulphate of iron, is employed for the precipitation of hydriodic acid, as pro- tiodide of copper. Ammonio-sulphate of copper, prepared in the same manner as the corresponding silver salt, is also em- ployed as a test for arsenic. (43.) Iodide of Potassium, KI. The commercial salt is tested for carbonate of potassa by treating it with hot alcohol, in which the latter salt is insoluble. It is a reagent for certain metals, particularly for lead and mercury, with which it forms charac- teristic precipitates. (44.) Neutral Acetate of Lead, PbO,C 4 H 3 O 3 . The best sugar of lead of commerce is dissolved in ten or twelve parts of distilled water ; it is useful for arranging acids into groups, and for the special detection of chromic acid. (45.) Basic Acetate of Lead, 3 PbO,C 4 H 3 O 3 . Seven parts of well-washed litharge and six of the best neutral acetate of lead are gently heated and agitated with thirty parts of water till the sediment has become perfectly white ; the liquid is then de- canted and preserved for use in a well-closed bottle. It has the same applications as the last- described salt, but its chief use is as a test for sulphuretted hydrogen. (46.) Protosulphate of Iron, EeO,S0 3 . Clean iron nails are digested with dilute sulphuric acid till hydrogen ceases to be evolved. The solution is filtered, and the crystals obtained washed with water slightly acidified with sulphuric acid and dried. This salt is a powerful deoxidizing agent, and is of es- pecial application as a test for nitric acid. It also precipitates gold in the metallic state, and forms a blue compound with ferridcyanide of potassium. (47.) Lime Water, CaO,HO. Fresh slaked lime is agitated with cold water, allowed to settle, and the clear fluid preserved in well-stopped bottles. It serves to detect carbonic acid, and as a means of distinguishing certain organic acids. (48.) Sulphate of Lime, CaO,S(> 3 . The precipitate formed on adding chloride of calcium to dilute sulphuric acid is well washed, digested, and agitated with water, and the fluid fil- tered for use. It is employed to distinguish between calcium, strontium, and barium. (49.) Chloride of Calcium, CaCl. Pure carbonate of lime is dissolved in dilute hydrochloric acid, the solution evaporated to perfect dryness, and the residue redissolved in distilled 56 QUALITATIVE ANALYSIS. water. It must be perfectly neutral. It is of great use for the classification of organic acids. (50.) ProtocUoride of Tin, SnCL Granulated tin is boiled with concentrated hydrochloric acid, the metal being in excess ; it is then diluted with four or five times its quantity of water, slightly acidulated with hydrochloric acid, and filtered. It must be kept in well-closed bottles containing fragments of metallic tin to prevent the protochloride from passing into the state of perchloride. It is a powerful reducing agent. It also serves to detect mercury, and, when mixed with nitric acid, it indi- cates the presence of gold. (51.) Bichloride of Platinum, Pt C1 2 . The solution of the metal in aqua regia is evaporated to dryness on the water-bath, and redissolved in eight or ten parts of water. It is of great use in analytical chemistry for the detection and estimation of potassium and ammonium. (52.) Sesquichloride of Iron, Fe 2 Cl 3 . Clean iron nails are digested with diluted hydrochloric acid ; the decanted acid li- quid is then boiled with successive additions of nitric acid, in a capacious vessel, till all effervescence ceases, and till it no longer tinges solution of ferridcyanide of potassium blue ; it is then precipitated with excess of ammonia, and the well- washed hydrated peroxide of iron is heated with hydrochloric acid, care being taken that it is not all dissolved, it being necessary that the test should not contain excess of acid ; it is then filtered for use. It is employed as a means of classifying organic acids, and is also of great use in the analysis of the phosphates of the alkaline earths. (53.) Sulphurous Acid, S0 2 . This is prepared by transmit- ting the gases produced by the action of six parts of oil of vitriol on one part of charcoal (carbonic and sulphurous acid gases) through ice-cold water till no more is absorbed. It must be kept in well-closed bottles, and should always smell strongly of the acid. It is a powerful means of reduction ; it precipitates mercury from its solution, converts chromic acid into oxide of chromium, arsenic acid into arsenious acid, etc. (54.) Hydrqfluosilicic Acid, HF,SiF 2 . Equal weights of a mixture of powdered fluor spar and quartz are gently heated, in a retort, with oil of vitriol, and the gas evolved passed into water, the extremity of the delivering tube dipping into mercury placed at the bottom of the jar, in ordor to prevent the tube from becoming choked up with the silicic acid which is ON REAGENTS. 57 precipitated the instant the gas comes into contact with water. The gelatinous mass is filtered through linen, and the filtrate preserved for use. Hydrofluosilicic acid forms an insoluble compound with potassium, which it is sometimes employed to separate from chloric acid. It is also used to discriminate be- tween barium and strontium, with the former of which it forms a crystalline precipitate. (55.) Chlorine Water. The gas evolved by heating finely- powdered peroxide of manganese with five or six times its weight of hydrochloric acid is conducted into cold water, until the fluid is saturated. It must be kept in a well-closed bottle and preserved from the light. It is employed to expel iodine and bromine from their combinations. (56.) Chloride of Mercury, HgCl. The corrosive sublimate of commerce is purified by crystallization, and dissolved for use in twelve or fourteen parts of water. It forms characteristic coloured precipitates with certain acids. (57.) Subnitrate of Mercury, Hg 2 O,N0 5 . One ounce of pure nitric acid is poured on one ounce of mercury and allowed to remain for twenty-four hours in the cold ; the crystals formed are dissolved in water acidified by nitric acid, and filtered. A small quantity of metallic mercury should be put into the bot- tle in which this reageut is preserved. (58.) Molybdate of Ammonia. This reagent, which is used as a test for phosphoric acid, is prepared by roasting sulphide of molybdenum until it ceases to liberate sulphurous acid, and be- comes completely converted into molybdic acid, which while hot is yellow, but white when cold. It is digested with ammonia, filtered, and then mixed with hydrochloric acid in quantity suf- ficient to redissolve the precipitate which at first forms. The solution should be perfectly colourless ; if it has a yellow tinge, it indicates the presence of phosphoric acid. (59.) Protonitrate of Cobalt, CoO,NO 5 .- It is not easy to obtain this reagent quite pure, though for blowpipe experiments it is a matter of great consequence that it should be so. Fre- senius gives the following directions for preparing it : an in- timate mixture of two parts of very finely-powdered cobalt, four parts of saltpetre, one part of effloresced carbonate of soda, and one part of dry carbonate of potassa, is projected in small portions into a red-hot crucible, which is then exposed to the strongest possible heat till the mass is fusing ; when cold it is reduced to powder, boiled with water, and the well- washed 58 QUALITATIVE ANALYSIS. mass dissolved in hydrochloric acid; the gelatinous mass is carefully evaporated to drynesa ; the residue boiled with water, filtered, and carbonate of ammonia (carbonate of potassa is better) added to the filtrate while kept at the boiling heat till all acid reaction ceases ; the filtered solution is precipitated by carbonate of potassa, and the precipitate obtained washed and dissolved in nitric acid. The solution is evaporated to dryness at a gentle heat, and one part of the residue dissolved in ten parts of water for use. Solution of nitrate of cobalt is em- ployed to distinguish certain metals in the oxidating flame of the blowpipe. Thus, alumina acquires a beautiful pale blue colour, magnesia a rose-red tint, and oxide of zinc a bright green. A few drops of the solution are placed on the sub- stance to be operated upon by means of a platinum wire or dropping tube. (60.) Distilled Water. No other water should be employed in the laboratory. It should give no precipitate or even tur- bidity with chloride of barium, nitrate of silver, oxalate of ammonia, or lime water, and should leave no residue on eva- poration. (61.) Alcohol, C 4 H 6 2 . Rectified spirits of wine, sp. gr. about '840, are sufficiently strong for most purposes. What is termed absolute alcohol, and which is sometimes required, is prepared by adding carbonate of potassa, that has recently been exposed to a red-heat, to ordinary alcohol until it ceases to dis- solve any more ; the whole is allowed to digest for twenty-four hours ; the liquid is then poured off, mixed with a sufficient quantity of quicklime to absorb the whole, and slowly distilled from a retort on a water bath at the temperature of about 180 ; it is then obtained of a sp. gr. of '7947. It must not redden blue litmus-paper, and must volatilize without leaving any residue. (62.) Ether, C 4 H 5 0. Sulphuric ether of commerce is suffi- ciently strong and pure for all purposes. In inorganic analysis it is employed to detect and isolate bromine. The student is recommended to assure himself by careful testing of the purity of his reagents ; by so doing he may not only save himself from much subsequent embarrassment, but he will be gaining much valuable experience in qualitative exa- minations. 59 CHAPTER III. ON THE COMPORTMENT OF THE PRINCIPAL METALLIC OXIDES WITH REAGENTS. 26. BY means of certain reagents this extensive class of compounds may be arranged into a series of Groups ; some of which, again, by other reagents, may be subdivided into Sec- tions, the whole forming a very convenient classification : Group I. Metallic oxides not precipitated from their solu- tions by liydrosulphnric acid, by sulphide of ammonium, or by alkaline carbonates. The Alkalies proper, viz. oxide of potassium (potassa), oxide of sodium (soda), oxide of lithium (lithia), and oxide of ammonium (ammonia). Group II. Metallic oxides not precipitated from their solu- tions by hydrosulphuric acid, but precipitated under certain circumstances, as salts, by sulphide of ammonium, and preci- pitated by the alkaline carbonates as carbonates. The Alkaline Earths, viz. oxide of barium (baryta), oxide of strontium (strontia), oxide of calcium (lime), and oxide of magnesium (magnesia).* Group III. Metallic oxides not precipitated by hydrosul- phuric acid, but precipitated, as oxides, by sulphide of am- monium. Oxides of aluminum, glucinum, chromium, thorinum, yttrium, cerium, zirconium, titanium, and tantalum. Group IV. Metallic oxides not precipitated from their acid * Oxide of magnesium is not precipitated by alkaline carbonates in the presence of ammoniacal salts ; but it is thrown down under these circum- stances by phosphate of soda. 60 QUALITATIVE ANALYSIS. solutions by hydrosulphuric acid, but precipitated, as sul- phides, by sulphide of ammonium : Oxides of zinc, nickel, cobalt, and iron, protoxide of manganese, and sesquioxide of uranium. Group V. Section A. Metallic oxides precipitated from their solu- tions, whether acid, alkaline, or neutral, by hydrosulphuric acid. Oxides of lead, silver, mercury, bismuth, cadmium, copper, palladium, rhodium, and osmium. This section may be further divided into two sub- sections by the comportment of its members with hy- drochloric acid ; oxide of silver, and suboxide of mercury, being completely, and oxide of lead partially , precipitated by that reagent. Section B. Metallic oxides precipitated from their acid solutions by hydrosulphuric acid, but not precipitated by that reagent from their alkaline solutions, their sulphides being soluble in alkaline sulphides. Oxides of antimony, arsenic, tin, platinum, iridium, gold, selenium, tellurium, tungsten, vanadium, and mo- lybdenum. A synoptical view of the comportment of the principal metallic oxides with the general reagents, and the colours of the precipitates occasioned thereby, is given in the following Table. ill 3 "tS '3 rrt -li "5 '3 .& 3^,3 ^ 5 ,c! -Q JE ** IJS SI i's il O ^ Vo 3 Ml Ma > |||ipj ! 1 iflMltl 1 s I S3iil*| iftfs^sl Bnouiran^oA mil % 6 62 QUALITATIVE ANALYSIS. GROUP I. The Alkalies proper: Potassa, Soda, Lithia, and Ammonia. 27. POTASSA. (KO.) General characters of Hydrate of Potassa (K 0, H 0). When pure, it is quite white, dissolving in water with the disengage- ment of heat, and attracting both water and carbonic acid from the atmosphere. It is highly caustic and eminently alkaline. Nearly all its salts being soluble in water, it is capable of being precipitated by a very few reagents. Its presence is however evinced by the following. (Solution of Chloride of Potassium (KC1) may be used.) Bichloride of Platinum (PtCl 2 ) produces a bright yellow crystalline precipitate of double chloride of platinum and po- tassium (KC1, PtCl 2 ). It is very sparingly soluble in water, and its formation is promoted by the presence of free hydro- chloric acid. It is quite insoluble in strong alcohol. Previous to applying this test, the operator must assure himself of the absence of ammonia, and the solution should be concentrated. Tartaric acid (2HO,C 8 H 4 10 or 2HO,T)added in excess produces a crystalline precipitate KCl + 2HO,i;=KO,HO,f+HCl soluble in strong acids and in carbonated and caustic alkalies, but insoluble in tartaric and acetic acids : the formation of this salt is greatly facilitated by agitation. Caroazotic or trinitrophenic acid, (HO,C P f ^rf\\ ^) ^is- ^1M U 4 _J 3 solved in alcohol produces a bright yellow crystalline precipi- tate (KO,C 12 H 2 (JS T O 4 ) 3 0), which is freely soluble in boiling water, but very sparingly so in cold. Perchloric acid (C1O 7 ) produces a sparingly soluble white crystalline precipitate (KO,C10 7 ). Before the blowpipe salts of potassa, if free from soda, heated on a platinum wire in the inner flame, tinge the outer flame violet. If soda salts be present, potassa may be detected by fusing a clear bead of borax with oxide of nickel, and then adding the mixture ; the brown colour of the bead is changed to blue. Characteristic. The reaction with bichloride of platinum, and the violet coloration of the blowpipe flame. METALLIC OXIDES. 63 28. SODA. (NaO.) General characters of Hydrate of Soda (NaO, HO). They are very similar to those of potassa. Its soluble salts afford no precipitates with any of the above reagents for potassa ; but Bi- metantimoniate of potassa (2 K 0, Sb 5 ) produces, in neutral and even in very dilute solutions, if well agitated, and provided no other oxide but potassa be present, a white crystalline precipitate. (Solution of Sulphate of Soda or of Chloride of Sodium may be used.) 2(NaO,SO s )+2KO,Sb0 6 =2(KO,Sq 8 )4-2NaO,SbO B . Before the blowpipe soda salts are distinguished by the strong yellow colour which they communicate to the outer flame, which reaction is not prevented by a very considerable excess of potassa. According to Kobell, one part of chloride of sodium may hereby be detected in twenty-five or thirty parts of chloride of potassium. 29. LITIIIA. (LiO.) General characters. "When pure this oxide is white. It is not so soluble in water as potassa or soda. Its solution rapidly absorbs carbonic acid when exposed to the air. Carbonate of lithia is very sparingly soluble in water ; when a solution of a salt of Hthia is mixed with one of phosphate of soda, the double phosphate of soda and lithia of very sparing solubility is precipitated. Phosphate of soda is therefore the character- istic test for this alkali. Before the blowpipe salts of lithia are detected by the fine crimson tinge which they communicate to the outer flame, when heated in the inner flame on platinum wire. Potassa salts do not interfere with this reaction; but soda salts destroy it, substituting for the crimson their own peculiar yellow colour. 30. AMMONIA (OXIDE OF AMMONIUM). (NH 4 O.) General characters. The solution of pure ammonia in water is, when concentrated, highly caustic and alkaline. It has a powerful and penetrating smell, by which its presence can generally be detected. It attracts carbonic acid from the at- mosphere. Most of its salts are soluble in water, and nearly all of them are totally volatilizable by heat. When present in an uncombined state, in quantity too small to be detected by the smell, its presence may be evinced by the production of white clouds when a feather dipped in strong acetic acid is held over the liquid. In salts of ammonia the compound 64 QUALITATIVE ANALYSIS. metal ammonium (NH 4 )may be regarded as replacing the sim- ple inetals in other salts, thus : Hydrochlorate of Ammonia. Chloride of Ammonium. NH 3 HO,SO 3 = NH 4 0,SO 3 Sulphate of Ammonia. Sulphate of Ammonium. Comportment of Ammoniacal Salts with reagents. Bichloride of platinum produces a yellow crystalline precipi- tate (NH 4 Cl,PtCl 2 ), having a great resemblance to the corre- sponding double salt of potassium. Tartaric acid produces in concentrated solutions a crystal- line precipitate (NH 4 0, HO, T) much more soluble than the corresponding potassa salt. If to a solution of corrosive sublimate (HgCl) a sufficient quantity of iodide of potassium be added to redissolve the pre- cipitate which is at first formed, and then considerable ex- cess of caustic potassa, a solution is obtained which will de- tect one drop of chloride of ammonium in a pint of water, by the production of a yellowish turbidity. The reaction is con- sidered to be expressed in the following equation : 4(HgI,KO)+3(KO,HO,)+NH 8 = GROUP II. The Alkaline Earths : Baryta, Strontia, Lime, and Magnesia. 31. BAEYTA. (BaO,HO.) General characters. When pure, it is of a greyish-white colour ; it combines with water with the evolution of great heat, and is completely dissolved. Its concentrated aqueous solution deposits crystals. It is powerfully caustic and alkaline. It combines greedily with carbonic acid, forming a white insolu- ble compound, which is poisonous. Comportment of soluble Barytic Salts with reagents. (Solution of Chloride of Barium may be used.) Alkaline carbonates produce a white precipitate (BaO,C0 2 ) soluble with effervescence ill hydrochloric acid. METALLIC OXIDES. 65 Ammonia and caustic alkalies do not produce any precipitate, provided atmospheric air be excluded. Sulphuric acid and soluble sulphates produce, even in very dilute solutions, a white precipitate (BaO,S0 3 ), quite insoluble in water, and very sparingly so in hydrochloric acid. Sydroflwosilicic acid (HF,SiF 2 ) produces after a time a colourless crystalline precipitate, almost entirely insoluble in free acids. BaCl + HF,SiF 3 = BaF,SiF 2 + HC1. Common phosphate of soda (2NaO,HO,P0 5 ) produces a white precipitate soluble in free acids. Oxalic acid (HO,C 2 O 3 + aq) and soluble oxalates produce, in tolerably concentrated solutions, a white precipitate (BaO,C 2 3 + aq) soluble in free acids; the formation of this precipitate is favoured by ammonia. Chromate of potassa (KOCrO 3 ) produces a yellow precipi- tate (BaOCrOg), very sparingly soluble in nitric and hydro- chloric acids. Before the blowpipe, salts of barium, particularly the chloride, when heated on platinum wire at the point of the blue flame, communicate to the external flame a pale apple-green colour, which is not suppressed by the presence of salts of strontium, calcium, or magnesium, unless they greatly predominate. Characteristic. The reactions with sulphuric and hydrofluo- silicic acids. 32. STEONTIA. (SrO.) General characters. It greatly resembles baryta ; but it is not so heavy ; neither is its hydrate so soluble in water ; its aqueous solution is consequently less caustic. Comportment of soluble Salts of Strontium with reagents. (Solution of Chloride or Nitrate of Strontium may be used.) Alkaline carbonates, the caustic alkalies, and phosphate of soda behave towards solutions of strontium salts precisely as towards solutions of salts of barium. Sulphuric acid produces a white precipitate (SrO,S0 3 ) not altogether insoluble in water; in very dilute solutions, there- fore, sulphate of lime and other soluble sulphates do not occa- sion an immediate precipitate. Hydrofluosilicic acid occasions no precipitate even in con- centrated solutions. PART i. F 66 QUALITATIVE ANALYSIS. Chromate of potassa in cold and dilute solutions produces no precipitate ; but, by boiling, a copious yellow precipitate (SrO,CrO 3 ) is determined. Oxalic acid produces a white precipitate (Sr 0,C 2 3 -f- aq) even in dilute neutral solutions ; in very dilute solutions the preci- pitate is immediately determined by the addition of ammonia. Before the blowpipe, sulphate of strontia fuses to an opales- cent mass, and colours the outer flame carmine-red. Chloride of strontium, heated on the ring of platinum wire at the apex of the blue flame, tinges the whole flame immediately deep-crim- son ; but as the assay fuses the colour disappears, by which it is distinguished from chloride of lithium. The presence of chlo- ride of barium prevents the production of the coloured flame. Soluble salts of strontium, digested with alcohol, and inflamed, give rise to an intense and characteristic carmine-red colour. Characteristic. The red coloration of alcohol flame ; the non- precipitation by hydrofluosilicic acid ; and the gradual precipi- tation by sulphate of lime. 33. LIME. (CaO.) "When pure, it is white and infusible. It has an acrid, caustic, alkaline taste. It has a powerful affinity for water, in combin- ing with which it emits great heat, and falls into a bulky powder. The hydrate of lime is far less soluble in water than the hydrates of the two preceding oxides, one part requiring for a perfect solution from 450 to 500 parts of water. The solution is slightly caustic, and gradually absorbs carbonic acid from the atmosphere, until the whole of the lime is converted into carbonate. Comportment of solutions of Salts of Calcium ivith reagents. (Solution of Chloride of Calcium may be used.) The caustic and carbonated alkalies, and phosphate of soda, behave with calcareous solutions precisely as with solutions ot barium and strontium. Sulphuric acid and the soluble sulphates occasion no precipi- tate in very dilute solutions ; but, on the addition of alcohol, a precipitate (CaO,SO 3 ) immediately takes place: in concen- trated solutions a bulky precipitate is produced, soluble, though not remarkably so, in nitric and hydrochloric acids. Hydrofluosilicic acid does not produce any precipitate in solutions of calcareous salts. METALLIC OXIDES. 67 Oxalic acid and the soluble oxalates occasion an immediate precipitate (CaO,C 2 3 ) in neutral solutions, soluble in the mi- neral acids, but insoluble in acetic acid. CaCl-f NH 4 0, C 2 p 3 =CaO, C 2 O 3 + NH 4 C1. The formation of this precipitate is increased and quickened by the addition of ammonia. Before the blowpipe, chloride of calcium, unless it has been fused, heated on the ring of the platinum wire, tinges the outer flame red, but the colour is more feeble than with chloride of strontium. Pure lime and the carbonate emit a very strong light. Soluble lime salts impart a yellowish-red tinge to the flame of alcohol. Characteristic. The reactions with oxalic and sulphuric acids. 34. MAGNESIA. (MgO.) General characters. It is a white infusible powder, possessed of a feeble but distinct alkaline reaction. Like lime, it is more soluble in cold than in hot water : 36,000 parts of boiling water, and 5142 parts at 32, being required to dissolve one part of the earth. Caustic magnesia does not emit any heat on being moistened with water. Behaviour of soluble Magnesium Salts ivith reagents. (Solution of Sulphate of Magnesia may be used.) Ammonia, in neutral solutions, occasions a white bulky pre- cipitate (MgO, HO). If the solution be acid, or if ammoniacal salts be present, no precipitate takes place, in consequence of the property possessed by magnesia of forming double salts with ammonia. Caustic potassa produces a voluminous flocculent precipitate, which disappears on the addition of chloride of ammonium ; but, on boiling with excess of potassa, the precipitate reappears, in consequence of the decomposition of the ammoniacal salt. Carbonate of soda produces in neutral solutions, and in the absence of ammoniacal salts, a white voluminous precipitate 2 (HO,MgO) +3 (MgO,CO 2 ), which is increased by boiling, in consequence of the expulsion of the carbonic acid, which, in the cold, keeps a portion of magnesia in solution. 5(MgO,S0 3 )+5(NaO,CO 2 )>4-2HO = 2(MgO,HO) +(3MgO,CO 2 )+2CO 2 . Carbonate of ammonia, by boiling, and in the absence of am- moniacal salts, occasions a slight precipitate. r 2 68 QUALITATIVE ANALYSIS. Sulphuric acid produces no precipitate (MgO,SO 3 ), being very soluble in water. Phosphate of soda alone does not produce any precipitate in very dilute solutions ; but, if ammonia be added, a crystalline precipitate of basic phosphate of magnesia and ammonia (2MgO,NH 4 0,P0 5 ) is formed in highly diluted solutions. 2 (MgO,S0 3 ) + 2NaO,HO,PO 5 + NH 3 = 2MgO,NH 4 0,P0 6 + 2(NaO,S0 8 ). This precipitate is insoluble in aminoniacal salts, but soluble in free acids. Oxalate of ammonia, in the absence of ammoniacal salts, forms a white precipitate (MgO,0 + 2aq). Before the blowpipe, in the absence of other metallic oxides, salts of magnesium, when ignited on charcoal, then moistened with solution of protonitrate of cobalt, and again strongly ignited, acquire a feeble red tint. Characteristic. The reactions with phosphate of soda and nitrate of cobalt. General Remarks on the Oxides of the Second Group. From the property of magnesia to form soluble double salts with ammonia, this oxide may be kept in solution by the addi- tion of chloride of ammonium and ammonia, while baryta, strontia, and lime are precipitated by carbonate of ammonia. The magnesia is detected in the filtered liquid by phosphate of soda. The immediate formation of a precipitate, on the addi- tion of sulphate of lime, and the gradual formation of a crys- talline precipitate on the addition of hydronuosilicic acid, are characteristic of baryta. Strontia, in combination with baryta, is detected by converting both earths into chlorides, and digest- ing with absolute alcohol, in which chloride of barium is almost insoluble. Chloride of strontium is detected in the alcoholic solution by the carmine-red flame it communicates to the alcohol when ignited ; and lime is detected by oxalate of am- monia. To discover the alkalies in the presence of the oxides of the second group, the baryta, strontia, and lime are first removed by boiling with carbonate of ammonia and caustic ammonia, and from the filtered liquid the magnesia is precipi- tated by baryta-water; the excess of baryta is removed by adding sulphuric acid in slight excess, and boiling ; the whole is then filtered, and the clear filtrate evaporated to dryness in a platinum dish, and ignited ; the residue is redissolved in METALLIC OXIDES. 69 water, and the solution tested for potassa, soda, and lithia, by- dividing it into three portions, and proceeding with each in the manner above directed for the discovery of the alkalies. A portion of the original solution is tested for ammonia, by heat- ing with caustic potassa, and applying a feather moistened with strong acetic acid to the mouth of the tube. GROUP III. Metallic Oxides not precipitated by Hydrosulphuric Acid, but precipitated as Oxides by Sulphide of Am- monium. (Oxides of Aluminum, Yttrium, Grlucinum, Thorinum, Zirconium, Chromium, Cerium, Titanium, and Tantalum.) 35. OXIDE or ALUMINUM (ALUMINA). (A1 2 O 3 .) General characters. When pure it is white, and in the state of powder it is light, and not at all compact. It has neither taste nor smell, but it adheres to the tongue, thereby occasion- ing a slight sense of astringency. By the heat produced by a stream of oxygen gas directed against the flame of a spirit-lamp, it slowly melts, and gives a limpid and colourless globule, which on cooling becomes crystalline. It is quite insoluble in water, although it possesses a powerful affinity for that fluid, from which it can only be deprived by heating to redness. It con- denses moisture from the atmosphere in a remarkable manner. The hydrate of alumina has a strong affinity for vegetable colouring principles. Comportment of soluble Aluminum Salts with reagents. (Solution of Sulphate of Alumina may be used.) Sulphide of ammonium produces a white voluminous preci- pitate of hydrate of alumina, soluble in potash, hydrosulphuric acid being evolved. Potassa and soda produce in neutral solutions a bulky pre- cipitate of hydrate of alumina (A1 2 3 3HO), entirely soluble in excess of the precipitant, but again precipitated by chloride of ammonium which destroys the solvent. 70 QUALITATIVE ANALYSIS. Ammonia occasions the same precipitate very sparingly so- luble in excess of the precipitant. The alkaline carbonates precipitate hydrate of alumina with the evolution of carbonic acid. Phosphate of soda produces a precipitate soluble in free acids and in potassa. Silicate of potassa (soluble glass) produces a precipitate (silicate of alumina) in solutions of this oxide in potassa, Before the blowpipe, alumina and many of its compounds may be detected by heating the assay on charcoal, then moist- ening it with solution of protonitrate of cobalt, and again heating it strongly in the oxidizing flame. A fine blue colour is produced. Characteristic. The reactions with sulphide of ammonium, potassa, and nitrate of cobalt. 36. OXIDE OF GLTJCINTJM. (G-1 3 3 .) General characters. This earth, when pure, has neither taste nor smell; it is insoluble in water, and infusible; but it does not harden in the fire like alumina, neither is the paste which it forms with water plastic. Its soluble salts have a sweet taste with a slight astringency. They do not yield an alum with sulphate of potassa. Comportment of Salts of Glucinum with reagents. Potassa and soda precipitate hydrate of glucina soluble in excess of the precipitants, but again precipitated by long boil- ing. Chloride of ammonium also precipitates the earth from its alkaline solution. Ammonia produces a voluminous precipitate, insoluble in excess of precipitant ; the presence of chloride of ammonium does not prevent the formation of this precipitate. The carbonated alkalies occasion a bulky precipitate, which is soluble in great excess of the precipitants, but more easily in carbonate of ammonia than in carbonate of potassa. Phosphate of soda produces a voluminous precipitate. Before the blowpipe, gluciua and its salts cannot well be de- tected ; they do not become blue when strongly heated with protonitrate of cobalt, like alumina. Sulphide of ammonium precipitates hydrate of glucina with the extrication of sulphuretted hydrogen. Glucina is distinguished from alumina by its reaction with METALLIC OXIDES. 71 alkaline carbonates ; by its salts not yielding alum with sul- phate of potash ; and by its not becoming blue when heated before the blowpipe with nitrate of cobalt. 37. OXIDE OF YTTRIUM. (YO.) General characters. This rare earth is of a pale-yellow co- lour. Its specific gravity is 4*842 ; it is therefore heavier than baryta, the specific gravity of which is 4'000. It is soluble in acids after ignition. It gradually absorbs carbonic acid from the atmosphere. Many of its salts, and especially the sulphate, have a faint amethyst-red colour, and a sweetish taste. Ac- cording to Mosander, three bases have been included under the name of yttria, one of which he calls erbia, and the other terbia. Comportment of Salts of Yttrium with their reagents. Potassa, soda, and ammonia produce white voluminous pre- cipitates, insoluble in excess of the precipitants, even by heat. The carbonated alkalies produce precipitates soluble in excess of the precipitants, particularly in carbonate of ammonia ; from the latter solution crystals of double carbonate of ammonia and yttria may be obtained. Sulphate of potassa produces, after a time, a precipitate, which is completely redissolved on the addition of water, even in the presence of sulphate of potassa. Phosphate of soda produces a precipitate soluble in hydro- chloric acid, from which it is again thrown down by boiling. Ferrocyanide of potassium (K 2 FeCy 3 ) occasions a white pre- cipitate. Before the blowpipe, yttria cannot with certainty be detected. According to Plattner, phosphate of yttria may be recognized by giving a regulus of phosphide of iron with boracie acid and iron, and from the difficulty with which it is dissolved by mi- crocosmic salt. 38. OXIDE OF THOBINUH. (ThO.) General characters. This rare earth is, when quite free from manganese, white. It is the heaviest of all the earths, its spe- cific gravity being 9'402. Its solutions have an astringent taste ; it absorbs carbonic acid from the air. When moist, the hydrate dissolves readily in acids ; but after having been dried, it is acted on with difficulty. The calcined earth is only at- tacked by hot sulphuric acid. Sulphate of thorina is, accord- 72 QUALITATIVE ANALYSIS. ing to Berzelius, distinguished from all other oxidized bodies known, by its property of being precipitated by boiling, and slowly redissolving on cooling. Behaviour of solutions of Oxide of Thorinum with reagents. Potassa and ammonia produce a quickly subsiding precipitate, insoluble in excess of the precipitants. The carbonated alkalies produce a precipitate dissolving rea- dily in excess of the precipitants. Sulphate ofpotassa produces a double salt, insoluble in water containing sulphate of potassa. Ferrocyanide of potassium occasions a heavy white precipi- ate, soluble in acids. Before the blowpipe, the reactions of thoriua have not been studied. 39. OXIDE OF ZIRCONIUM. (Z 2 3 .) General characters. When pure and calcined, glucina is a white infusible powder; when ignited, it becomes brilliantly incandescent : it is sufficiently hard to scratch glass. Its spe- cific gravity is 4'3. After having been ignited, it is soluble only in concentrated sulphuric acid. Its soluble salts have a purely astringent taste, without any sweetness. Comportment of Salts of Oxide of Zirconium with reagents. Potassa and ammonia produce precipitates insoluble in excess of the precipitants. The carbonated alkalies produce precipitates slightly soluble in excess of the precipitants ; the hydrate of zirconia is solu- ble in carbonate of ammonia, but very sparingly so in the car- bonates of the fixed alkalies. Ferrocyanide of potassium occasions a white precipitate. Sulphate of potash produces a white double salt, almost in- soluble in water. Before the blowpipe, the reactions of zirconia are similar to those of glucina, from which it is distinguished by the vivid white light which it emits when ignited. 40. SESQUIOXIDE or CHROMIUM. (Cr 2 O 3 .) General characters. After ignition it is of a fine green co- lour, and is only soluble in hot sulphuric acid : the hydrate is of a greyish-green colour, and is readily soluble in acids, form- ing green solutions under reflected, and red by transmitted, METALLIC OXIDES. , 73 light ; but if the hydrate has been strongly dried, but not ig- nited, it dissolves in acids with difficulty. Comportment of solutions of Salts of Chromium with reagents. (Solution of Sesquichloride of Chromium, Cr 2 Cl 3 , may be used.) Sulphide of ammonium produces a bluish-green precipitate of hydrate of sesquioxide of chromium, sulphuretted hydrogen being disengaged. O 2 C1 3 4- 3NH 4 S + 6 H O = Cr 3 O 3 ,3HO + 3NH 4 C1 + 3HS. Potassa produces a bluish-green precipitate (Cr 2 3 ,3HO), readily soluble in an excess of the precipitant, forming a green solution, from which the green anhydrous oxide is re-precipi- tated by boiling, either alone or with chloride of ammonium. Ammonia and carbonate of ammonia produce a bluish-green precipitate, partially soluble in the precipitant, to which it im- parts a red colour ; but the precipitation is complete by boiling the ammoniacal solution. Carbonate of potassa or soda produce a bluish-green precipi- tate, soluble completely in considerable excess of the precipi- tant, and not re-precipitated by boiling. Phosphate of soda produces a light green precipitate. Any compound of oxide of chromium when fused with nitre gives rise to the formation of chromate of potassa (KO,Cr0 3 ), which is soluble in water, and to which it communicates a yel- low colour. Before the blowpipe, the presence of oxide of chromium is easily detected by the beautiful green bead obtained when it is heated with borax or microcosmic salt, both in the inner and outer flame : oxide of copper gives also a green bead, but only in the outer flame. Characteristic. The colour of its salts ; their conversion into chromic acid ; and the reaction in the blowpipe flame. 41. TITANIC ACID. (TiO 2 .) General characters. It is a white, insipid, infusible powder ; when heated, it assumes a fine yellow colour, but again becomes colourless on cooling. It reddens infusion of turnsole even after having been exposed to a red-heat, though the calcination renders it insoluble in acids. Comportment of solutions of Titanic Acid with reagents. Ammonia precipitates a white gelatinous hydrate, soluble with great readiness in acids, and soluble also in small quantities 74 QUALITATIVE ANALYSIS. in the carbonated alkalies : it is precipitated from its solution in carbonate of ammonia by long boiling : its solution in carbonate of potassa or soda is precipitated by boiling with sal-ammoniac. Titanic acid is precipitated as a heavy white powder from its acid solutions by continued boiling : this precipitate cannot, however, be washed on a filter with pure water (Rose) : ac- cording to Berzelius, it can be completely precipitated from its solution in sulphuric acid by long-continued boiling. The caustic alkalies and sulphide of ammonium do not preci- pitate titanic acid in the presence of a sufficient quantity of tartaric acid. Ferrocyanide of potassium produces a red-brown precipitate. Sulphite of ammonia (NH 4 O,SO 2 ), aided by a gentle heat, completely precipitates titanic acid. If, in a solution of titanic acid, or an alkaline titanate, in hydrochloric acid, a bar of zinc, tin, or iron be placed, the titanic acid will be reduced to ses- quioxide of titanium (Ti 2 3 ), and the liquor will acquire a blue or violet colour ; after a time, however, a violet powder is pre- cipitated, and the liquid becomes colourless. This is charac- teristic of titanic acid. Before the blowpipe, pure titanic gives in the reduction flame, with microcosmic salt, a violet bead ; the reaction is observed better on adding metallic tin ; in the presence of peroxide of iron the glass appears, when strongly heated in the reducing flame, yellow, and on cooling, red ; with borax no alteration is produced by the presence of iron. General Remarks on the Oxides of the Third Group. Alumina and glucina are both dissolved readily by caustic potassa ; but the two earths are distinguished from each other by the latter being precipitated from its alkaline solution by boiling, and by its hydrate being soluble in carbonate of am- monia. Yttria, thorina, and zirconia are not soluble in caustic potassa ; yttria is distinguished from the other two earths by the double salt which it forms with sulphate of potassa being soluble in solution of sulphate of potassa ; whereas the double salts formed by thorina and zirconia are not soluble in sulphate of potassa. These two latter earths are not very easily distin- guished from each other ; but the precipitate produced by car- bonated alkalies in solutions of thorina is much more soluble in j carbonate of potassa than the corresponding precipitate in so- i lution of zirconia. Thorina is, moreover, more than double the i density of zirconia, which is distinguished again by the glaring METALLIC OXIDES. 75 white light which it produces when strongly ignited. The colour of the salts of chromium, and their behaviour before the blowpipe, is quite sufficient to distinguish this oxide from all the other members of the group. Titanic acid is distinguished by the blue colour produced on bringing a rod of zinc or iron into contact with its acid solutions. This group may be subdivided into two sections by the com- portment of its members with caustic potassa ; thus alumina, glaucina, and oxide of chromium, are soluble in this alkali : the other members are insoluble. GROUP IV. Metallic Oxides not precipitated from their acid solutions by Hydrosulphuric Acid, but completely precipitated by Sulphide of Ammonium as Sulphides. (Oxides of Zinc, Nickel, Cobalt, Manganese, Iron, and Uranium.) 42. OXIDE OF Zrac. (ZnO.) General characters. When pure it is white ; it becomes yellow when heated,but on cooling its whiteness usually returns, though sometimes it retains it yellow tinge. It is sometimes obtained crystalline, and is then always yellow. When the metal is burned in the air, the oxide is obtained of snowy whiteness, and in light flocks. In this state it has been called " philosophical wool." It has a remarkable affinity for alumina ; a combination of the two oxides in atomic proportions is met with in the mineral kingdom crystallized in regular octahedra, and is known under the name of Oahnite. Comportment of solutions of Salts of Zinc with reagents. (Solution of Sulphate of Zinc may be used.) Potassa and ammonia produce in neutral solutions a white gelatinous precipitate (ZnO, HO), readily soluble in an excess of the precipitants. Hydrosulphuric acid produces in neutral solutions a white precipitate (ZnS); in acid solutions no precipitate is formed. Sulphide of ammonium completely precipitates salts of zinc, as ZnS, insoluble in an excess of the precipitant, as well as in potassa and ammonia. The carbonates of the fixed alkalies precipitate basic carbonate 76 QUALITATIVE ANALYSIS. of zinc (3HO,ZnO 4-2ZnO,C0 3 ), quite insoluble in an ex- cess of the precipitants, but soluble in potassa and in ammonia. The presence of any salt of the latter prevents the formation of this basic salt, a soluble double salt of zinc and ammonia being formed. Carbonate of ammonia produces a white precipitate, soluble in an excess of the precipitant. Phosphate of soda produces a precipitate soluble in acids, and in potassa and in ammonia. Oxalic acid and binoxalate of potassa occasion precipitates which are soluble in acids and in fixed alkalies, but the forma- tion of which is not prevented by sal-ammoniac. Ferrocyanide of potassium produces a white gelatinous pre- cipitate, insoluble in free hydrochloric acid. Ferridcyanide of potassium produces a yellowish-red precipi- tate, soluble in hydrochloric acid. Before the blowpipe, zinc salts are easily detected. Heated with carbonate of soda on charcoal in the reducing flame me- tallic zinc is produced, which volatilizes, and on coining into contact with the air is again oxidized, and the charcoal becomes covered with a sublimate, which, when hot, is yellow, but on cooling white, and gives when heated in the oxidating flame with a few drops of protonitrate of cobalt a beautiful and cha- racteristic green colour. Characteristic. The reactions with potash and sulphuretted hydrogen, and before the blowpipe. 43. OXIDE OP NICKEL. (NiO.) General characters. The pure oxide is of a deep ash-grey colour ; it is not magnetic ; it dissolves readily in acids ; its hy- drate is of an apple-green colour. Comportment of solutions of Salts of Nickel with reagents. (Solution of Sulphate of Nickel may be used.) Potash produces a bright green precipitate, insoluble in an excess of the precipitant, but soluble in carbonate of ammonia. Ammonia precipitates the same green hydrate, but an excess redissolves it, forming a clear blue solution, from which potassa again precipitates the hydrate. Hydrosulphuric acid does not precipitate acid solutions of j nickel. In neutral solutions, after a time, an inconsiderable j black precipitate is formed ; but in the presence of an alkaline j METALLIC OXIDES. 77 acetate, aided by a gentle heat, sulphuretted hydrogen effects a complete precipitation. Sulphide of ammonium produces a black precipitate (NiS), par- tially soluble in an excess of the precipitant ; hence, after the subsidence of the precipitate, the fluid remains black. Alkaline carbonates produce a pale green precipitate, soluble in carbonate of ammonia. Phosphate of soda occasions a very pale yellow precipitate. Ferrocyanide of potassium produces a pale yellowish-green precipitate. Cyanide of potassium throws down a greenish-white precipi- tate (NiCy), which an excess redissolves into a brownish-yellow liquid (Ni Cy, K Cy + H Cl = Ni Cy + K 01 + H Cy) ; on the ad- dition of a mineral acid, cyanide of nickel is again precipi- tated, and hydrocyanic acid set free. Before the blowpipe, salts of nickel, heated in the outer flame with borax or microcosmic salt, give a reddish-coloured bead. The addition of nitre changes the colour to dark purple or blue : heated in the inner flame on charcoal with carbonate of soda, reduction takes place, and renders the bead grey. Characteristic. The reactions with ammonia and cyanide of potassium. 44. OXIDE OF COBALT. (CoO.) General characters. As obtained by the calcination of the carbonate, it is of an ash-grey colour ; as obtained by the com- bustion of the metal, it is blue or greyish-blue ; as precipitated from its solutions by caustic potassa, it has a fine blue colour. When this precipitate is boiled for some time, it assumes by degrees a violet, and finally a dirty red tinge, which, according to Proust, is the hydrate : the blue precipitate is considered by some chemists to be a basic salt. It dissolves by fusion with vitreous fluxes, communicating to them a magnificent blue colour, or if in excess, black, Comportment of solutions of Salts of Cobalt with reagents. Potassa produces a blue precipitate, insoluble in an excess of the precipitant, but soluble in carbonate of ammonia; the pre- cipitate becomes green by exposure to the air, and dingy-red when boiled. Ammonia produces a blue precipitate, which an excess redis- solves, forming a solution which is at first green, but which by 78 QUALITATIVE ANALYSIS. exposure to the air becomes brown ; if sal-ammoniac be present in sufficient quantity, neither potassa nor ammonia produce any precipitate, though, if air have access, the solution gradually becomes brown. The alkaline carbonates produce a red precipitate, which upon being boiled becomes blue. Phosphate of soda occasions a blue precipitate. Ferrocyanide of potassium gives a green precipitate, which gradually turns grey. Ferridcyanide of potassium gives a dark reddish-brown pre- cipitate. Hydrosulphuric acid in acid solutions occasions no precipi- tate : in neutral solutions, after a time, a slight black precipi- tate, the solution acquiring a dark colour. Cyanide of potassium throws down CoCy as a brownish white precipitate, easily soluble in excess of the precipitant ; on the addition of hydrochloric acid no precipitate takes place, the whole of the cobalt passing into the state of cobalto-cyanide of potassium; thus: 2CoCy + 3KCy + HCy = (Co 2 Cy 6 K 3 )+ H, which is not decomposable by weak acids. Sulphide of ammonium produces a black precipitate, insolu- ble in excess. Before the blowpipe, salts of cobalt are distinguished by the beautiful blue colour they communicate to borax and microcos- mic salt, in both oxidating and reducing flames. Heated with carbonate of soda in the reducing flame, a grey powder (metallic cobalt) is produced. Characteristic. The reactions with potash, ammonia, and cyanide of potassium, and before the blowpipe. 45. OXIDE OE MANGANESE. (MnO.) General characters. It is of a greyish-green colour. When prepared by igniting the carbonate or oxalate in an atmosphere of hydrogen, it absorbs oxygen from the air, gradually becom- ing brown. The oxide, prepared by fusing the chloride with anhydrous carbonate of soda, undergoes no alteration by ex- posure to the air. The hydrate when first precipitated is white, but on exposure to the air it gradually becomes brown. This oxide possesses, in common with magnesia and oxide of iron, the property of only being partially precipitated by ammonia, and of carrying with it a portion of silicic acid when precipi- METALLIC OXIDES. 79 tated from a liquid holding this acid in solution. The salts of oxide of manganese are sometimes colourless, and sometimes of a pale rose-colour. Comportment of solutions of Salts of Manganese with reagents. (Solution of Sulphate of Manganese may be used.) Potash and ammonia produce precipitates which at first are white, but soon become coloured, first yellow, then brown, and finally nearly black : if ammoniacal salts be present, ammonia occasions no precipitate, and potassa only a partial one, which slowly becomes brown by exposure to the air. A clear am- moniacal solution containing oxide of manganese gradually gets turbid by exposure to the air, brown hydrated sesquioxide of manganese being deposited. The carbonated alkalies produce a white precipitate of car- bonate of manganese, which is permanent in the air, and spa- ringly soluble in sal-ammoniac. Phosphate of soda produces a white precipitate persistent in the air. Ferrocyanide of potassium produces a pale red precipitate, soluble in free acids. Ferridcyanide of potassium gives a brown precipitate,insoluble in free acids. Hydrosulphuric acid does not precipitate either acid or neutral solutions. Sulphide of ammonium produces a flesh-red precipitate (Mn 8), insoluble in excess, but soluble in mineral acids and even in strong acetic acid ; by exposure to the air it becomes oxidized, finally assuming a brownish-black colour. Puce-coloured oxide of lead (binoxide), heated with pure dilute nitric acid, produces when added to a solution of a salt of manganese, a fine crimson colour (permanganic acid, Mn 2 O 7 HO). By fusing any salt of manganese on platinum foil with a mixture of nitre and carbonate of soda, mancjanate of soda of a fine green colour is produced ; the smallest portion of manga- nese may in this manner be detected. Before the blowpipe, fused in the oxidating flame with borax or microcosmic salt, salts of manganese give an amethyst- coloured bead, the colour of which disappears in the reducing flame, but may again be produced in the oxidating flame. Characteristic. The reactions with sulphide of ammonium and binoxide of lead, and with carbonate of soda with heat. 80 QUALITATIVE ANALYSIS. 46. PROTOXIDE OF IRON. (EeO.) General characters. This oxide, which is only obtained pure with extreme difficulty, is black, and has frequently a metallic lustre ; it is brittle ; fuses at a high temperature ; and on cool- ing is converted into a brittle, brilliant, but not vitreous mass. It is dissolved in acids with great difficulty, after having been exposed to a red-heat ; but the salts formed are identical with those obtained by dissolving the metal itself in the respective acids. It is very feebly magnetic, by which it is distinguished from the ferrosoferric oxide (FeO,Fe 2 O 3 ), which is strongly magnetic. It combines with water, forming a hydrate which, when pure, is white ; but by contact with the atmosphere it speedily becomes coloured ; first grey, then green, then bluish- black, and finally yellow. When boiled in an hermetically closed vessel, it parts with its water, and becomes black. It is the basis of all the protosalts of iron. Comportment of Salts of Protoxide of Iron with reagents. (Solution of Protosulphate of Iron may be used.) Potassa and ammonia produce a nocculent precipitate of hy- drated protoxide of iron, which at first is nearly white, but which readily becomes coloured by exposure to the air. The presence of ammoniacal salts prevents the precipitation of oxide of iron by ammonia, and in some degree by potassa. Alkaline carbonates produce a white carbonate, gradually becoming coloured, though not so readily as the oxide. It is soluble in sal-ammoniac, but a coloured precipitate makes its appearance by exposure to the air. Phosphate of soda produces a white precipitate, which after a time becomes green. Ferrocyanide of potassium produces a precipitate, which, if air be entirely excluded, is white (K,Fe 3 Cfy 2 ) ; but, if air or a small quantity of sesquioxide of iron be present, it has a blue tinge ; by exposure to the air, or by contact with nitric acid or chlorine, it absorbs oxygen, and gives rise to the formation of Prussian blue. The transformation will be understood from the following equations : (a.) 3FeCl + 2K 2 , (b.) 3(K,Fe 3 Cfy 2 )+40 Prussian blue is insoluble in hydrochloric acid ; its colour is discharged by the fixed caustic alkalies. METALLIC OXIDES. 81 Ferridcyanide of potassium (K 3 Cfy 2 ) produces a beautiful blue precipitate, insoluble in acids, but easily decomposable by alkalies. The composition of this substance is Cfy 2 F 3 , and it is thus formed : 3FeCl + K 3 Cfy 2 = 3KC1 + Cfy 2 Fe 3 (Turnbull's blue). Hydrosulphuric acid does not precipitate acid solutions of oxide of iron ; and neutral compounds very incompletely. Sulphide of ammonium produces a black precipitate, speedily becoming brown by exposure to the air. It is insoluble in alkalies and alkaline sulphides, but easily soluble in mineral acids. Before the blowpipe, protosalts of iron heated on charcoal with borax or microcosmic salt in the oxidating flame give dark- red beads, becoming lighter on cooling ; in the inner flame the colour produced is green, which disappears on cooling, if the metal be not present in too large quantity. When fused with soda on charcoal in the reducing flame, a metallic magnetic powder is obtained. Characteristic. The reactions with ferrocyanide andferrid- cyanide of potassium. 47. SESQUIOXTDE or IKON. (Fe 2 O 3 .) General characters. Its colour and physical appearance differ according to its mode of preparation. It is met with in nature of a grey colour, and crystalline ; as prepared by the calcination of the subsulphate of the sesquioxide, it has a fine red colour; from the sulphate its colour is deeper, and when made from the nitrate it is brownish-black ; and it is sometimes met with quite black. By the action of a high heat it is con- verted into ferrosoferric oxide (FeO,Fe 2 3 ) with the disengage- ment of oxygen gas. It does not dissolve very readily save in concentrated acids, after having been strongly heated, though much more easily than the protoxide. It is not easily preci- pitated from its solutions by means of an alkali or an earth. If too little alkali be added, a subsalt is thrown down ; if too much, a portion is precipitated with the oxide. When iron is oxidized by degrees in contact with a large quantity of water, a hydrate of the sesquioxide of a clear orange colour is formed. PART I. 82 QUALITATIVE ANALYSIS. Comportment of solutions of Salts of Sesquioxide of Iron with reagents. (Solution of Sesquichloride of Iron may be used.) Potassa and ammonia produce a voluminous reddish-brown hydrate, insoluble in excess. The precipitation is prevented by the presence of organic acids, sugar, etc. Fe 2 Cl 3 + 3NH 3 + 6HO = Fe a 8 ,3HO + 3NH 4 C1. The carbonated alkalies throw down precipitates of rather a lighter colour, carbonic acid being at the same time disengaged. Phosphate of soda produces a white precipitate of phosphate of sesquioxide of iron of varying composition, which becomes brown, and finally dissolves on the addition of ammonia. Ferrocyanide of potassium produces a beautiful blue precipi- tate (Fe 4 Cfy 3 ) by the following reaction : Sulphocyanide of potassium produces a rich blood-red colour, even in very dilute solutions (FeCyS 2 ). Hydrosulphuric acid produces in neutral solutions a white deposit of sulphur, the sesquioxide being reduced to protoxide thus: Sulphide of ammonium produces in neutral solutions a black precipitate, insoluble in excess, and becoming brown by expo- sure to the air. This precipitation is preceded by the reduction of the sesquioxide into protoxide. Before the blowpipe, salts of sesquioxide of iron behave in the same manner as those of the protoxide. Characteristic reactions. Those wiihferrocyanide andsulpho- cyanide of potassium. 48. SESQUIOXIDE OF URANIUM. (Ur 2 3 .)- General characters. The hydrate is a beautiful yellow powder, soluble in acids, forming fine yellow solutions ; heated to about 300, it loses its water and becomes anhydrous ; it then has a bright brick-red colour. Heated above that tem- perature, it loses oxygen and becomes converted into uranoso- uranic oxide (UrO,IJr 2 3 ) of a deep-grey colour. Sesquioxide of uranium reddens moistened turnsole paper, though it blues paper stained red with infusion of logwood. It produces, therefore, the reaction of acid and base. According to Berzelius, this substance should more properly be called uranic METALLIC OXIDES. 83 acid; its properties being rather those of an electro-negative than of an electro-positive oxide. It forms definite compounds with bases, all of which are insoluble in water ; and, when pre- cipitated from its solution in an acid by means of an alkali, the latter is divided into two portions, and the precipitate ob- tained is a uranate. Comportment of solutions of Salts of Sesquioxide of Uranium with (Solution of the Nitrate of the Sesquioxide of Uranium may be used.) Caustic alkalies precipitate uranates of the bases of a pale yellow colour. Alkaline carbonates produce pale yellow precipitates, soluble in excess of carbonate of ammonia, but again precipitated by boiling. Sulphite of ammonia produces at a boiling temperature a yellow precipitate. Ferrocyanide of potassium gives a reu crown precipitate. Hydrosulphuric acid reduces salts of sesquioxide with the deposition of sulphur ; but sesquioxide is again produced by ex- posure to the atmosphere, or by the action of oxidizing agents. Sulphite of ammonia produces a black precipitate. Before the blowpipe, heated alone on charcoal, sesquioxide of uranium is converted into protoxide ; heated with microcosmic salt on platinum wire in the oxidating flame, it dissolves, pro- ducing a clear yellow glass, which on cooling becomes green. Characteristic. The reactions with the caustic and carbo- nated alkalies and with ferrocyanide of potassium. General Remarks on the Oxides of the Fourth Group. Of the seven oxides constituting this group, oxide of zinc alone is soluble in caustic potassa. It is thus readily separated from the other six, and it is distinguished from alumina, oxide of chromium, etc., members of the third group, by its being again precipitated from its alkaline solution by sulphuretted hydrogen. Oxides of zinc, cobalt, iron, and manganese form with ammonia soluble double salts not precipitable by alkalies. Oxide of iron is easily eliminated from the other three oxides, by converting it into sesquioxide, by boiling with nitric acid or with chlorate of potash. It is then completely precipitated by ammonia, provided no non-volatile organic matter be present. The same is the case with sesquioxide of uranium. Hydrate of G 2 84 QUALITATIVE ANALYSIS. oxide of nickel is dissolved by ammonia, but again precipitated by potassa. Hydrated oxide of nickel and Tiydrated oxide of cobalt are both soluble in carbonate of ammonia, which is not the case with Jiydrated protoxide of manganese, which may thus, therefore, be separated from them. The perfect separation of oxide of nickel from oxide of cobalt is attended with great difficulties. The presence of nickel may be recognized by the behaviour of the solution of its cyanide in cyanide of potas- sium with hydrochloric acid, which precipitates it, the same not being the case with cyanide of cobalt. In the former case a double cyanide of nickel and potassium (NiCy,KCy) is formed ; but in the latter, with the presence of free hydro- cyanic acid, a cobaltocyanide of potassium is formed. Thus : Now, though the solution of cyanide of cobalt in cyanide of potassium is not precipitated by an acid, it may be so in the presence of nickel, and if that metal be present in the propor- tion of three equivalents to two equivalents of cobalt, the whole of the latter will be precipitated, the three equivalents of nickel replacing the three equivalents of potassium in the cobaltocyanide of potassium, and giving rise to cobaltocyanide of nickel (N 3 Co 2 Cy 6 ). Thus : 3(NiCy + KCy) + Co 2 K 3 Cy 6 = 6KCy + Co 2 Ni 3 Cy 6 . The blowpipe is, moreover, an infallible test of the presence of oxide of cobalt. GROUP V. Section A. Metallic Oxides precipitated from their solutions, whe- ther acid, alkaline, or neutral, as Sulphides by Hydro- sulphuric Acid. (Oxides of Lead, Silver, Mercury, Bismuth, Cadmium, Copper, Palladium, Iridium and Osmium.) 49. PKOTOXIDE OF LEAD. (PbO.) General characters. When pure, this oxide is yellow, but its powder has a reddish tint. "When certain lead salts, as the sub- nitrate and oxalate, are decomposed by heat without fusion, they furnish an oxide of a pure sulphur-yellow colour, which by tri- turation becomes red : in this state it is sometimes called METALLIC OXIDES. 85 massicot. By allowing a solution of oxide of lead, in caustic soda, to remain for some months exposed to the air, white, semi- transparent, dodecahedron crystals of anhydrous oxide may be obtained. The same crystals, according to Payen, are formed on mixing a dilute solution of acetate of lead, with great ex- cess of caustic ammonia, and exposing to the rays of the sun. Protoxide of lead becomes of a deep-red colour when heated, regaining its primitive colour on cooling : at a red heat it fuses and cools in semitransparent, deep brick-red, crystalline scales ; at a still higher temperature it undergoes partial decomposition, the metal being reduced. It absorbs carbonic acid slowly from the air ; hence an effervescence generally attends its solution in an acid. Pure water is capable of retaining in solution from ToVo * 12000 f oxide of lead. The hydrate is white, and ab- sorbs carbonic acid rapidly from the air. It loses its water at about 300, Protoxide of lead combines with alkalies and earths ; its combinations with potassa and soda are crystalliza- ble. It enters into fusion with, and dissolves, the earths with great facility. Its best solvent is nitric or acetic acid. Comportment of solutions of Salts of Lead with reagents. (Solution of Nitrate or Acetate of Lead may be used.) Potassa and ammonia produce white precipitates, which are basic salts, soluble in a great excess of potassa, but insoluble in ammonia. The carbonated alkalies produce white precipitates, soluble in potassa. Phosphate of soda occasions a white precipitate, soluble in potassa. Oxalic acid produces, in neutral solutions, a white preci- pitate. Ferrocyanide of potassium produces a white precipitate, solu- ble in potassa. Iodide of potassium produces a yellow precipitate, soluble in great excess by heat, and separating on cooling, in magnificent yellow spangles. (Pbl.) Chromaie of potassa produces a fine yellow precipitate, solu- ble in potassa, but insoluble in dilute nitric acid. (PbO,CrO 3 .) Hydrochloric acid and soluble chlorides produce a heavy white precipitate, soluble in boiling water, out of which it separates, on cooling, in brilliant crystals. This precipitate is soluble in potassa. (PbCl.) 86 QUALITATIVE ANALYSIS. HydrosulpJiuric acid produces a black precipitate, both in acid and neutral solutions. (PbS.) Sulphide of ammonium produces a black precipitate, insoluble in excess. Sulphuric acid and soluble sulphates produce a white precipi- tate (PbO,S0 3 ) sparingly soluble in dilute acids, but soluble in solution of potassa, and assuming a black colour when moist- ened with hydrosulphuret of ammonia. Before the blowpipe, on charcoal, mixed with carbonate of soda, salts of lead are immediately reduced, furnishing a me- tallic globule, which gradually sublimes, leaving a yellow resi- due. The metallic globule can easily be flattened under the hammer. Characteristic reactions. Those with sulphuric and hydro- chloric aoidsand with sulphuretted hydrogen. 50. Poisoning ly Lead. Detection of the Metal in Organic Mixtures. The vegetable acids and alkaline and fatty substances dis- solve protoxide of lead, forming salts of a poisonous character ; and as the glaze of earthenware vessels employed for culinary purposes usually contains oxide of lead, the accidental adultera- tions of food and liquids which have been prepared or allowed to remain for some time in such vessels, are not of uncommon occurrence. Pies which have been baked in a newly glazed vessel have been known to become contaminated with oxide of lead, and to have occasioned the peculiar symptoms of lead poi- soning in the persons partaking of them. The presence of the deleterious metal in the food, in cases of this kind, is proved .by moistening the articles with sulphide of ammonium, by which they become of a more or less brown colour ; and lead is detected in the glaze of the vessels by boiling in them acetic acid or caustic potass, by which the oxide becomes dissolved. Cases of poisoning by the accidental impregnation of articles of drink with oxide of lead would probably be more frequent, but for the property possessed by some vegetable substances, particularly those containing tannin or gallic acid, of forming insoluble compounds ; but, as a general rule, no substances used for food should be kept in leaden vessels, neither should liquids containing free acid, or in which a vegetable acid is likely to be formed, be kept for any length of time in a newly glazed earthenware vessel. Dr. Christison relates a case in METALLIC OXIDES. 87 which lead colic, paralysis, and death were occasioned by eat- ing the cream which formed on milk kept in a leaden cistern. Wines have been accidentally impregnated with oxide of lead, in consequence of the bottles having been cleaned with shot, and some of the pellets left behind ; the impregnation is not very likely to occur to any injurious extent in good wine, such as port or sherry, but in acid wines, such as currant or goose- berry, a sufficient quantity of the oxide might be dissolved to produce serious consequences. Wines have been also designedly adulterated with oxide of lead, to remove their acidity, and to give them a sweet taste ; this fraud was formerly practised to a considerable extent. It was detected and exposed in France by Fourcroy. In this country it does not appear to have been pursued to any material extent ; though, before the poisonous nature of lead compounds was generally known, British wines must have been frequently adulterated, for, in an old cookery book, quoted by Sir G. Baker, the following recipe occurs : " To hinder Wine from turning. Put a pound of melted lead in fair water into your cask, pretty warm, and stop it close." Cider, also, is apt to become impregnated with salts of lead, in consequence of the metal being used for various purposes in the construction of the cider house ; and although the source of the contamination was long ago pointed out by Sir George Baker, so recently as 1841 a set of cases which presented the incipient symptoms of lead colic were traced by MM. Chevalier and Ollivier to cider having been adulterated with lead to the amount of nearly two grains and a half per quart, in consequence of a publican having kept his cider for two days in a vessel lined with lead. The poisonous salt was found to be malate. New rum is often found to contain lead, which it derives from the worm of the still ; old rum, on the contrary, rarely con- tains the poison ; a circumstance which affords a good illustra- tion of the action of tannin in precipitating the nietal in an insoluble form, for Dr. Traill states that the rum intended for immediate consumption is collected at once from the still into glass bottles, while that intended to be kept is received into oak casks. When oxide of lead has to be looked for in liquids containing organic substances, a little nitric acid should be added previous to nitration, in order to redissolve any insoluble compound formed by the salts of lead with albumen and other organic principles, sulphuretted hydrogen is then transmitted through 88 QUALITATIVE ANALYSIS. the clear liquor : if a dark-coloured precipitate is formed, the whole is boiled and the precipitate collected on a filter. It is washed and dried, and then boiled with strong nitric acid, by which the sulphide of lead is wholly or in part converted into nitrate, which being soluble in water is filtered, neutralized by ammonia, and submitted to the general reagents as above. Should it be necessary to examine the insoluble matters, they are to be cut in small pieces and incinerated in a crucible with four parts of black flux ; any lead that may be present will be found fused into a small button at the bottom of the crucible, and may be removed from the carbonaceous matters by washing with water. If the metal is in the form of dry oxide, it is readily detected by the blowpipe on charcoal, or by mixing it with paste, spreading the mixture on a piece of card, drying and burning it, when metallic lead is immediately produced. Poisoning by carbonate of lead is very common amongst white- lead manufacturers and painters ; the poison finds its way into the system continuously and insidiously in minute quantities, inducing the complaint known as colica pictorum and paralysis. Cases of this form of poisoning have not been so frequent in manufactories of white lead since the practice has been adopted of grinding the carbonate in water. Dr. A. T. Thompson con- siders that the carbonate is the only compound of lead which possesses poisonous properties; and that if any other salt in small doses becomes so, it is in consequence of its conversion into carbonate in the body. This opinion is not, however, adopted by Dr. Christison, or by Dr. Alfred Taylor. Carbonate of lead is known by the following properties : (1) it is blackened by sulphuretted hydrogen ; (2) it is soluble with effervescence in nitric acid ; if however, as is frequently the case with the com- mercial carbonate, it contains sulphate of lead or sulphate of baryta, the two latter substances remain undissolved by the nitric acid ; the presence of oxide of lead in the nitric acid solution is proved by the tests mentioned above ; (3) it is de- composed when ignited on platinum foil, leaving oxide of lead of a yellow colour ; (4) it is reduced on charcoal by the blow- pipe, metallic lead being produced. 51. Action of Water on Lead. Gmelin, in his elaborate ' Handbook of Chemistry,' sums up the state of our knowledge of the circumstances under which METALLIC OXIDES. 89 water becomes impregnated with lead, thus : 1. Clean lead, in contact with water and air, free from carbonic acid, yields a solution of oxide of lead, which turns reddened litmus-paper blue, is turned brown by sulphuretted hydrogen, and gives a white precipitate with sulphuric acid. 2. Water, freed from air by boiling, does not dissolve lead when kept in contact with it in a close vessel. 3. Water which has been agitated with air takes up, in two hours, a quantity of oxide of lead amount- ing to between 12 Q 00 and -j--^-^ ^ then reddens blue litmus- paper, and gives a brownish-black precipitate with sulphuretted hydrogen. 4. Spring water, containing If grains of saline matter in two pints, arid no carbonic acid, when passed through a leaden tube 150 feet long, dissolves a quantity of lead suf- ficient to give a brown colour with sulphuretted hydrogen. 5. Distilled water, in contact with lead and air, free from car- bonic acid, dissolves 70 i 00 of oxide, acquires an alkaline re- action, and becomes turbid on exposure to the air. Morveau was the first to notice that the presence of small quantities of carbonic acid, sulphuric acid, or of various salts, either prevents or greatly diminishes the quantity of oxide of lead dissolved ; and Yorke found that when clean lead was exposed to spring water, 10 pounds of which contained 1'21 grain of chloride of sodium and chloride of calcium together with 64 grains of car- bonate of lime, a slight deposit of brown oxide took place on the surface of the metal, but no oxide was dissolved. In rela- tion to the important subject of the contamination of water by lead, Christison remarks: "Leaden pipes should not be used for conducting wa,ter, unless lead remains untarnished after twenty-four hours' immersion ; they are unfit for the purpose if the water contains less than ^oVoth of its weight of salts. If the quantity of salts exceeds this limit, and the salts consist mainly of sulphates and carbonates, leaden pipes may be used ; but if they consist chiefly of chlorides, even 1 part in 4000 is not sufficient to prevent the solution of the lead." That it is dangerous, however, to rely on the " preservative action " of sulphates and carbonates, even when the water contains such salts in considerable quantities, will be evident from the follow- ing analysis, made by the author, of a water, the action of which on the leaden cistern was so great that the bottom of the vessel was, after six months, eaten into holes. One gallon of this water contained nearly 78 grains of solid matter, the composition of which was 90 QUALITATIVE ANALYSIS. Silica 0-24 Carbonate of lime 15'09 Carbonate of magnesia 18'97 Sulphate of lime 15-32 Sulphate of potassa 6*79 Sulphate of soda 10-77 Chloride of sodium 11-46 Organic matters 4-10 77-74 If the organic matters dissolved in water contain nitrogen, nitric acid may be formed by the oxidation of that element, which may be greatly promoted by the presence of alkaline carbonates. Water impregnated with nitrates acts powerfully on lead, and if carbonates and sulphates are not also present, the water may hold in solution a very injurious quantity of lead salt. This is illustrated by the following analysis, made by the author, of a well-water from Highgate, in 1848, which acted strongly on the leaden cistern in which it was stored. An im- perial gallon contained Silica 0-896 Sulphate of potash 17-044 Sulphate of soda 9'515 Chloride of calcium 5'920 Chloride of sodium 9'632 Nitrate of lime 40-120 Nitrate of magnesia 17 '064 Organic matters none 100-191 In their chemical report on the supply of water to the me- tropolis, addressed to the Secretary of State, in June, 1851, Professors Graham, Miller, and Hofmann give the following as the most practical conclusions which they arrived at after a long inquiry, undertaken to illustrate the action of water on lead, a subject, they observe, of great difficulty, and still far from being exhausted : Certain salts, particularly sulphates, to which a protecting effect is usually ascribed, did not appear to exercise uniformly that useful property. Some salts, on the other hand, such as chlorides, and more particularly nitrates, increase the solvent action of water. Of all protecting actions, that of carbonate METALLIC OXIDES. 91 of lime, dissolved in carbonic acid, appeared to be the most considerable and surest. The soluble oxide of lead is converted into the carbonate, which, although not absolutely insoluble, appears to be the least soluble of all the salts of lead. Pure water did not dissolve a quantity of carbonate of lead greater than gL of a grain to the gallon ; while water, on the other hand, which contained already so much as 6 grains of lead to the gallon, had the quantity reduced to -^ of a grain by free exposure to the atmosphere for twenty-four hours, the lead being deposited as carbonate by the absorption of carbonic acid. Carbonic acid is usually present in well, river, and lake waters, in quantity sufficient for protection ; and the immunity of such waters from lead impregnation is probably to be ascribed more often to their carbonic acid than to the salts which they may also contain. Organic matter in soft water is doubly danger- ous, as the rapid corrosion which it occasions may be followed by solution of the lead salt formed. As an illustration of the action of soft water, nearly free from earthy carbonates, on lead, even when no soluble organic matter is present, I may refer to a well-water from Hatton, which was analysed by the author, at the request of the Lord Chief Baron, in consequence of its destructive action on the leaden cistern. This water con- tained 38'47 grains of saline matter per imperial gallon, con- sisting of Carbonate of lime 2'120 Carbonate of magnesia .... '880 Carbonate of soda 15-196 Sulphate of potassa traces Sulphate of soda 10'456 Chloride of sodium 9'288 Protocarbonate of iron .... '480 Silica -050 38-470 52. Recognition and Quantitative Estimation of Lead in Water. This, fortunately, is attended with no difficulties, though cer- tain precautions are necessary when the water contains (as almost all natural waters do, more or less) soluble organic matter. We possess, in sulphuretted hydrogen, a test, by which the presence of % of a grain of lead is indicated in one gallon of distilled water by the production of a brown tinge. In y$J QUALITATIVE ANALYSIS. making the experiment, the beaker should be placed on a sheet of white paper, and a second beaker, to which no sulphuretted hydrogen has been added, should be placed by its side, also on white paper, so that the faintest brown tinge may be rendered evident by comparison ; a few drops of pure hydrochloric acid should be added previous to the application of the test. Dr. Smith, in his investigation of the action of the waters of the Dee and the Don on lead pipes and cisterns, employed the tint produced by sulphuretted hydrogen as a means of estima- ting, quantitatively, the amount of lead deposited in water. 1*6 grains of nitrate of lead were dissolved in 1000 grains of distilled water; a solution was thus prepared, 1000 grains of which contained 1 grain of metallic lead. From an accurately graduated measure, this solution was dropped into a gallon of water containing no lead, until a transmission of sulphuretted hydrogen, the same depth of tint, was developed as in the par- ticular ease on trial. Very accurate results are said to have been obtained, until the quantity of lead exceeds i grain per gallon, when the colour gets so dark that slight differences cannot be discriminated : -j^- of a grain of lead to a gallon of Eure water gave a tint quite perceptible, that is, one of metallic jad in seven millions, and a less quantity could readily be distin- guished by careful comparison, even in specimens previously containing a considerable proportion of lead, the difference of Y^-Q-th of a grain was plainly visible. The presence of organic matter in solution in water interferes materially with the ac- tion of sulphuretted hydrogen on the lead which it may con- tain ; and, as Dr. Smith informs us that nearly one-half of the solid matter in the water of the river Dee consisted of organic matter, it would appear that the method he describes could not give very reliable quantitative results. Previous to applying the sulphuretted hydrogen test, the water should be carefully evaporated to dryness and the residue ignited in a small porcelain capsule, whereby the organic matter is destroyed. The saline matter should be moistened with nitric acid, and then warmed with the addition of acetic acid and water, and, if necessary, filtered. To estimate the amount of lead quantitatively, a gallon of the water should be employed ; it should be gently evaporated to dryness, and the residue having been moistened with a few drops of nitric acid and ignited, k digested with dilute hydrochloric acid and filtered. The solution is neutra- lized with. carbonate of soda, and acidified with acetic acid; a METALLIC OXIDES. 93 small quantity of bichromate of potash is then added, and the liquid set aside for some hours. If a yellow precipitate forms, it is cliromate of lead, which is redissolved in hydrochloric acid, and tartaric acid and excess of ammonia having been added, sulphuretted hydrogen is passed through the solution. The precipitated sulphide of lead is washed by decantation, and converted into sulphate by evaporating it to dryness with a little fuming nitric acid and a drop or two of sulphuric acid. The ignited residue contains 73*6 per cent, of oxide of lead. According to Taylor, sulphuric acid fails to detect T \ of a grain of acetate of lead in 1 ounce of water, and begins to form a decided precipitate only when the quantity amounts to \ of a grain. Iodide of potassium forms a rich yellow pre- cipitate with soluble salts of lead, which is soluble to some ex- tent in boiling water, separating again, as the solution cools, in the form of magnificent golden-coloured shining plates ; it fails, however, according to Taylor, to indicate the presence of ^ a grain of lead in 1 2 ounces of water. 53. OXIDE OF SILYEE. (AgO.) General characters. As obtained by dropping solution of nitrate of silver into caustic potassa, it is a greyish-brown powder ; but, if the solutions are concentrated and boiling, the oxide precipitates as a heavy black powder. Exposed to the rays of the sun, it disengages a certain quantity of oxygen, and turns black. It is entirely reduced by ignition. It is slightly soluble in pure water, and in water of barytes. It reacts alkaline to test paper, and displaces from their combina- tions with the alkalies a portion of the acids, with which it forms insoluble compounds. It combines with caustic ammonia, giving rise to a dangerous substance (fulminating silver). It readily dissolves in nitric and other acids. Comportment of solutions of Salts of Silver with reagents. (Solution of Nitrate of Silver may be used.) Potassa and ammonia produce a light brown precipitate, readily soluble in ammonia (AgO, HO). The carbonated alkalies produce a white precipitate, soluble in carbonate of ammonia (AgO,C0 2 ). Phosphate of soda produces, in neutral solutions, a yellow precipitate, soluble in ammonia (3AgO,P0 5 ), Solution of 94 QUALITATIVE ANALYSIS. ignited phosphate of soda (2NaO,P0 5 ) gives a white precipi- tate (2AgO,P0 5 ). Ferrocyanide of potassium gives a white precipitate. Ferridcyanide of potassium produces a reddish-brown precipi- tate. Chromate of potassa produces a rich brown precipitate. Protosulphate of iron produces a white precipitate, consisting of metallic silver. Hydrochloric acid and the soluble chlorides produce a white curdy precipitate even in exceedingly dilute solutions. This precipitate becomes violet, and finally black, without, however, suffering decomposition by exposure to light. It is insoluble in dilute acids, but readily soluble in ammonia. "When heated it fuses, without decomposition, into a horny mass. If the solution of silver be exceedingly dilute, hydrochloric acid pro- duces an opalescent appearance. A bar of metallic zinc precipitates silver from its solution in the metallic state. Hydrosulphuric acid produces a black precipitate, both in acid and in neutral solutions. (AgS.) Sulphide of ammonium gives a black precipitate, insoluble in excess. Before the blowpipe, mixed with carbonate of soda, and heated on charcoal, salts of silver are readily reduced, while no incrustation takes place. Characteristic reaction. That with hydrochloric acid. 54. SUBOXIDE or MERCURY. (Hg 2 O.) General characters. It is a black powder which a very gentle heat converts into metallic mercury and oxide of mercury, and a stronger heat into mercury and oxygen gas. The black pow- der obtained by the long-continued agitation of the metal, and which was supposed to consist of this oxide, is probably only the metal in a state of very fine division. The soluble salts of this oxide redden litmus-paper, and are decomposed, when mixed with much water, into soluble acid, and insoluble basic salts. Comportment of solutions of Salts of Suboxide of Mercury with (Solution of Nitrate of Suboxide of Mercury may be used.) Potassa and ammonia produce a black precipitate insoluble in excess (Hg 2 O,HO). METALLIC OXIDES. 95 Alkaline carbonates produce a dirty-yellow precipitate which turns black by boiling. Phosphate of soda produces a white precipitate. Ferrocyanide of potassium produces a white gelatinous pre- cipitate. Ferridcyanide of potassium gives a reddish-brown precipitate, which gradually becomes white. Iodide of potassium produces a greenish-yellow precipitate, soluble in excess. (Hg 2 I.) Chr ornate of potassa produces a red precipitate. Hydrosulphuric acid produces a black precipitate, both in acid and neutral solutions. (Hg 2 S.) Sulphide of ammonium produces a black precipitate, insoluble in excess, but decomposed by potassa into sulphide of mercury and metallic mercury. It is not decomposed or dissolved by boiling nitric acid, but easily by aqua regia. A bar of metallic zinc throws down an amalgam of zinc and mercury. Hydro chloric acid and soluble chlorides produce a white pre- cipitate (Hg 2 Cl), insoluble in acids, but rendered black by po- tassa and ammonia, the suboxide being formed. Protochloride of tin produces a grey precipitate, which, boiling, resolves into globules of metallic mercury. A drop of solution of a salt of suboxide of mercury, rubbed with a rag on a piece of bright copper, leaves a silvery stain, which disappears when it is heated to redness. Before the blowpipe -, mixed with carbonate of soda, and heated in a glass tube, metallic mercury sublimes in the form of a grey powder, which, on being rubbed with a glass rod, is resolved into globules. Calomel (Hg 2 Cl) should be used for this experiment. Characteristic reactions. Those with hydrochloric acid, po- tassa, and ammonia. 55. OXIDE OF MEECTJKY. (HgO.) General characters. It is a brick-red crystalline powder; but when finely pulverized has a yellow tinge. At a slightly elevated temperature it turns black, but it regains its red colour on cooling. At a red heat it is resolved into oxygen gas and metallic mercury, and is entirely volatilized. In this manner the presence of impurities, red-lead or brick-dust, may be detected. It dissolves readily in acids, but is not acted on by ammonia. QUALITATIVE ANALYSIS. Comportment of solutions of Salts of Oxide of Mercury with reagents. (Solution of Corrosive Sublimate (HgCl) may be used.) Potassa produces a yellow precipitate (HgO,HO) insoluble in excess. If an insufficient quantity of the alkali be added, the precipitate is reddish-brown. The presence of ammoniacal salts causes the formation of a white precipitate, which is a compound of amidide of mercury with undecomposed mercury salts. Thus: 2HgCl+NH 3 =(HgNH 2 HgCl)+HCL Fixed carbonated alkalies produce a reddish-brown precipi- tate, insoluble in excess, but converted, in the presence of ammoniacal salts, into the above white product. Carbonate of ammonia produces a white precipitate. Phosphate of soda produces a white precipitate. Ferrocyanide of potassium produces a white precipitate, which eventually becomes blue. Ferridcyanide of potassium produces, in solutions of the ni- trate and sulphate, a yellow precipitate ; in solutions of the chloride, none. Hydrosulphuric acid and sulphide of ammonium give rise to different-coloured precipitates according to the quantity of the reagent added. If it be added in small quantity, and the solu- tion agitated, the precipitate is white, being a compound of sulphide of mercury and undecomposed salt ; the addition of larger quantities causes the precipitate to assume, successively, a yellow, orange, brown, red, and black colour. (HgS.) The sulphide of mercury is soluble in solution of potassa, but not in boiling nitric acid, though it dissolves readily in aqua regia. ! Iodide of potassium produces a cinnabar-red precipitate, so- j luble in excess. (Hgl.) It crystallizes out of a hot solution in | magnificent crimson spangles. Protochloride of tin added in small quantity to a solution of j: chloride of mercury, gives rise to the precipitation of subchloride. 2 Hg C1 + Sn Cl = Hg 2 C1 + Sn C1 3 . When the reagent is added in excess, the subchloride is in its turn decomposed, a grey powder being formed, which may, by j boiling with a few drops of hj drochloric acid, be united into ji globules of metallic mercury. Hg 2 Cl + Sn Cl = Hg 2 + Sn C1 2 . The Galvanic Tests. On placing a drop of a strong solution of j METALLIC OXIDES. 97 corrosive sublimate on a gold coin, and touching the latter through the solution with an iron point, the mercury will be deposited on the coin in the form of a bright silvery stain. This elegant test has been modified in a variety of ways. By the followingprocess of Deverge distinct indications are obtained where the poison does not form more than an 80,000th part of the water. A thin plate of gold, and another of tin, a few lines broad, and two or three inches long, are closely applied to one another by silk threads at the ends, and then twisted spi- rally ; this small galvanic pile is left for twenty-four or twenty- six hours in the solution previously acidified with hydrochloric acid, upon which the gold is found whitened, and mercury may be obtained in globules by heating the gold in a tube. For facility of application, an important condition is, that the quantity of fluid should not exceed three or four ounces, because, in a larger quantity, a pile of the size stated cannot remove the whole of the mercury. Characteristic reactions. Those with potassa and ammonia. 56. Poisoning ly Mercury. Detection of the Metal in Organic Mixtures. From the researches of M. Boullay and Professor Orfila, it appears that various vegetable fluids, extracts, etc., possess the power of decomposing corrosive sublimate, a portion of the chlorine being gradually disengaged in the form of hydro- chloric acid ; the salt is consequently converted into calomel, which is deposited in a state of mixture, or combination with vegetable matter. It has been further shown by Professor Taddei, of Florence, that gluten possesses the same property in a remarkable degree, and that that vegetable principle is capa- ble of removing the poison completely from its solution, form- ing with it a ternary compound of protochloride of mercury and gluten. Among soluble vegetable principles, albumen, casein, and gelatin possess similar properties ; hence solution of albumen is an antidote against the effects of the poison. Various insoluble animal principles have been likewise shown to act on corrosive sublimate in the same manner as vege- table gluten, and this chemical action appears to be the source of the corrosive property of the poison. "With these facts before us, it is evident that in examining cases of poi- soning bv corrosive sublimate, we must not always expect to find it in a state of solution, even though it may have been re- PAET I. H 98 QUALITATIVE ANALYSIS. ceived into the stomach in considerable quantities. The atten- tion of the toxicologist must, therefore, be directed to both the liquid and the solid portions. 1. Examination of the liquid portion. It is filtered from the solid matter, and a portion tested with protochloride of tin or by the galvanic test. Should evidence of the poison be thus obtained, the remainder of the solution is mixed with half its volume of ether and well agitated : on allowing it to stand some time, the greater part of the ether rises to the surface, and may be removed by a pipette. It should then be filtered, and eva- porated to dryness, or distilled ; and the residue, being redis- solved in water, is submitted to the proper tests. Dr. Taylor does not find this process to answer unless the poison is present in a moderately large proportion. He therefore recommends to acidulate the suspected fluid with hydrochloric acid, and to immerse in it a narrow slip of finely laminated zinc, round which a spiral slip of fine gold-foil has been twisted : if at the end of five or six hours the gold still retains its bright colour, no corrosive sublimate is present. Should it, on the other hand, be tarnished, it is washed, first in ether, and then in water, dried, and heated in a small reduction tube to obtain a sublimate of the metal. 2. Examination of the insoluble matters. They are boiled with distilled water, and the liquid having been filtered is tested as above described. Should no trace of the poison be thus obtained, Dr. Christison triturates the whole of the insoluble matters with protochloride of tin, and collects and washes the coagulum formed ; he then boils it with a moderately strong solution of caustic potassa in a glazed porcelain vessel till all the lumps have disappeared. The animal and vegetable matters^ and oxide of tin united with them are thus dissolved, and a powder of metallic mercury, generally easily discernible with a small magnifier, subsides. To separate it, the solution is allowed to remain at rest at a temperature little short of ebul- lition for fifteen or twenty minutes ; the vessel is then filled up with hot water, and the fatty matters which float on the sur- face skimmed off". After two or three affusions, the black pow- der is transferred to a smaller vessel, and finally dried, and sub- limed in a small tube. By this mode of operating, Dr. Chris- tison has detected a quarter of a grain of corrosive sublimate mixed with two ounces of beef, or with five ounces of new milk, or porter, or tea, made with a liberal allowance of cream and sugar. He has also thus detected a tenth part of a grain METALLIC OXIDES. 99 in four ounces of the last mixture, that is in 19,200 times its weight. Dr. Taylor directs that the coagulum obtained by tri- turating the organic mixture with protochloride of tin should be well boiled in strong hydrochloric acid previous to digesting it with caustic potassa, in order to separate the mercury from the oxide of tin and organic matters, and to estimate it quanti- tatively. In cases where the quantity of poison is small, he has found this mode of treatment to be troublesome and not unat- tended with risk. In such cases, he prefers the galvanic test, by which he states that he has detected one-sixteenth part of a grain of corrosive sublimate dissolved in one ounce of organic liquid, and obtained metallic mercury from it in less than half an hour ; and he does not think it possible to conceive a case where, in an analysis of this kind, the galvanic test would not be immediately applicable. In the * Comptes-Eendus,' March 31, 1845, MM. Danger, and Plandin have described their method of extracting mer- cury from the liver of an animal poisoned by corrosive sub- limate : the process is nothing more than a modification of the galvanic test, but the experiments are valuable as proving the absorption of the poison. Their process is as follows : The animal substance finely cut up is heated to about 212, with one-third or half its weight of strong sulphuric acid ; in an hour or two, the whole forms a dark carbonaceous-looking liquid. It is allowed to cool, and chloride of lime gradually added. The liquid becomes whiter and more viscid. The quan- tity of chloride used is about equal to the weight of the sul- phuric acid; it is added until the whole appears like a white calcareous mass. The dried residue is then digested in abso- lute alcohol, which dissolves the mercurial compound. It is now diluted with water, and the earthy residue repeatedly washed, the liquids being afterwards mixed and concentrated. The concentrated liquid is placed in a funnel terminating at an angle of 90 in a capillary point the galvanic plates of gold and tin being introduced into the contracted part of the funnel. In this way every drop of the liquid comes in contact with the metals, and the gold is slowly covered with mercury. They state that they have thus detected the metal in a solution con- taining the 100,000th part. According to Dr. Taylor, it is easy to detect corrosive sublimate in organic solids by simply boiling them with copper gauze and a few drops of hydrochloric acid. It must be observed that Orfila takes an exception to the em- H2 100 QUALITATIVE ANALYSIS. ployment of tin and gold as the galvanic elements, and on these grounds : " Some tin" he says, " may, after a time, be dis- solved, and precipitated on the gold, and thus simulate the ap- pearance of mercury ; the whitened gold, when exposed to heat, may even recover its golden colour ; the whitening of the gold does not, therefore, absolutely demonstrate the existence of a mercurial compound in the suspected substance ; and, if the fluid metal cannot be afterwards obtained in distinct globules, the evidence must be regarded as inconclusive." When corrosive sublimate is contained in organic substances (albumen, etc.), insoluble in water, Rose recommends to digest them in ammonia in preference to potassa, the mercury being subsequently precipitated on copper much more readily from an ammonical than from a potassa solution ; and, in cases where ammonia fails to effect a solution, he prefers the following me- thod of treatment to the ordinary one of digesting with nitric acid: The dried substance mixed with about a quarter of its weight of carbonate of potassa is introduced into a capacious retort, and distilled at a heat gradually increasing to redness : the mercury rises and is condensed in the neck of the retort, where it can easily be distinguished from the brown viscid em- pyreumatic oil which passes over at the same time. If the or- ganic matter contained only slight traces of mercury, the whole of the metal is found in the neck of the retort, none passing with the oil into the receiver. 57. TEROXIDE OF BISMUTH. (BiO 3 .) General characters. When pure, it is of a straw-yellow colour, and melts at a strong red-heat to an opaque glass, which, while hot, is dark-brown or black, but on cooling becomes yellow ; when melted with silica, alumina, or metallic oxides, it dissolves them readily. The oxide precipitated by water re- tains nitric acid, from which it may be freed by caustic potassa or soda, which convert it into a hydrate. It is easily reduced by ignition with organic substances or charcoal powder. Its salts are colourless. Comportment of solutions of Salts of Bismuth with reagents. (Solution of Nitrate of Teroxide of Bismuth may be used.) Potassa and ammonia produce white precipitates, insoluble in excess. Alkaline carbonates and phosphate of soda produce white pre- cipitates. METALLIC OXIDES. 101 Ferrocyanide of potassium occasions a white precipitate, in- soluble in hydrochloric acid. Ferridcyanide of potassium produces a light yellow precipi- tate, soluble in hydrochloric acid. Iodide of potassium produces a brown precipitate, readily so- luble in excess. ; ^ , J , , J Chromate ofpotassa produces a yellow^ pfSej^itate, sol ub}.3 in dilute nitric acid. ^ J ' Sydrosulphuric acid and sulpkidt of/ammotyWJfi gijodixoe? a black precipitate both in acid and neutral' s^Mfons', Insoluble in excess, and in dilute nitric acid, but soluble in boiling nitric acid. (BiS 3 .) Before the blowpipe, heated on charcoal in the reducing flame, salts of bismuth are easily reduced to brittle globules, which spring to pieces under the hammer ; the charcoal at the same time becomes covered with a yellow incrustation. Characteristic reaction. The neutral salts of bismuth are distinguished by their property of being decomposed by water into a soluble acid, and an insoluble basic salt. The chloride of bismuth exhibits this property in the most marked manner. The insoluble basic bismuth salt is distinguished from the basic salt of antimony formed under similar circumstances by its being insoluble in tartaric acid. 58. OXIDE OF CADMIUM. (CdO.) General characters. The colour of this oxide varies accord- ing to its state of aggregation. It is sometimes of a deep red- brown, sometimes clear brown, and occasionally black. It is infusible, and does not volatilize at exceedingly high tempera- tures ; but, when mixed with powdered charcoal, it is reduced by heat, and the metal burns and volatilizes. By long-continued gentle ebullition of the metal, the oxide may be obtained, ac- cording to Herapath,in long purple needles, opaque, and grouped in rays. Its hydrate is white ; it loses its water by heat, and absorbs carbonic acid from the air. It is not soluble in the fixed alkalies, but it dissolves in caustic ammonia. It dissolves easily in acids, forming colourless solutions. Comportment of solutions of Salts of Cadmium with reagents. (Solution of Chloride of Cadmium may be used.) Potassa and ammonia produce a white precipitate (HO, CdO), insoluble in potassa, but easily soluble in ammonia. 102 QUALITATIVE ANALYSIS. The carbonated alkalies produce a white precipitate (CdO, C O 2 ), insoluble in excess : ammoniacal salts do not prevent the formation of this precipitate. Phosphate of soda produces a white precipitate. Oxalic acid produces a precipitate soluble in ammonia. <-Fsrrocyanid_e.. of potassium produces a slightly yellow precipi- tajt&j Soluble,. 'in* hydrochloric acid. JFerridcyanide of potassium produces a yellow precipitate, also jsokik ) being f jrmed, '"', which produces, with persalts of iron, a very deep blood- red colour. Taylor recommends the following method of applying this test : "Place the diluted hydrocyanic acid in a,watch-glass, and invert over it another watch-glass, holding in its centre one drop of sulphide of ammonium containing sulphur in excess ; there is no apparent change in the sulphide, but if the watch- glass be removed after the lapse of from- half a minute to ten minutes, according to the quantity of hydrocyanic acid present, sulphocyanide of ammonium will be obtained. On gently heat- ing the drop of sulphide, and evaporating it to dryncss, the ad- dition of a drop of solution of a persalt of iron to the dried residue, brings out the blood-red colour instantly, which is in- tense in proportion to the quantity of sulphocyanide present ; the warmth of the hand may be employed to expedite the evolution of the vapour. This test is even more delicate and expeditious than the nitrate of silver test, in whicii the vapours are received in a solution of that salt, and will, according to Taylor, detect in five minutes, hydrocyanic acid not exceeding _^._ of a grain, in ten drops of a liquid. JSolid cyanides when heated with concentrated sulphuric acid evolve carbonic oxide, sulphate of the metallic base and sulphate of ammonia being formed. Thus: 102. Detection of Hydrocyanic Acid in Organic Mixtures. When hydrocyanic acid is to be sought for in organic mix- tures, Taylor urges the importance of obtaining evidence of its presence before resorting to the process of distillation, in order to avoid an objection which may be raised, to the effect that the acid might have been a product of distillation. He gives PAJBT I. M 162 QUALITATIVE ANALYSIS. the following method, which he finds to succeed with the blood, muscles, and viscera, as well as with the liquid contents and solid substance of the stomach. The liquid is acidulated with sulphuric acid and placed in a wide-mouthed bottle, the aper- ture of which is capable of being covered with a watch-glass or a saucer of white porcelain. The inside of the glass or saucer is moistened with nitrate of silver and placed over the mouth of the bottle : if hydrocyanic acid be present, the spots of nitrate will become white in the course of a few seconds, or in from ten minutes to a quarter of an hour, according to the quantity of poison present, and the closeness of the test to the liquid. No heat need be applied to the liquid ; a temperature of 64 suffices for the result, but the warmth of the hand ac- celerates the action. Taylor mixed two-thirds of a grain of anhydrous hydrocyanic acid with eight ounces of porter, and obtained a well-defined deposit of cyanide of silver in the watch-glass in a quarter of an hour ; and by substituting in the watch-glass weak caustic potassa for nitrate of silver, and after a few minutes adding sulphate of iron and hydrochloric acid, he obtained Prussian blue from two-thirds of a grain of anhy- drous hydrocyanic acid mixed with eight ounces of porter ; the process was equally successful with the stomach of a dog, after it had been thoroughly washed with water, and had been re- moved from the body twenty-four hours. These experiments prove the great volatility of the poison, and the diffusibility of its vapour. The silver test serves the analyst as a guide, for Taylor finds, that unless the white film is formed on the nitrate of silver, the application of the Prussian blue test will also fail. Prussian blue may even be procured from the cyanide of silver : for this purpose, caustic potassa is added to the film of cyanide in the watch-glass, and heat applied ; the cyanide is dissolved, and brown oxide of silver precipitated ; to the fil- tered liquid sulphate of iron is added, and after a time dilute sulphuric acid, Prussian blue is immediately produced. When it is found necessary to distil the fluid, Lassaigne and Chris- tison acidulate with sulphuric acid, and distil from a vapour bath till one-eighth part of the fluid has passed into the re- ceiver. The tests are then applied to the distillate ; if the quantity of poison be small, nitrate of silver, or potassa, may be placed in the receiver to fix the acid as it passes over. Poisoning by .Hydrocyanic Acid. Liebig makes the following remarks on this fearful poison : " Its rapid action on the blood INORGANIC ACIDS. 163 is very remarkable. Comparatively large quantities of the acid in aqueous solution may be taken into the digestive ap- paratus without producing any very perceptible noxious effects; while the same quantity inhaled as vapour causes immediate death. Thus, a cat can bear the administration of from two to three drops of anhydrous hydrocyanic acid diluted with from four to six ounces of water, without being in the least aifected by it. If two drops of the anhydrous acid be inserted into the mouth of the same animal, taking care at the same time to prevent it from breathing by stopping its mouth and nostrils, no perceptible effect is produced; but the cat dies the very in- stant it is permitted to breathe, and consequently as soon as the vapour of the acid gets into the lungs." With respect to the quantity of hydrocyanic acid requisite to destroy lite, the matured opinion of Taylor is, that a quan- tity of Scheele's acid (at five per cent.), above twenty drops (i. e. one grain of anhydrous acid), or an equivalent portion of any other acid, would commonly prove fatal. Even less than this seven-tenths of a grain might under favourable circum- stances, destroy life. The largest dose from which an adult has been known to recover is forty minims at 3^ per cent., which is equivalent to about a grain and a third of anhy- drous acid. It may be well to observe that the acid of commerce differs much in strength, according to the process by which it has been prepared, and independently of decom- position by keeping. The medicinal acid long used in this country was intended to be an imitation of that of Vauquelin, which contains 3'3 per cent. ; but the London College of Phy- sicians have improperly (as Dr. Christison observes) altered the strength to 2 per cent. : that of Giese, which keeps well, is of the same strength as the first ; that of Schrader contains only 1 per cent ; that of Gobel, 2*5 per cent. ; that of Ittner, 10 per cent. Of the alcoholic solutions the best known are that of Schrader, which contains about T5 per cent, of pure acid ; that of the Bavarian Pharmacopoeia, which contains 4 per cent. ; that of Duflos, 9 per cent. ; that of Pfaff, 10 per cent. ; and that of Keller, 25 per cent. The medicinal dose is from a minim to two minims of Scheele's acid, and from three to five minims of the London Pharmacopoeia! acid gradually increased. Every grain of anhydrous acid yields 5 grains of cyanide of silver. Suppose 100 grains of a sample to have yielded 45 2 M 164 QUALITATIVE ANALYSIS. grains of cyanide, it follows that the acid contains 9 grain* of anhydrous acid to 91 grains of water. To procure an acid con- taining 2 per cent, of anhydrous acid, would require the addi- tion of 9 grains of anhydrous acid to 441 grains of water ; now, as the acid in question contains already 91 grains of water in every 100 grains, it is evident that we need only add 350 grains of water to every 100 grains of the acid, in order to prepare an acid of 2 per cent, strength (44191=350). 103. HYDROSULPHOCYANIC ACID. (HCyS 2 .) The sulphocyanides, most of which are soluble both in water and in alcohol, are recognized by the following reactions : Nitrate of silver produces a white precipitate (AgCyS 2 ), in- soluble in dilute nitric acid, and in ammonia, and furnishing metallic silver when ignited. Sesquichloride of iron gives an intense blood-red colour; if the red liquor be acidulated with hydrochloric acid, and frag- ments of zinc added, sulphuretted hydrogen is evolved and the colour disappears. A mixture of protosulphate of iron with a salt of copper, oc- casions a white precipitate. Subnitrate of mercury gives a white precipitate. Sulpho cyanide of sodium (or of ammonium) exists in the saliva, and may be detected by allowing a little dilute solution of perchloride of iron to remain for a few seconds in the mouth : on expelling it, it will be found to have acquired a dis- tinct red tinge. 14. HYDROFERROCYANIC ACID. (H 2 Cy 3 e; orH 2 Cfy.) This acid, when exposed to the air, absorbs oxygen and be- comes blue ; its solution when boiled is decomposed into hy- drocyanic acid and a white compound, which becomes blue by exposure to the air. With alkalies it forms salts which are soluble in water ; most of the other ferrocyanides are in- soluble. Comportment of soluble Ferrocyanides with reagents. (Solution of Ferrocyanide of Potassium (K Cfy) may be used.) Protosulphate of iron produces a pale-blue precipitate. (See Protoxide of Iron.) Sesquichloride of iron, a deep-blue precipitate (Prussian blue. (See Sesquioxide of Iron.) INORGANIC ACIDS. 165 Sulphate of copper a reddish-brown precipitate (Cu 3 Cfy). Protonitrate of cobalt a yellowish-green precipitate. With salts of zinc, lead, silver, and mercury, white precipi- tates. Insoluble ferrocyanides are decomposed by fusion with alka- line carbonates, soluble alkaline ferrocyanides being formed. When an alkaline ferrocyanide is distilled with dilute sulphuric acid, hydrocyanic acid is formed; but if it be heated with ex- cess of concentrated sulphuric acid carbonic acid is formed by the following reaction : One equivalent of ferrocyanide of potassium and nine equiva- lents of water from the oil of vitriol and the water of crystalli- zation of the ferrocyanide, give rise to six equivalents of car- bonic oxide, two equivalents of potasoa, one of protoxide of iron, and three of ammonia; the sulphuric acid, oxide of iron, potassa,and ammonia arrange themselves into a crystalline anhy- drous iron alum, having the following composition (Fownes) : 105. HYDROFERRIDCYANIC ACID. (H 3 Cy 6 Fe 2 ; or H 3 Cfdy.) The potassium compound of this acid is obtained by passing a current of chlorine through a solution of ferrocyanide of potassium, until the liquid ceases to give a precipitate with ses- quichloride of iron. Ferridcyanide of potassium is formed by the following reactions : 2(K 2 Cy 3 Fe) + 01 = (K 3 Cy 6 Fe 2 ) + KC1. The alkaline ferridcyanides have a red colour ; that of potas- sium crystallizes in fine right-rhombic prisms; Protosulphate of iron produces with ferridcyanide of potas- sium a fine blue colour (TurnbulPs blue). K 3 (Cy 6 Fe 2 ) + (3FeO,SO 3 )=Fe 3 (Cy 6 Fe 2 ) + 3(KO,S0 3 ). Sesquichloride of iron occasions no precipitate but only gives a darker colour to the solution. Sulphate of copper produces a yellowish-green precipitate. Protonitrate of cobalt a dark reddish-brown precipitate. Nitrate of silver produces a dark orange- coloured precipitate soluble in ammonia. Subnitrate of mercury a brownish-red precipitate. The in- soluble ferridcyanides are decomposed by fusion with alkaline carbonates, soluble ferridcyanides being formed. 166 QUALITATIVE ANALYSIS. 106. HYDROSULPHURIC ACID (SULPHURETTED HYDROGEN). (HS.) This acid is a colourless gas, which under strong pressure becomes liquid; it has a highly offensive odour, resembling that of putrid eggs; its taste is acid, astringent, and bitter; it is in- flammable, burning with a blue flame, aud disengaging sul- phurous acid; it explodes violently when mixed with oxygen or atmospheric air and ignited ; when mixed with chlorine, sulphur is deposited and hydrochloric acid formed ; it is highly delete- rious when inspired, even when mixed with a large quantity of air ; it is absorbed by water, which acquires its peculiar smell aud a nauseous sweet taste ; the solution reddens litmus-paper, and decomposes by exposure to the air, sulphur being depo- sited ; most metallic oxides are decomposed by hydrosulphuric acid, sulphides of the metals, and water, being formed. The alkalies and alkaline earths, and the oxides of chromium, tan- talum, and titanium, do not exchange their oxygen for sulphur in the moist way. The sulphides of the metals of the alkalies, and the alkaline earths, are soluble in water ; they are decom- posed by dilute mineral acids with the evolution of sulphuretted hydrogen, which is readily recognized by its odour and by its action on paper moistened with solution of lead. KS + HO,S0 3 =KO,S0 3 + HS. The sulphides of the metals of the fourth group are insoluble in water, but are likewise decomposed by dilute mineral acids. EeS + HO,SO 3 =FeO,86 3 + HS. All the other metallic sulphides but sulphide of mercury sue de- composed by strong nitric acid, sulphuric acid being formed, and sulphur generally separated ; aqua-regia effects their de- composition more easily, it also dissolves sulphide of mercury. As some metallic oxides, when dissolved in acids, are precipi- tated by sulphuretted hydrogen, while others are not, and as the precipitated sulphides differ in colour, and in other proper- ties, sulphuretted hydrogen becomes a valuable reagent for de- tecting and separating metallic compounds. Most metallic sulphides are decomposed by heat with access of air, evolving sulphurous acid ; the alkaline sulphides thus treated are con- verted into sulphates, When hydrosulphuric acid, or a solution of an alkaline or earthy sulphide, is brought into contact with nitrate of silver or acetate of lead, black precipitates are formed. s + HS = AgS -f HO,N0 5 . INORGANIC ACIDS. 16? Either of these salts are therefore certain tests of the presence of this acid in the gaseous state ; a strip of paper moistened with solution of subacetate of lead is generally employed. Chlorine, bromine, sulphurous acid, nitrous acid, iodic acid, chromic acid, and the sesquisalts of iron all decompose hydro- sulphuric acid with the separation of sulphur. An alcoholic solution of iodine (or an aqueous solution of iodine in iodide of potassium) is converted by hydrosulphuric acid into hydriodic acid with the separation of sulphur. All the compounds of sulphur when mixed with soda and ignited on charcoal before the blowpipe in the reduction flame, yield liver of sulphur, which evolves sulphuretted hydrogen when treated with a mineral acid, and produces a brownish-black spot, when moistened and placed on a piece of polished silver. III. Acids not precipitated by either Chloride of Barium or Nitrate of Silver. 107. NITEIO ACID. (NO 5 .) General characters. This acid has been obtained in an an- hydrous state by passing perfectly dry chlorine gas over crystals of well-dried nitrate of silver at a temperature not ex- ceeding 200 Fahr. The hydrated acid, in its most concentrated form, is a colourless, fuming liquid, very easily decomposed, mere exposure to the light causing it to become yellow, and to disengage oxygen gas. It is one of the strongest of the acids, and constitutes a reagent of the greatest value, from the facility with which it parts with a portion of its oxygen. The yellow fuming acid, containing nitrous acid, possesses generally the greatest oxidizing power. The most highly concentrated acid is, in many cases, without action on bodies on which a diluted acid acts with energy ; thus the concentrated acid does not attack lead or tin, while the addition of a small quantity of water causes an energetic action to take place. Organic sub- stances are for the most part resolved by the concentrated acid into carbonic acid and water, the action in many cases being sufficiently energetic to cause them to take fire ; the diluted acid generally converts them into oxalic, malic, and carbonic acids. Nearly all metallic oxides are dissolved by nitric acid, the exceptions are oxides of tin and antimony, tellurous and tungstic acids. All the neutral salts of nitric acid are soluble 168 QUALITATIVE ANALYSIS. in water, and all are decomposed at a strong red-heat, the nature of the products depending on the nature of the base; thus the alkaline nitrates yield oxygen and nitrogen, while other metallic nitrates give oxygen and an inferior oxide of nitrogen. When nitrates are ignited in the presence of other substances susceptible of oxidation, they take the liberated oxygen, and in some cases the combination is attended by de- flagration ; thus a nitrate thrown on red-hot charcoal causes the latter to throw off brilliant scintillations ; a mixture of a nitrate with cyanide of potassium deflagrates vividly when heated on a platinum plate ; and phosphorus and sulphur brought into contact with a heated nitrate, occasion a violent detonation. 108. Detection of Nitric Acid in solutions of Nitrates. (Solution of Nitrate of Potash (KO,NO 5 ) may be used.) Hydrochloric acid added in excess to nitric acid or to a solu- tion of a nitrate, gives it the property of dissolving gold leaf, in consequence of the liberation of chlorine. When a solution containing a nitrate is mixed with half its volume of concentrated sulphuric acid, allowed to cool, and a crystal of protosulphate of iron added, the liquid round the crystal assumes a reddish-brown colour, in consequence of the following reaction. 10(FeO,S0 8 ) Or, the solution to be tested having been mixed with -a larger volume of concentrated sulphuric acid and allowed to cool, a strong solution of protosulphate of iron is carefully poured upon it, when a violet amethystine-red or blackish-brown colour, according to the quantity of nitric acid present, will be seen at the place of contact, which increases on careful agitation, but disappears on heating. By this mode of operating, and in the absence of chlorine, the smallest traces of nitric acid may be detected. When a solution of a nitrate is concentrated, and heated in a test-tube with copper filings or turnings, and concentrated sulphuric acid, nitric oxide gas is set free, which, combining with the oxygen of the air in the tube, forms ruddy fumes of hyponitric acid. (204(HO,N0 5 )+3Cu==3(OuO,NO 5 )+4HO + N0 2 . When a solution of a nitrate is mixed with sulphuric acid, INORGANIC ACIDS. 169 and a sufficient quantity of a solution of indigo in sulphuric acid added to give it a distinct blue colour, and the mixture heated, the colour is either discharged, or turned yellow, in con- sequence of the oxidation of the indigo at the expense of the nitric acid. If a small crystal of sulphate of manganese be placed in a porcelain capsule, covered with phosphoric acid, and heated to dryness, a colourless residue is obtained ; but if the smallest quantity of nitric acid, or a nitrate be present, it has a violet, crim- son, or dark-red colour, according to the quantity of nitric acid. This is probably the most delicate test of this acid at present known. If a solid nitrate be heated with concentrated sulphuric acid, nitric acid distils over, which produces a white cloud with ammonia, and does not occasion any turbidity when passed into solution of nitrate of silver ; sometimes the brownish-red fumes of hyponitric acid are perceptible, in consequence of the decomposition of nitric acid by the heat. All nitrates when heated on charcoal deflagrate ; violently with cyanide of potassium, in consequence of the decomposition of the cyanogen, carbonic acid and nitrogen being evolved. 109. NITROUS ACID. (N0 3 .) At ordinary temperatures this acid is a gas of a dark yel- lowish-red colour ; at J^ahr. it condenses into a blue and very volatile liquid, which is decomposed when brought into contact with water. Thus : Nitrous acid forms a class of salts called nitrites. The alkaline nitrites are soluble in alcohol ; the corresponding nitrates are insoluble. "When submitted to gentle distillation, the nitrites expel ni- tric oxide gas, nitrates being formed. The neutral nitrites colour the protosalts of iron a blackish- brown. They reduce ter chloride of gold and subsoils of mercury to the metallic state. With nitrate of silver alkaline nitrites produce a sparingly soluble crystalline nitrite of silver. "When a fragment of nitrite of potash is moistened with a so- lution of sulphate of copper, a brilliant-green colour is produced. Acetic acid decomposes solutions of nitrites, the liquid then 170 QUALITATIVE ANALYSIS. striking an olive-green colour with solution of protosulphate of iron. 110. CHLORIC ACID. (HO,C1O 5 .) When coDcentrated, it is a yellowish oily -looking liquid, very sour, reddening, and finally bleaching, litmus-paper ; it is very unstable, being resolved by heat into perchloric acid, oxygen, and chlorine : it converts sulphurous into sulphuric acid, and is at the same time itself reduced to chlorine; it also converts sulphuretted hydrogen into sulphuric acid, sulphur, and water; all its compounds with bases, are soluble in water; they do not possess bleaching properties ; but when mixed with sul- phuric acid they are decomposed, perchloric acid, chlorine, hypo- chloric acid, and oxygen being formed ; the solution becomes yellow, and it then possesses the power of destroying the co- lour of blue vegetable infusions, and by the application of heat decolorizes solution of indigo. The alkaline chlorates, when ignited, disengage oxygen, and become converted into chlorides ; most other chlorates are resolved by heat into metallic oxides, and a mixture of oxygen and chlorine. Concentrated sulphuric acid in contact with the chlorates, is coloured, first brown and then yellow, whilst a very explosive greenish-yellow gas (Cl 4 ) escapes. 3(KO,CLO) + 4(HO,SO 3 )= 2(KO,HO,2S0 8 ) + KO,C10 7 + 2C1O 4 + 2HO. When triturated with sulphur or phosphorus, chlorates deto- nate with dangerous violence ; a mixture of a chlorate with sugar bursts into a flame on being touched with a drop of con- centrated sulphuric acid. The compounds of perchloric acid (HO,C10 7 ) with bases, are distinguished from chlorates by their greater stability, not being decomposed by acids or reducing agents ; their solutions do not therefore become yellow on being mixed with sulphuric acid. The perchtorate ofpotassa is remarkable for its difficult solubility in water. The compounds of hypochloric acid (CIO) with bases have bleaching properties. Bleaching salts (hypochlorite of lime and hypochlorite of soda) are always mixed with chlorides. Chlo- ride of lime contains also some hydrate of lime. They are de- composed on the addition of an acid, with the evolution of chlo- rine only. CaCl+2(HO,S0 3 )==2(CaO,S0 3 ) ORGANIC ACIDS. 171 They are powerful oxidizing agents, and are decomposed by heat and by exposure to light ; they decompose salts of man- ganese, and lead, precipitating from the former, hydrated per- oxide of manganese, and from the latter, first chloride, and then brown peroxide. Pure hypochlorites, when treated with sulphuric acid, evolve hypochlorous acid. If a solution of arsenious acid in hydrochloric acid be coloured blue by solution of indigo, and mixed gradually with a dis- solved bleaching salt, no decolorization of the liquid will take place until all the arsenious acid has been converted into ar- senic acid. On this reaction is founded a method of estimating the value of bleaching-powder. B. ORGANIC ACIDS. I. Acids precipitated by Chloride of Calcium. (Oxalic, Tartaric, Paratartaric, Citric, and Malic Acids.) 111. OXALIC ACID. (H O, C 3 O 3 ; or H O, O.) General characters. It crystallizes in four-sided prisms, which are colourless, very soluble in water, very acid, and poi- sonous. These crystals contain three equivalents of water of crystallization, which when sharply heated they lose, the dry acid subliming; at a high temperature, oxalic acid is decom- posed into water, carbonic, and. formic acids, without blackening, by which it is distinguished from most other organic acids. The alkaline oxalates are soluble in water, as are also some other oxalates with a metallic base ; they are all decomposed at a red-heat, alkaline and earthy oxalates being thereby converted into carbonates. Comportment of solutions of Oxalic Acid, and Oxalates, with reagents. (Solution of Oxalate of Ammonia (NH 4 OO) may be used.) Chloride of calcium and all soluble lime salts produce a white precipitate even in highly dilute solutions (Ca O, O + 2aq), insoluble in water as well as in oxalic and acetic acids, but rea- dily soluble in hydrochloric and in nitric acids ; the presence of ammonia promotes the precipitation of oxalic acid by salts of lime. 172 QUALITATIVE ANALYSIS. Chloride of barium produces a white precipitate (BaO, O + aq) almost insoluble in water, but soluble in nitric and in hydro- chloric acids. Nitrate of silver gives a white precipitate, soluble in nitric acid and in ammonia. Acetate of lead produces a white precipitate. "When heated with concentrated sulphuric acid, oxalic acid and dry oxalates are decomposed, and the oxalic acid resolved into carbonic acid and carbonic oxide gases, HO,C 2 3 = C0 2 +CO + HO, which escape with effervescence. The latter gas may, if present in sufficient quantity, be kindled ; it burns with a blue flame : if the mixture become black, it is a proof that it contains some ! other organic substance. 112. TARTAKIC ACID. (2HO,C 8 H 4 10 ; or 2HO,T.) General characters. This acid crystallizes in large rhombic prisms, which are soluble in water and have a pleasant acid taste ; the solution decomposes by keeping, becoming covered with a mouldiness. "When heated, the crystallized acid loses water, and gives rise to the formation of a series of new com- pounds ; and when treated at a high temperature with a strong solution of hydrate of potassa, it is converted into acetate and oxalate of potassa. Thus : 2HO,C 8 H 4 10 = HO,CjH 8 O s + 2(HO,C 2 3 ). Tartaric acid. Acetic acid. Oxalic acid. The alkaline tartrates are soluble in water; all the salts of tartaric acid that are insoluble in water are easily soluble in hydrochloric acid. Comportment of Tartaric Acid, and solutions of Tartrates, with reagents. (Solution of Tartrate of Soda (2NaO,T) may be used.) Chloride of calcium produces a white precipitate almost in- soluble in water, but soluble in ammoniacal salts, the presence of which, therefore, prevents its formation ; it is soluble also in cold potassa, but excess must be avoided, or hydrate of lime will be precipitated ; if the potassa solution be boiled, tartrate of lime separates as a gelatinous mass, which redissolves as the solution cools. Lime water produces in solutions of neutral tartrates a white ORGANIC ACIDS. 173 precipitate, dissolving in tartaric acid, and also in ammoniacal salts. Chloride of barium produces a white precipitate soluble in dilute acids. Acetate of lead occasions a white precipitate of tartrate of lead, which when ignited out of access of air is decomposed, metallic lead in a fine state of division being formed, which burns when projected into the air. Nitrate of silver produces a white precipitate of tartrate of silver, which by boiling is reduced to a shining mirror, adhering to the glass by agitation ; the reduced metal separates in thin lamina. When a salt of potassa (the acetate answers best) is added to free tartaric acid, and the mixture agitated, a sparingly so- luble crystalline bitartrate of potassa separates ; the addition of alcohol further diminishes the solubility of this salt, which is freely dissolved in alkalies and in mineral acids. Tartaric acid possesses the property of preventing the pre- cipitation of several metallic oxides by alkalies, in consequence of the formation of soluble tartrates not decomposed by al- kalies; among the metallic oxides which it thus affects are alumina, peroxide of 'iron , and. protoxide of manganese. Tartrates, when heated, carbonize, emitting a very peculiar odour ; the tartrates of the alkalies and alkaline earths are thus converted into carbonates. , /:' i PAEATAETAEIC ACID, OB BACEMIC ACID. This acid is isomeric with tartaric acid, from which it is dis- tinguished by the insolubility of its lime salt in excess of acid and in ammoniacal salts, and by the formation after a time, of a precipitate, with solution of gypsum. 113. CiTEicAciD. (3HO,C 12 H 5 O n ; or 3HO,Ci.) General characters. This acid forms large transparent crys- tals, very soluble in water, and of a strong but pleasant acid taste; its solution, like that of tartaric acid, decomposes by keeping. The acid itself carbonizes when heated to redness, emitting a pungent acid vapour ; when heated with sulphuric acid in excess, it is decomposed into carbonic acid, carbonic oxide, acetic acid, and water. Thus : U i2 n 5 u n "r ojcu/ss-awf, -f-zuu, -f- Z(,u 4 n 8 u 3 ;, -r ^JCLW. The alkaline citrates are soluble in water ; the insoluble salts 174 QUALITATIVE ANALYSIS. of citric, and in fact of all organic acids are decomposed by boiling with carbonate of soda, soluble alkaline salts being thus obtained. Comportment of Citric Acid, and solutions of soluble Citrates with reagents. (Solution of Citrate of Soda (3NaO,Ci) may be used.) Chloride of calcium produces, in solutions of citrates, but not in citric acid, a white precipitate of neutral citrate of lime, insoluble in potassa, but readily soluble in sal-ammoniac, from which, however, a basic citrate of lime is precipitated by boiling. Lime water produces no precipitate in the cold ; but by boiling, basic citrate of lime is deposited, which is redissolved as the solution cools ; this reaction serves to distinguish citric acid from most other organic acids. Acetate of lead produces a white precipitate, sparingly solu- ble in ammoniacal salts and in ammonia, but readily soluble in citrate of ammonia. Citric acid, like tartaric acid, prevents the precipitation of certain metallic oxides by alkalies. 114. MALIC ACID. (2HO, C 8 H 4 8 ; or 2HO,M.) This acid, which occurs in several acid fruits, crystallizes with some difficulty, and deliquesces rapidly when exposed to the air ; at the temperature of 230 it gradually decomposes into fumaric acid (2 HO, C 8 H 2 O 6 ,), crystallizing in micaceous scales ; at a higher temperature it is in a great measure con- verted into maleic acid (2 H 0, C 8 H 2 6 ,), which rises as a crys- talline sublimate, fumaric acid remaining in the retort. The salts which this acid forms with bases are mostly soluble in water. Comportment of Malic Acid, and solutions of Malates with reagents. Chloride of calcium does not produce any precipitate till al- cohol is added, when white malate of lime is deposited. Lime water produces no precipitate either in hot or cold so- lutions of malates, by which this acid is distinguished from tartaric, racemic, citric, and oxalic acids. Acetate of lead produces a white precipitate (2PbO,C 8 H 4 O 8 ,6aq), which, by repose, crystallizes in needles, and which ORGANIC ACIDS. 175 fuses under the boiling-point of water. The best test of malic acid is probably its comportment under heat. II. Acids precipitated by Sesquichloride of Iron. (Succinic, Benzoic, Tannic, and Gallic Acids.) 115. SUCCINIC ACID. (2HO,C 8 H 4 6 ; or 2HO,Su.) This acid is crystalline and volatile ; its vapour is exceedingly acrid and penetrating. It is very stable, resisting the action of nitric acid, chlorine, and even of a mixture of hydrochloric acid and chlorate of potash. It combines with anhydrous sul- phuric acid, forming sulpliosuccinic acid, which is tribasie. It was originally obtained from amber, whence its name. The salts which it forms with bases are mostly soluble in water. Comportment of Succinic Acid and solutions of Succinates with (Solution of Succinate of Ammonia (2NH 4 O,C 8 H 4 O 6 ) may be used.) Sesquichloride of iron produces a brownish-red voluminous precipitate, soluble in acids and decomposed by ammonia, hy- drated sesquioxide of iron being formed. Acetate of lead produces a white precipitate of succinate of lead, soluble in excess and in tartaric and acetic acids. As the alkaline and earthy succinates are insoluble in alcohol, the acid after the addition of alcohol and ammonia is precipi- tated by cTiloride of barium. 116. BENZOIC ACID. (HO,C 14 H 5 O 3 ; orHO,5z.) This acid is obtained by sublimation, in the form of light flexible pearly scales, or by precipitation as a crystalline powder. When pure, it has no smell. It is fusible and volatile, its vapour being like that of succinic acid, very irritating. It is not very soluble in cold water, more so in hot, and readily so- luble in alcohol. Its principal source is gum-benzoin. Most of its salts are soluble in water. Comportment of Benzole Acid and solutions of Benzoates with reagents. (Solution of Neutral Benzoate of Ammonia (NH 4 O,Ci 4 H 5 O 3 ) maybe used.) On the addition of hydrochloric acid to the solution of a benzoate in water, benzoic acid separates as a white crystalline powder. 176 QUALITATIVE ANALYSIS. Sesquickloride of iron produces a pale-yellow precipitate, de- composed by ammonia, and by strong acids, which combine with sesquioxide of iron, benzoic acid being precipitated. Acetate of lead does not immediately precipitate free ben- zoic acid ; but it produces a white flaky precipitate in solutions of fixed alkaline benzoates. 117. TANNIC ACID. (C 54 H 22 O 34 , Wrecker.} When pure, this acid is nearly white, but not at all crystalline. It is very soluble in water ; the solution absorbs oxygen from the air, and is converted into gallic acid. It has a most astrin- gent but not bitter taste. It is almost insoluble in ether. It is obtained in a state of purity from the gall nut, of which it constitutes nearly two-thirds of the weight. On the ad- dition of a mineral acid, a precipitate composed of tannic acid, and the acid employed, is determined. Solution of gelatine produces an insoluble curdy precipitate. The whole of the tannic acid in a solution may be removed by a piece of animal membrane. Solution of starch, and most of the vegetable bases also, pro- duce precipitates. Sesquichloride of iron gives a dark blue-black precipi- tate. 118. GALLIC ACID. (3HO,C 14 H 3 O 7 2aq, Strecker.) This acid forms beautiful prisms of a silky lustre, and a slightly yellow colour. It is not very soluble in cold, but dis- solves in three parts of boiling water. Its alkaline solution, when exposed to the air, becomes first yellow, then green, red, brown, and finally nearly bbick, by the absorption of oxygen. It is not precipitated by gelatine ; its solution in hot sulphuric acid is precipitated by water as a reddish-brown, crystalline powder, possessing colouring properties, and which, when heated, yields fine red prisms. By the action of heat it is con- verted into pyrogallic and metagallic acids. Solutions of gallic acid give, with a mixture of the salts of protoxide and sesquioxide of iron, a dark-blue precipitate; with salts of oxide of iron the precipitate is black. ORGANIC ACIDS. 177 III. Acids not precipitated by Chloride of Calcium, or by Sesquichloride of Iron. (Acetic, Formic, Uric, and Meconic Acids.) 119. ACETIC ACID. (HO,C 4 H 3 3 ; or HO,Ac0 3 ; or HO,A.) The normal acid at temperatures below GO is a crystalline solid. It melts at 62 or 63, forming a liquid of a pungent, peculiar, and agreeable smell, and a burning acid taste. It has a powerful action on the skin, on which it raises a blister, pro- ducing a painful sore. It boils at 248, and its vapour is in- flammable. It is decomposed by anhydrous sulphuric acid, and also by chlorine, two new acids, sulphacetic and cJiloracetic acids, being formed. It is also decomposed when passed in a state of vapour through a red-hot tube, the products being car- Ionic acid and acetone. Thus : Acetone. Acetates, when heated, are decomposed with the same trans- formation. The greater number of the acetates are soluble in water. Acetate of silver, and acetate of suboxide of mercury are crystalline, and only sparingly soluble. When heated with dilute sulphuric acid, acetates are decomposed with the libera- tion of acetic acid ; when heated with concentrated sulphuric acid and alcohol, acetic ether is disengaged, which may be known by its peculiar odour. When an acetate is heated with potassa and arsenious acid, oxide of Tcakodyl (C 4 H 6 As,0) highly poisonous, is disengaged. With excess of oxide of lead, acetic acid forms a solution which has an alkaline reaction. Ses- quichloride of iron exhibits no reaction with acetic acid ; but in solutions of neutral acetates, peracetate of iron is formed, which imparts to the solution a blood-red colour. 120. FORMIC ACID. (HO,C 2 H0 3 .) This acid in its most concentrated state, fumes in the air, and has a very pungent acid smell. At a low temperature, it crys- tallizes in brilliant scales. It is highly corrosive, acting power- fully on the skin, and producing painful sores. It boils at 212, and its vapour is inflammable ; all its salts are soluble in water. They have a general resemblance to the acetates, but are, never- theless, quite distinct. When heated to rediness, they give off PART i. N 178 QUALITATIVE ANALYSIS. carbonic acid and carbonic oxide, leaving the metal reduced, or carbonic oxide, leaving a metallic oxide. By the property which this acid and its salts possess, of reducing the oxides of the precious metals, it is distinguished from acetic acid. When nitrate of silver, or subnitrate of mercury are added to con- centrated solutions of alkaline formiates, sparingly soluble pre- cipitates are formed, and on the application of heat, a reduction instantly takes place, carbonic acid and water being formed. Thus : By gently heating chloride of mercury with an alkaline for- miate, it is reduced first to subchloride, and then to metallic mercury. When formic acid is heated with concentrated sulphuric acid, it is decomposed into water and carbonic oxide. Thus : The sulphuric acid withdraws from the formic acid the elements of water, and a transposition of its atoms takes place. The same decomposition occurs, on heating a formiate with sulphuric acid. 121. URIC ACID. (2 HO,C 10 H 2 N 4 4 .) This acid is separated from its compounds by mineral acids as a white crystalline powder. It is very sparingly soluble in water, the solution reddens litmus ; all its salts are likewise very sparingly soluble. When uric acid is dissolved in nitric acid, the solution evaporated to dryness, and ammonia added, a beautiful crimson colour is obtained. This is the characteristic test for uric acid : the substance to be examined is warmed in a watch-glass with a drop of nitric acid, and evaporated to dry- ness at a gentle heat; the residue will have a red colour if uric acid be present, becoming crimson when exposed to the vapour of ammonia. This beautiful compound is murexid or purpurate of ammonia ; on the addition of a drop of caustic potassa a magnificent purple colour is produced, which vanishes on the application of heat. By fusion with alkalies, uric acid disengages ammonia. 122. MECONIC ACID. (3HO,C 14 HO n , 6aq.) This acid, which is found only in opium, crystallizes, when pure, in the form of beautiful white silvery scales. It is so- luble in water and in alcohol. The solution is decomposed by ORGANIC ACIDS. 179 boiling; it is also decomposed by hydrochloric acid, and en- tirely so when heated with excess of potassa ; oxalic acid, car- bonic acid, and a colouring matter being the products. Its dis- tinguishing character is that of causing a deep Hood-red colour in solutions of per salts of iron, without, however, any precipi- tation taking place. 180 CHAPTEE V. ON THE COMPOETMENT OF THE PRINCIPAL ALKALOIDS WITH REAGENTS. Nicotina or Nicotine, the poisonous alkaloid of Tobacco. Conia or Conicine, the poisonous alkaloid of Hemlock. Morphia or Morphine, ~\ Narcotina or Narcotine, > the alkaloids of Opium. Codeia or Codeine, 3 Strychnia or Strychnine, ~) Srucia or Brucine, { the P oisonous a^aloids of Nux vomica. Aconitina or Aconitine, the poisonous alkaloid of Monkshood. Atropia or Atropine, the poisonous alkaloid of the Deadly Nightshade. 123. NICOTINA. (C 10 H 7 N.) General characters. It is a colourless oily liquid, which by exposure to the air becomes first yellow, then brown, and finally solid, by absorbing oxygen. It is inflammable, has an acrid taste, and an extremely irritating odour. It is soluble in water, in alcohol, and in ether ; its solution in alcohol turns turmeric-paper brown. Its salts are soluble in water, and in alcohol, but only very sparingly soluble in ether. Reactions with reagents. Terchloride of gold produces in aqueous solutions a reddish- yellow precipitate. Bichloride of platinum causes a yellow precipitate. Infusion of galls, a copious white precipitate. Chloride of mercury, protochloride of tin, and acetate of lead, white precipitates. Chlorine gas decomposes nicotina with the evolution of hy- drochloric acid gas, and the formation of a blood-red liquid, which, on exposure to the rays of the sun, becomes colourless. Solution of iodine produces with solution of nicotina a ALKALOIDS. 181 reddish-brown precipitate, the colour of which is destroyed by potassa and by ammonia. The volatility of nicotina is its most marked test. 124. CONIA. (C 16 H 16 N.) General characters. This alkaloid is likewise a volatile, oily, colourless liquid, becoming yellow by exposure to the air. It has a powerful odour, like that of fresh hemlock. It is very soluble in alcohol and ether, but sparingly so in water. Its vapour fumes when brought into contact with vapours of ni- tric, hydrochloric, and acetic acids. Its reactions with reagents are very similar to those of nicofina, from which it is distin- guished by its odour and by its sparing solubility in water. 125. MORPHIA. (C S4 H 19 N"O 6 , + 2aq.) General characters. This alkaloid occurs in small, colourless, rectangular prisms ; by heat the two equivalents of water which it contains are expelled, and the alkaloid fuses into a re- sinous mass which on cooling crystallizes. It is soluble ill about 1000 times its weight of cold, and in about 400 times its weight of hot water ; the solution has a bitter taste, and an al- kaline reaction. It is sparingly soluble in cold alcohol, but abundantly so in hot. In ether it is very sparingly soluble ; the fixed alkalies dissolve morphia in considerable quantities, ammonia dissolves it less readily. The salts of morphia are so- luble in water and alcohol, especially in presence of excess of acid, but insoluble in ether. Reactions with reagents. Nitric acid produces with morphia, or its salts, a dark-orarcyd colour, gradually passing into a yellow. A mixture of nitric and sulphuric acids colours morphia green. On adding the alkaloid or one of its salts to a solution of iodic acid, iodine is set free, imparting a brown colour to the liquid ; if some starch be previously added to the iodic acid, the colour becomes blue. Sesquichloride of iron produces a Hue precipitate, soon pass- ing into a dingy green^ and finally to a brown. Bichloride of platinum produces a granular orange-yellow precipitate. Erom a solution of terchloride of gold the metal is gradually reduced. 183 QUALITATIVE ANALYSIS. 126. NABCOTINA. (C 46 H 25 NO 14 2aq.) This alkaloid, which crystallizes in large, lustrous, right rhombic prisms, possesses a very feeble alkaline reaction ; its salts are all acid, and have a very bitter taste. It is nearly in- soluble in water, but abundantly so in alcohol and in ether, especially in the latter, by which property, and by its not strik- ing a blue colour with sesquichloride of iron, narcotina is dis- tinguished from morphia. The most characteristic test is how- ever the following : If a crystal of narcotina, or of either of its salts, be placed in concentrated sulphuric acid, and a drop of nitric acid added, a blood-red colour will be produced, which on the addition of more nitric acid disappears. 127. STRYCHNIA. (C 42 H 22 ]Sr 2 O 4 .) This powerful alkaloid crystallizes in white lustrous prisms, or in octohedra. Cold water does not dissolve more than T~O\O f it s weight, but nevertheless acquires a distinctly bitter taste ; in ether and in absolute alcohol strychnia is insoluble, but in hot dilute alcohol it dissolves, crystallizing out on cool- ing. Its salts are soluble in alcohol. The sulphate, nitrate, and hydrochlorate dissolve in water, their solubility being much increased in the presence of free acid. Strychnia is insoluble in the caustic alkalies, but soluble in the essential oils and in chloroform. Reactions with reagents. Terchloride of gold gives a reddish-yellow precipitate. Bichloride of platinum, a yellow granular precipitate. Infusion of galls, a white precipitate. Sulphocyanide of potassium, a white crystalline precipitate. Nitric acid added to strychnia becomes yellow, if the acid be concentrated, it acquires at the first moment a rose-red tint, soon however becoming light-yellow. Chlorine water gives a white precipitate soluble in ammonia. Bichromate ofpotassa and sulphuric acid give rise to a beau- tiful violet tint, soon fading into a pale rose-colour. " If," ob- serves Otto, " this experiment be executed with the necessary care, the strychnia deposited from one drop of an ethereal solu- tion can be very distinctly recognized ; but to obtain a distinct reaction it is requisite that the amount of bichromate be pro- ALKALOIDS. 183 portional to the quantity of the alkaloid, and this is best at- tained by using the former in the shape of a solid piece, and neither as a powder nor in solution. After solution of the strychnia has been effected, the acid is spread out over the surface of a porcelain dish, and then the crystal of bichro- mate of potassa placed in it. On inclining the dish, violet stripes are noticed flowing from the salt, and on moving the crystal backward and forward by means of a glass rod, the whole of the liquid assumes a rich violet colour. Other oxidizing agents produce a similar effect. If a crystal of strychnia, or either of its salts, be placed on a slip of clean platinum-foil, and moistened with strong sul- phuric acid, and if the foil be then connected with the positive pole of a small galvanic battery (two of Grove's cells are quite sufficient) and the wire from the negative pole be brought into momentary contact with the liquid (taking care that it does not touch the metal), the violet colour will instantly appear. For the physiological test for strychnia, we are indebted to Dr. Marshall Hall. This consists in throwing fresh young frogs into tetanic convulsions by the action of the poison. The back of the animal should be cleaned with blotting-paper, pre- vious to placing them in the liquid to be tested. Dr. Hall states that by this method B0 * 00 of a grain of strychnia may be detected, and Dr. Harley found that when a solution contain- ing only T ^Vo"o f a g raui f acetate of strychnia was injected into the lungs of a very small frog, the animal became vio- lently tetanic in nine and a half minutes, and died in two hours. The best solvent for strychnia in complicated organic fluids is chloroform ; the liquid having been previously made alkaline ; on expelling the chloroform by gentle evaporation, the alkaloid remains behind, and may then be dissolved in dilute hydro- chloric acid, and subjected to the above tests. Animal charcoal possesses the property of removing alkaloids from solutions ; it should never therefore be used as a decolorizing agent. Hofmann and Graham dissolved half a grain of strychnia in half a gall n of beer, and left two ounces of animal charcoal in contact with the liquid for twenty-four hours; the charcoal was then separated by filtration and boiled for half an hour with eight ounces of alcohol, fresh portions being added from time to time to replace the loss by evaporation. The alcohol was then removed by distillation, the residue mixed with caus- tic potassa, agitated with ether, and the ethereal solution allowed 184 QUALITATIVE ANALYSIS. to evaporate spontaneously in a watch glass. On the addition of a drop of sulphuric acid, and a particle of bichromate of potassa, the characteristic violet colour was produced. 128. BRUCIA. (C 46 H 26 N 2 8 ,8aq.) This alkaloid is more soluble in water than strychnia ; it is soluble also in strong alcohol, by which it may be separated from strychnia; it crystallizes in small oblique rhombic prisms which are insoluble in ether ; its taste is intensely bitter ; it is less poisonous than strychnia ; its salts are more soluble in water than those of strychnia. Reactions ivith reagents. Nitric acid produces a bright red colour, which soon becomes yellowish-red, and finally yellow; the addition of protochloride of tin changes the colour to violet. Concentrated sulphuric acid produces a rose-red colour, which soon vanishes. On adding sulphuric acid and bichromate of potash to a solu- tion containing brucia, the blue, violet, and purple colours ob- served in strychnia are not produced. 129. ACONITINA. This intensely poisonous alkaloid generally occurs in whitish granular masses, without any distinct crystalline structure ; it is sparingly soluble in water, more soluble in ether, and readily dissolved by alcohol. Its solution has a strong alkaline re- action, and it forms, with acids, salts which do not crystallize. There are no reliable chemical tests whereby to identify this alkaloid, but Dr. Headland has recommended (' Lancet,' March 29th, 1856) as a physiological test, the application of a spiri- tuous extract of the acid contents of the stomach. If -^ of a grain be obtained it will be sufficient. He states that 3-^ of a grain will poison a mouse with characteristic symptoms ; yi-Q a small bird ; T oVo ^ a g ra ^ n causes tingling and numb- ness of the tip of the tongue; y^o of a grain dissolved in spirit, and rubbed into the skin, causes loss of feeling lasting for some time. 130. A.TEOPIA. (C^ It occurs in white, silky needles, or as a vitreous mass. It is sparingly soluble in water, more so in ether, and dissolving ALKALOIDS. 185 freely in alcohol and in chloroform. Its salts, which are easily soluble in water, are very unstable, suffering gradual decom- position by exposure to the air. Like aconitina, its reactions with reagents show nothing characteristic, but if a few drops of a solution of one part of sulphate of atropia in 9600 parts of water, be introduced into the eye, considerable dilatation of the pupil is produced, lasting from half an hour to an hour. 131. Detection of the Poisonous Alkaloids in Organic Mixtures. The following process, recommended by Stas, has met with the approval of most toxicologists. It is founded on the ob- servation, that the alkaloids form acid salts, which are soluble in water, and in alcohol, and that on decomposing a solution of this kind by means of an alkali, and agitating it with a sufficient quantity of ether, the liberated base dissolves in the ether. If the contents of a stomach, food, etc., are to be examined, they are mixed with double their weight of the strongest alcohol, from 10 to 30 grains of tartaric or oxalic acid are added, and the mass heated in a retort or flask to about 170 Fahr. Or- gans, such as liver, lungs, etc., must be cut into small shreds, moistened with alcohol and acid, then pressed, and this opera- tion several times repeated. When cold, the liquid is strained through filtering-paper, and the residue washed with strong alcohol. The filtrate is evaporated, nearly to dryness, at a temperature not exceeding 100 Fahr.. in a current of air, or still better in vacuo over sulphuric acid. The residue is ex- hausted with cold anhydrous alcohol; the extract is evapo- rated ; the residue is dissolved in the smallest possible quantity of water, and the solution saturated with bicarbonate of soda. The liquid is then agitated with from four to six times its volume of pure, rectified ether ; the alkaloid will be found in the ethereal solution. A small portion is tested for the volatile alkaloids (nicotine, conia,) by allowing it to evaporate spon- taneously in a watch-glass. 132. (1) A Volatile Alkaloid is indicated. In this case the whole ethereal solution is agitated with a strong solution of caustic potassa, and allowed to stand ; the ether which has collected on the surface is then poured off, and agitated with pure dilute sulphuric acid, and allowed to remain at rest for some time ; the volatile alkaloids are hereby con- verted into sulphates and are contained in the aqueous solution ; 186 QUALITATIVE ANALYSIS. this is now to be agitated with caustic soda or potassa and again treated with ether. The ethereal solution contains the alkaloid now in a pure state as regards animal matters, but containing perhaps ammonia, from which it is freed by spontaneous eva- poration, or in vacuo over sulphuric acid ; the ammonia being expelled, the residue is in a state to be chemically examined. 133. (2) A Fixed Alkaloid is indicated. The solution, after saturation with bicarbonate of soda, is completely extracted by repeated agitation with pure rectified ether, and the ethereal solution exposed to spontaneous evapo- ration. The residue, which is distinctly alkaline and contains much animal matter, is dissolved in a small quantity of dilute alcohol, and allowed spontaneously to evaporate; the residue is treated with dilute sulphuric acid and allowed to stand ; the clear liquid is concentrated in vacuo over sulphuric acid, a very concentrated solution of carbonate of potassa is added, and then anhydrous alcohol ; the alcoholic solution on evaporation yields the alkaloid in crystals the nature of which is examined. By this process, Stas has succeeded in isolating from mix- tures with foreign substances, morphia, codeia, strychnia, brucia, veratria, aconitina, atropia, and several other alkaloids. When the presence of opium has to be established, the suspected matters are treated with dilute alcohol and a little hydrochloric acid, the extract is concentrated by evaporation, the residue dissolved in water, filtered, and the filtrate mixed with magnesia in excess, heated to ebullition and then again filtered. The resulting liquid contains meconate of magnesia, which, when acidulated with hydrochloric acid, strikes the characteristic Hood-red colour with sesquichloride of iron ; but as some kinds of opium do not contain meconic acid, it may be advisable to treat the suspected substance for morphia in the manner above described. 187 CHAPTEE VI. ON SYSTEMATIC QUALITATIVE ANALYSIS. 134. IN the Third Chapter the chemical comportment of a great number of substances with reagents has been described. Amongst the bases there have been omitted only a few me- tallic oxides of rare occurrence ; and amongst the acids, nearly all in the inorganic, and a great number of those in the organic kingdom that are of anything like frequent occurrence, have been included. Now when the student has made himself ex- perimentally acquainted with the relations of these substances to reagents, and when he has become familiar with all the leading reactions that have been described, lie is in a position certainly to pronounce a correct opinion as to the nature of almost any arrangement of a single base with a single acid (or element replacing it), that may be presented to his notice ; but if he be called upon to decide upon the nature of a compound, consisting perhaps of a mixture of several substances, it is clear that a mere knowledge of the action of reagents alone, will not serve him, but that he must possess the knowledge also of a method of separating the various substances from each other, and of determining the absence or presence of certain bases and acids : in short, he must know how to proceed on a sys- tematic method of examination, without which chemical analy- sis would be little better than mere guess-work, and its cor- rectness or incorrectness quite a matter of chance. Even in the examination of a compound which is known to consist of a single acid and a single base only, the greatest advantage is derived from the systematic method, a great saving of time and labour is effected, and the result is arrived at in a far more satisfactory manner than it could be by the indiscriminate ap- plication of test after test, following no rule or order, but 188 QUALITATIVE ANALYSIS. merely throwing out baits, as it were, with the hope of eventu- ally meeting with appearances, and of observing phenomena, that may enable us to recognize the presence of some particular substance. In this Chapter our object will be to point out the manner in which this systematic method of qualitative analysis is to be conducted ; and our attention will be confined to those substances and combinations only which are of general occur- rence. The subject will be thus rendered far more inviting, and less embarrassing, to the beginner ; and, after the student has made himself thoroughly acquainted with the principles on which general examinations are conducted, and has acquired a dexterity in conducting the manipulatory processes, he will afterwards find no difficulty in encountering the analysis of compounds in which the rarer substances occur, to the exist- ence of which a careful preliminary examination will, generally speaking, arouse his attention. PRELIMINARY EXAMINATION. 135. Before submitting a compound to analysis in the wet way, there are certain dry operations to be performed upon it which should never be neglected, as they frequently furnish valuable hints for the subsequent examination, and always give some insight into the nature of the substance ; though it must particularly be borne in mind that the absence of many of the reactions about to be described does not necessarily imply the absence of particular bodies, since when several substances are present together, one reaction frequently masks or altogether destroys ;mother. Nevertheless, although the wet analysis of a compound will almost invariably be necessary, the prelimi- nary dry assay is not without its value, as it draws special at- tention to the presence of various bodies, when the operator is performing the systematic analysis. 136. In the first place, the physical characters of the sub- stance are noted, such as its form, density, colour, hardness, etc. It should then be reduced to powder, and subjected to the action of heat, both alone, and also together with certain fluxes. The experiments are conducted in the following manner : I. A small portion of the substance reduced to a state of fine division, is introduced into a tube of hard glass open at both ends, and about one-third of an inch in diameter, and heated in the flame of a gas or spirit lamp. SYSTEMATIC QUALITATIVE ANALYSIS. 189 A. It does not change. No organic matter is present ; the absence of hydrates ; of salts containing water of crystallization; of readily fusible sub- stances, and of volatile substances, is also indicated. B. It changes colour. (1) It is yellow while hot, but becomes white on cooling. Oxide of zinc is probably present. (2) When hot it assumes a yellowish-brown colour, and on cooling a permanent dirty yellow. Binoxide of tin is probably present. (3) While hot the colour is reddish-brown; the colour becomes yellow on cooling. Oxide of lead is probably present. (4) When hot deep-yellow ; dull-yellow on cooling. Oxide of bismuth is probably present. C. It emits gases or fumes, or intumesces. f (1) A splinter of wood, red-hot at the end, burns vividly when introduced into the upper end of the tube (O). Per- oxides, chlorates, nitrates, iodates, and bromates either, or all, may be expected to be present. (2) A smell like that of a burning sulphur-match is recog- nized ; the evolved vapour bleaches litmus-paper. The presence of one of the lower oxides of sulphur is indicated. (3) Vapours of a reddish-brown colour are emitted. The presence of nitrates (especially those of the heavy metals) is indicated. (4) A gas is disengaged which extinguishes the flame of a burning body, and which, when passed into lime water or baryta water causes turbidity. The presence of carbonic acid is in- dicated. (5) A gas is disengaged which burns with a blue flame (CO), the substance not necessarily changing colour. The presence of oxalic acid is indicated ; that acid, and the oxalates, decompo- sing into CO and C0 2 by heat. (6) An inflammable gas, burning with a peach-coloured flame, and emitting the peculiar smell of oil of bitter almonds, is evolved. Some cyanogen compounds are present. (7) A gas emitting a strong odour of rotten eggs is evolved ; a piece of paper moistened with a solution of a lead salt is blackened when introduced into the gas. Hydrosulphuric acid 190 QUALITATIVE ANALYSIS. (sulphuretted hydrogen), arising probably from moist sulphides, is indicated. (8) A pungent gas is evolved which restores the blue colour of reddened litmus-paper, and causes dense fumes when a glass rod moistened with strong acetic acid is brought into contact with it. Ammonia is indicated, which may arise from the de- composition of ammoniacal salts, from the decomposition of nitrogenous organic matters, or possibly from the decomposition of some cyanogen compounds. (9) A vapour rises from the heated substance which con- denses in the cooler parts of the tube. a. The condensed liquid does not affect the colour of litmus-paper : it is pure water. Some salt or salts containing water of crystallization, (alums, borates, alkaline phosphates, etc.,) are present. ft. The condensed liquid has an alkaline reaction. Am- monium compounds are indicated. y. The condensed liquid has an add reaction. Some volatile acid is indicated (sulphuric, sulphurous^ hydrofluoric, hydrochloric, hydrobromic, hydriodic, nitric, etc.). (10) The substance sublimes with, or without odour, and is deposited on the cold sides of the tube. Various volatile salts may be present, also arsenious acid, benzoic acid, etc.; also sulphur (condensing in brown oily drops.) II. A small portion of the substance is heated in a closed tube with carbonate of soda. A. It does not blacken. (1) A metallic mirror is formed on the cooler part of the tube, which cannot be resolved into metallic globules by rubbing it with a glass rod, and which, on further heating, is volatilized, and again sublimed in the form of octahedral crystals. Arse- nical compounds are indicated. (2) A metallic mirror is formed, resolvable into metallic globules by rubbing with a glass rod. Mercurial compounds are indicated. (3) No metallic sublimate, but an aromatic odour of frank- incense emitted. A salt of benzoic acid probably present. (4) No metallic sublimate, but powerfully irritating fumes. Succinic acid is probably present. y. 8. SYSTEMATIC QUALITATIVE ANALYSIS. 191 B. It blackens. (1) A peculiar odour (acetone) is evolved. Presence of acetates. (2) A colourless residue, soluble with effervescence in hy- drochloric acid. One or more of the organic acids in combination with alkalies or alkaline earths may be present. 111. A small portion of the substance is heated on charcoal in the inner blowpipe flame. (1) The substance colours the outer flame a. Golden yellow, indicating sodium. ft. Violet, indicating potassium. . Yellowish-green, indicating barium. . Crimson, indicating strontium. e. Reddish-yellow, indicating calcium. . Green, indicating copper, phosphoric acid, boracic acid. 77. Blue, indicating antimony, lead, arsenious acid, chloride of copper. (2) The substance deflagrates, indicating nitrates, or chlorates. The substance decrepitates, indicating chloride of sodium. The substance intumesces, indicating alums, borates. (3) The substance fuses readily and is absorbed by the char- coal, indicating alkaline salts. (4) The substance does not fuse, but leaves a white residue. Salts of barium, strontium, calcium, magnesium, aluminum, zinc, and silicic acid may be present; if it become highly luminous on ignition, oxides of strontium, calcium, magnesium, and zinc must be looked for. If on moistening the cold residue with nitrate of cobalt and again strongly igniting, it assume a fine blue colour, alumina, or earthy phosphates, are present ; if it become pale-red, magnesia ; if green, oxide of zinc. (5) The substance does not fuse, but leaves a coloured residue. Heat a small fragment on a clear borax bead on platinum wire, in the outer flame of the blowpipe. a. The bead is blue when hot, and blue when cold. Characteristic of oxide of cobalt. ft. The bead is green when hot, and blue when cold. Characteristic of oxide of copper. y. The bead is green when hot, and green when cold. Characteristic of sesquioxide of chromium. 8. The bead is brownish-red when hot, and orange when cold. Characteristic of sesquioxide of iron. 192 QUALITATIVE ANALYSIS. c. The bead is reddish-brown, becoming lighter on cool- ing. Characteristic of oxide of nickel. . The bead is yellowish-brown while hot, the colour be- coming very faint on cooling. Characteristic of oxide of bismuth. rj. The bead is amethyst colour while hot, becoming lighter or altogether vanishing on cooling. Cha- racteristic of oxide of manganese. IV. A little of the substance is mixed with carbonate of soda, or with a mixture of carbonate of soda and cyanide of potassium, and heated in the inner blowpipe flame. (1) The substance is reduced to the metallic state without in- crustation. a. The metal appears in white shining scales. Silver ; tin. /?. The reduced metal is red. Copper, y. The reduced metal is yellow. Gold. (2) The substance is reduced to the metallic state with incrus tat ion. a. The reduced metal is brittle, the incrustation is white. Antimony. /?. The reduced metal is brittle, the incrustation is brown-yellow. Bismuth, y. The reduced metal is not brittle, the incrustation is yellow. Lead. (3) The substance is not reduced to the metallic state, but it furnishes an incrustation. a. The incrustation is white. Oxide of zinc. ft. The incrustation is brownish-red. Oxide of cad- mium. y. The incrustation is white, and garlic-smelling fumes are emitted. Compounds of arsenic. (4) An alkaline sulphide is left which stains metallic silver black. Compounds of sulphur. 137. If the substance to be analysed be a liquid, the first step is to evaporate a portion of it to dryness on plati- num foil, in order to see whether it really contain anything in solution, and then to ascertain, by means of test-papers, whether it has an acid, or alkaline, or neutral reaction ; if the first, it does not necessarily follow that the solution con- tains free acid, for it must be borne in mind that saline SYSTEMATIC QUALITATIVE ANALYSIS. 193 solutions of many of the heavy metallic oxides redden litmus- paper. It is not difficult, however, to decide whether the acid reaction is occasioned by a metallic salt or by a free acid ; the addition of a drop or two of solution of carbonate of po- tassa or soda generally, in the first case, renders the liquid turbid, while, in the latter case, it remains clear. If the solution be alkaline, it can contain no metallic oxides that are insoluble in alkaline liquids ; and if it be perfectly neutral, it probably con- sists only of salts of the alkalies, or alkaline earths. 138. These preliminary experiments having been made, and some insight into the general nature of the substance under examination having thereby been obtained, its relation to sol- vents is next to be examined. It is reduced to a fine state of division by trituration in an agate mortar, and a portion di- gested for some time in a test-tube with distilled water ; if it be entirely dissolved, the qualitative analysis of the aqueous solution may at once be proceeded with, taking about twenty- five or thirty grains of the substance. If a complete solution do not take place, a drop or two of the clear liquid is to be eva- porated to dryuess on platinum foil, in order to see whether water has dissolved anything. It the evaporated liquid leave a re- sidue, then the substance must a second, and a third time, be di- gested w r ith fresh portions of distilled w r ater, filtered, and the filtrate set aside for subsequent examination. The undissolved residue is next gradually heated to boiling, with dilute hydro- chloric acid, and if complete solution be not effected, the fluid is decanted, and the residue digested with concentrated hydro- chloric acid, particular attention being paid to the nature of the gases evolved. If an effervescence take place, carbonic acid, sul- phuretted hydrogen, and hydrogen gases may be looked for ; the first betrays the presence of carbonates, the second that of sul- phides, the third that of certain metals ; the evolution of chlo- rine indicates peroxides or chromates, and hydrocyanic acid points out the existence of certain cyanides. If complete solution of the substance be not effected by concentrated hydrochloric acid, a fresh portion is to be digested with nitric acid; a few of the metals only, escape solution by this acid (gold, pla- tinum, palladium, etc.) ; tin and antimony are oxidized by it, but the oxides are not dissolved ; it moreover decomposes all sulphides, with the single exception of sulphide of mercury, setting free sulphur, which is easily recognized by its colour and lightness, and its being completely evaporated when heated PART I. O 194 QUALITATIVE ANALYSIS. on a strip of platinum foil. If nitric acid fail to dissolve the substance completely, aqua-regia maybe tried, and if this leave a residue, recourse must be had to fusion with an alkali or an alkaline carbonate, the following preliminary experiments being first performed upon it. (1) A small portion is heated on a slip of platinum foil ; the odour of sulphurous acid is conclusive as to the presence of sul- phur, and, if no other substance be present, the whole will be dissipated by a very moderate heat ; in this case, a prolonged digestion of the residue in aqua-regia will, generally speaking, give a clear solution, the whole of the sulphur becoming oxi- dized into sulphuric acid. (2) A small quantity is moistened with sulphide of ammo- nium ; if it immediately become black, an insoluble silver, mer- cury, or lead salt is present. (3) A small portion is mixed with carbonate of soda, and heated in the inner blowpipe flame ; a metallic reduction, ac- companied by a yellow incrustation, indicates lead. (4) A small quantity is digested with carbonate of potassaor soda ; if it become black, subchloride of mercury is present, which is further confirmed by mixing a portion of the residue with slightly moistened carbonate of soda, and heating it in a glass tube ; small globules of mercury are easily recognized with the aid of a lens. (5) Another small quantity may be digested with ammonia, and the clear ammoniacal liquid supersaturated with nitric acid ; the formation of a white precipitate, insoluble in the nitric acid, indicates silver. Besides insoluble compounds of silver, lead, and mercury, the residue insoluble in aqua-regia, may contain sulphates of the alkaline earths, silicates, fluorides, certain phosphates and arse- niates, and the insoluble modifications of oxides of tin, antimony, and chromium. Sesquioxide of iron also is dissolved only with great difficulty, after having been strongly ignited. All these compounds may, however, be brought into a soluble state by fusion with carbonate of soda, or better with a mixture of car- bonate of soda and carbonate of potassa, or of carbonate of soda and cyanide of potassium, their constituents being thereby transposed in such a manner as to admit of their easy detection by a proper subsequent treatment. For example, if the ori- ginal substance undissolved by nitric acid, or by aqua-regia, con- tain sulphates of the alkaline earths, the result of a fusion with SYSTEMATIC QUALITATIVE ANALYSIS. 195 carbouate of soda is to bring them to the state of carbonates, which, though insoluble in water, are readily dissolved by dilute hydrochloric acid, and the bases are then easily detected in the hydrochloric solution : the sulphuric acid with which these bases were previously combined, is found in the aqueous solution of the fused mass, in the form of sulphate of soda. Again, if the insoluble residue contain chloride, iodide, or bromide of silver or lead, the result of the fusion with an alkaline carbonate, which in this case must be performed in a porcelain, and not in a pla- tinum crucible, is to transfer the chlorine, bromine, or iodine to the metal of the alkaline base, forming compounds easily solu- ble in water, and therefore easy of detection in the aqueous so- lution of the fused mass ; while the metal or metals with which the chlorine, iodine, or bromine was originally combined, being of course insoluble in water, may be dissolved in nitric acid, and detected by the appropriate tests. If baryta, strontia, and lime be present, besides the oxides of silver and lead, the two latter metals are precipitated from the nitric solution by sulphuretted hydrogen : the liquor filtered from the precipitate, is nearly neutralized with carbonate of soda, and mixed with hydrofluosilicic acid, by which the baryta, after a time, is preci- pitated. To the liquor filtered from this precipitate, solution of sulphate of potassa is added, which precipitates the strontia in company, perhaps, with a little of the lime, and on filtering off the precipitated sulphate of strontia, adding to the filtrate am- monia, and then oxalate of ammonia, a white precipitate, indi- cating the presence of lime, is determined. Silicic acid, if pre- sent, passes into the state of silicate of soda, and is discovered by adding excess of hydrochloric acid to the aqueous solution of the fused mass, evaporating to perfect dryness, and treating the residue with water ; by this treatment the silicic acid is brought to its insoluble condition, and remains undissolved. 139. It thus appears that when we have before us a complex substance, which is partly soluble in water, and not wholly dis- solved by either hydrochloric acid, nitric acid, or aqua-regia, three distinct analytical operations may be performed on it, viz.: 1. An analysis of the aqueous solution. 2. An analysis of the acid solution. 3. An analysis of the insoluble residue. If the operator be desirous of obtaining a precise knowledge of the mode of arrangement of the different acids and bases in- such a complex mixture, he must make these three distinct analyses ; but, if his object be only to ascertain what acids and 196 QUALITATIVE ANALYSIS. bases are present, he may generally omit a distinct analysis of the aqueous solution, and after a careful preliminary examina- tion, he may confine his attention to the acid solution (the sub- stance not having been previously exhausted by water), and to the mass obtained by fusion of the residue with an alkali. I. Examination for Bases : principles upon which the method depends. 140. In the Third Chapter, it was shown that by means of certain general reagents, metals may be arranged into groups ; some of which, by other reagents, may again be subdivided into minor groups, each of the individuals composing which may be recognized by their comportment with certain other reagents, which may be called special. 141. The general reagents are hydrochloric acid, hydrosul- phuric acid (sulphuretted hydrogen}, sulphide of ammonium, ammonia, and the carbonates of soda, ammonia, and potassa ; they are of the greatest use to the analytical chemist in conducting a systematic qualitative analysis. Thus, if on the addition of hydrochloric acid to a solution, no precipitate or turbidity should occur, the operator is assured that no salt of silver or of suboxide of mercury is present ; neither can the liquid con- tain any considerable quantity of a salt of lead ; again, if on the addition of sulphuretted hydrogen water in excess, no alteration should be perceptible, no member of the fifth group can be present ; and if, after having neutralized the liquid with ammonia, no precipitate, or decided change of colour should occur, on the addition of sulphide of ammonium, even on the application of heat, absence of all members of the fourth group is indicated, and the solution can contain no bases but some one or more of the first or second groups ; and lastly, the addi- tion of an alkaline carbonate will show at once, whether any of the alkaline earths are present. If this reagent should give no precipitate, then the solution in question can contain no metallic oxides but magnesia, and those contained in the first group; and {{'phosphate of soda should occasion no precipitate or turbi- dity, even after agitation and standing, it must be a solution of some salt, or salts, of one or more of the alkalies proper. 142. Let us assume, however, that the reagents have given indications of the presence of some one or more of the mem- bers of all the groups ; the solution must then be submitted to a careful analysis, in accordance with the following scheme. SYSTEMATIC QUALITATIVE ANALYSIS. 197 A. Hydrochloric acid has produced a precipitate, which does not disappear when the acid is added in excess, and the solution warmed. The precipitate must be chloride of SILVER, chloride of 'LEAD, subchloride of MERCURY, either or all, possibly also benzoic acid and uric acid; and if the acid has not been added in sufficient quantity and the solution wanned, oocychlorides of BISMUTH, ANTIMONY, and TIN; but as the absence of all these bodies is secured by heating the well-acidified solution, regard need only be had to SILVER, MERCURY, and LEAD. Analysis of the Precipitate by Hydrochloric Acid. It is thrown on a filter and washed thoroughly with boiling water. (1) The washings may contain chloride O/*LEAD, which, if pre- sent in quantity, crystallizes out as the solution cools in white needles. Confirmation. Sulphuric acid produces a ichite precipi- tate ; bichromate of potassa produces a, yellow precipitate. LEAD is present. (2) The insoluble portion on the filter may be chloride oj SILVER and subchloride of MERCURY ; wash the precipitate from the filter with ammonia, add that reagent in excess, and allow the insoluble portion (if any) to subside. The solution may contain chloride of SILVER. Confirmation. Dilute nitric acid added in excess pro- duces a white curdy precipitate. SILVER is present. The insoluble -black residue is subchloride of mercury. Confirmation. 3Jry at a gentle heat ; mix with dry car- bonate of soda, and heat the mixture in a bulb tube of hard glass ; metallic g lobules are deposited in the upper part of the tube. MERCURY is present. B. Hydrosulphuric acid water (sulphuretted hydrogen) pro- duces with the filtrate from the precipitate by hydrochloric acid (A) a coloured precipitate ; thoroughly saturate the solution with hydrosulphuric acid gas. The precipitate may contain sulphides of MERCURY, LEAD, BISMUTH, COPPER, CADMIUM, TIN, ARSENIC, ANTIMONY (GOLD and PLATINUM); the last two metals are to be tested for spe- cially, in a portion of the original solution. Throw on a filter ; wash the precipitate until it is quite 198 QUALITATIVE ANALYSIS. free from hydrochloric acid ; transfer the precipitate to a flask, boil it for some time with solution of caustic potassa, and filter. (1) Analysis of the portion of the precipitate by Hydrosulphuric Acid soluble in Caustic Potash. It may contain ANTIMONY, ARSENIC, and TIN. Add hydrochloric acid in excess, and thoroughly saturate with hydrosulphuric acid. The above metals are re-precipitated as sulphides. Filter and wash. Dissolve in as little aqua-regia as possible, and introduce the solution into a gas-evolution bottle, together with a slip of zinc, and some dilute sulphuric acid, (both free from arsenic) gas is evolved, which is washed by pass- ing through a second bottle containing a dilute solution of acetate of lead, and is conducted into a precipitate glass con- taining a solution of nitrate of silver. If a precipitate occur, allow the gas to pass until no further effect is produced. This precipitate may contain ANTIMONY : collect it on a filter, wash thoroughly, and then boil it with tartaric acid ; filter, aci- dify the filtrate with hydrochloric acid, and saturate with hy- drosulphuric acid. An orange-coloured precipitate is produced. ANTIMONY is present. The filtrate from the precipitate formed in the nitrate of silver may contain ARSENIC. Add a few drops of acetate of A light-yellow precipitate is formed (arsenite of silver). ARSENIC is present. If any TIN be present it will have been precipitated on the slip of zinc in the gas-evolution bottle, in the form of a black powder, which is easily removed from the zinc. It should be dissolved in a small quantity of strong hydrochloric acid, diluted with water and mixed with a little solution of chloride of mercury. A grey precipitate is formed (metallic mercury). TIN is present. (2) Analysis of the portion of the precipitate by Hydrosulphuric Acid insoluble in Caustic Potash. It may contain MERCURY, LEAD, BISMUTH, COPPER, and CAD- MIUM. Thoroughly wash ; then boil with strong nitric acid as long as red fumes are evolved ; add dilute sulphuric acid as long as a precipitate continues to form. It" this precipitate be white, it contains probably sulphate of lead; if black, it contains sul- SYSTEMATIC QUALITATIVE ANALYSIS. 199 phide of mercury ; if pale-yellow and flocculent, it is probably only sulphur. Filter, wash, and digest with acetate of ammonia, and filter ; add to the filtrate acetic acid in excess, and then solution of chromate of potash. A yellow precipitate is formed. LEAD is present. The residue insoluble in acetate of ammonia (if black) is sulphide of MERCURY. Confirmation. Dry at a temperature not above that of boiling water, mix with dry carbonate of soda, and heat in bulb tube of hard glass. Metallic globules are deposited in the upper part of the tube. MERCURY is present. The portion soluble in nitric acid may contain COPPER, CADMIUM, and BISMUTH. Digest with excess of ammonia, and filter. The solution may contain CADMIUM and COPPER; add hy- drochloric acid in excess, and then saturate with hydrosulphuric acid. Filter ; wash with hydrosulphuric acid water ; boil with dilute sulphuric acid, and filter rapidly. The filtrate may contain CADMIUM. Thoroughly saturate with hydrosulphuric add. A. yellow precipitate is formed. CADMIUM is present. The residue on the filter (if any) is sulphide of COPPER. Dis- solve in nitric acid and add excess of ammonia : a blue solution is formed. Confirmation. Add acetic acid, and then ferrocyanide of potassium : a chocolate-coloured precipitate is produced. COPPER is present. If ammonia has failed to redissolve the whole of the portion insoluble in caustic potash, but soluble in nitric acid, the in- soluble portion is probably BISMUTH. Confirmation. Eedissolve in dilute nitric acid, add a few drops of hydrochloric acid, evaporate nearly to dryness, and then add a large quantity of water : the liquid becomes milky. BISMUTH is present. C. Analysis of the filtrate from the precipitate occasioned ly hydrosulphuric acid. It may contain all the metals of ihefirst, second, third, and 200 QUALITATIVE ANALYSIS. fourth groups, likewise phosphoric acid in combination with sesquioxides of iron, chromium, and aluminum, and with baryta, strontia, lime, and magnesia. Evaporate till the smell of sulphuretted hydrogen has dis- appeared ; add a few drops of concentrated nitric acid, and continue the evaporation to perfect dryness. Should organic matter, or any of the organic acids have been indicated in the preliminary examination, the dry residue must be heated to redness to destroy them, since their presence in the solution to be analysed would prevent the precipitation of iron and alu- mina, but in the absence of organic matter, ignition of the residue may be dispensed with ; in any case, it should not be heated too strongly, as certain oxides (sesquioxide of iron and sesquioxide of chromium} are dissolved with difficulty after having been ignited. Warm the dry, or the ignited residue with concentrated hydrochloric acid, then add water, and heat until it is entirely redissolved, or at any rate until no further action takes place. Should a white residue be left, it is pro- bably silicic acid, possibly sulphates of baryta and strontia; should the residue be coloured, it probably contains sesquioxide of iron or sesquioxide of chromium. A. little experience will enable the operator to decide whether the insoluble residue re- quires further examination ; should he deem such to be neces- sary, he must proceed with it according to the instructions given under E. Filter from the insoluble residue (if any), add abundance of chloride of ammonium, then ammonia in ex- cess, warm gently, and filter rapidly. I. Analysis of the portion precipitated by Ammonia. It may contain sesquioxides of IKON, CHROMIUM, and ALUMI- NUM, also the phosphates of those sesquioxides : it may also con- tain the phosphates of BABYTA, STRONTIA, LIME, and MAGNESIA. The further treatment of the precipitate depends upon the pre- sence or absence of phosphoric acid. Dissolve a small portion in hydrochloric acid, and test with molybdate of ammonia. o. Phosphoric Acid is present. Boil the rest of the precipitate with oxalic acid ; when cold, add acetate of potash, shake well, allow to stand for some time, and then filter. The insoluble portion may contain BARYTA, STRONTIA, and LIME, now in combination with oxalic acid. It is collected upon a filter, well washed, and then dried and ignited, by SYSTEMATIC QUALITATIVE ANALYSIS. 201 which the oxalates are converted into carbonates : the carbo- { nates are dissolved in hydrochloric acid, and the solution is j mixed with that to be examined under D. The soluble portion may contain the sesquioxides of IRON, } ALUMINUM, and CHROMIUM ; MAGNESIA and PHOSPHORIC ACID. Boil for a few minutes with caustic potash, and filter. All he phosphoric acid passes into the alkaline solution in combi- lation with potassa and (perhaps) alumina, but as this acid is mown to be present, alumina is the only substance the pre- ence of which requires confirmation. Dissolve in hydrochlo- ric acid, then add slight excess of ammonia. A colourless gelatinous precipitate is produced. ALUMINUM is present. The portion insoluble in caustic potash may contain the ses- quioxides of IRON and CHROMIUM, and MAGNESIA : wash and boil vith chloride of ammonium. The solution may contain magnesia. Confirmation. Add ammonia and phosphate of soda, a white crystalline precipitate is formed. MAGNESIUM is present. The portion insoluble in chloride of ammonium may contain esquioxide of IRON, and sesquioxide of CHROMIUM. Fuse a >ortion of the dried residue w r ith carbonate of soda and nitre on latinum foil. Dissolve the fused mass in hot water and filter, 'he solution is yellow. Confirmation for Chromium. Acidulate with acetic acid, and add solution of acetate of lead : a yellow precipitate is produced. CHROMIUM is present. If anything be left insoluble in water, on the platinum foil, it s probably sesquioxide of IRON. Confirmation. Dissolve in hydrochloric acid, add acetate of potash, and a few drops of ferrocyanide of potassium : a blue precipitate is formed. IRON is present. . Phosphoric Acid is not present. Boil the remainder of the precipitate with caustic potash, and filter while hot. The alkaline solution may contain alumina. Confirmation. Add slight excess of hydrochloric acid, then ammonia, and heat : a white gelatinous precipitate is produced. ALUMINUM is present. 202 QUALITATIVE ANALYSIS. The portion insoluble in caustic potash may contain sesqui- oxide of IRON and sesquioxide of CHROMIUM. Examine for these oxides as under (a). II. Analysis of the portion not precipitated by Ammonia. It may contain NICKEL, COBALT, ZINC, and MANGANESE ; also all the metals of the first and second groups. Add excess of sulphide of ammonium, boil, and filter : a precipitate is formed. o. Treatment of the Precipitate occasioned by Sulphide of Ammonium. It may contain NICKEL, COBALT, ZINC, arid MANGANESE, as sulphides. Wash thoroughly ; redissolve in as little nitro-hydro- chloric acid as possible ; dilute (if necessary) ; filter ; then add caustic potassa (free from iron and lead) in slight excess, and boil. The alkaline solution may contain ZINC. Confirmation. Add a little sulphuretted hydrogen water : a white precipitate is produced. ZINC is present. The portion insoluble in caustic potassa, may contain oxides of COBALT, NICKEL, and MANGANESE. Thoroughly wash; re- dissolve in hydrochloric acid ; nearly neutralize with ammonia ; add excess of acetate of potassa; pass a stream of sulphuretted hydrogen quickly through the solution, and filter. The solu- tion may contain manganese. Confirmation. Add chloride of ammonium, ammonia, and sulphide of ammonium. A. flesh-coloured precipitate is produced. MANGANESE is present. Confirm also by blowpipe. The precipitate may contain COBALT and NICKEL as sulphides. Thoroughly wash ; redissolve in a little aqua-regia ; dilute with water ; nearly neutralize with carbonate of soda, then add a quantity of a solution of pure cyanide of potassium sufficient to redissolve the precipitate which is at first formed. Boil briskly for some time, till the evolution of hydrocyanic acid ceases ; filter (if necessary) ; when cold, add a strong solution of hypo- chlorite of soda ; allow to stand for some minutes, when, if no black precipitate or coloration appear, add a drop of dilute hydrochloric acid, and heat gently. If a precipitate takes place, it is sesquioxide of NICKEL. Confirm by blowpipe. The solution may contain COBALT in the form of cobalticya- SYSTEMATIC QUALITATIVE ANALYSIS. 203 nide of potassium (H 3 Co 2 Cy 6 ). Evaporate to dryness and confirm by the blowpipe. D. Analysis of the filtrate from the precipitate by sulphide of ammonium. It may contain the ALKALINE EARTHS, and the ALKALIES. Add ammonia and carbonate of ammonia, warm and filter. I. Treatment of the Precipitate produced by Carbonate of Ammonia. It may contain BARYTA, STRONTIA, and LIME, in the form of carbonates. Thoroughly wash ; add a small quantity of hydrochloric acid to the washed residue, and rinse the filter with hot distilled water. To a small portion of the clear solution, add a few drops of solution of sulphate of lime. o. No Precipitate is produced even after standing for some time. Neither BARYTA nor STRONTIA can be present, but LIME may be present. Confirmation. To a small portion of the hydrochloric solution of the precipitate by carbonate of ammonia, add ammonia in excess, and oxalate of ammonia. A white pre- cipitate is produced. LIME is present. /3. Sulphate of Lime occasions no immediate Precipitate, but after remaining for some time in the cold, a Precipitate is formed. BARYTA is absent, STRONTIA is present, possibly LIME also. To another portion of the hydrochloric solution of the precipi- tate by carbonate of ammonia, add strong solution of sulphate of potash, and evaporate to dryness. Add a small quantity of water to the residue, and filter. Add to the filtrate, ammonia and oxalate of ammonia. A white precipitate is produced. LIME is present. 7. Sulphate of Lime occasions an immediate Precipitate. BARYTA is present, and possibly STRONTIA and LIME. To the remaining portion of the hydrochloric solution of the pre- cipitate by carbonate of ammonia, add hydrofluosilicic acid, and warm ; baryta is thereby precipitated in the form of hydrofluo- silicate of baryta. Evaporate to dryness ; pour methylated spirit on the residue ; filter ; evaporate to dryness, and re- dissolve in water. Divide the solution into two parts. To 204 QUALITATIVE ANALYSIS. one portion add sulphate of lime. A white precipitate is formed after a time. STRONTIA is present. Dilute the other part of the solution with water, add a drop of ammonia and oxalate of ammonia. A white precipitate is produced. LIME is present. II. Treatment of the Filtrate from the Precipitate by Carbonate of Ammonia. It may contain MAGNESIA, and the ALKALIES. Add a few drops of carbonate of ammonia, to be assured that everything precipitable by that reagent has been removed. Should the solution remain clear, to a small portion add ammonia and phos- phate of soda, and agitate. A white crystalline precipitate is formed. MAGNESIA is present. Evaporate the remainder of the solution to dryness, and ig- nite till ammoniacal fumes cease to be evolved. Heat a small portion of the ignited residue on platinum wire, in the inner flame of the blowpipe. The outer flame is tinged yellow. SODA, is present. Bedissolve the remainder of the ignited residue in a little hot water ; add a drop of hydrochloric acid, arid a few drops of bichloride of platinum, and stir. A yellow crystalline precipi- tate is formed. POTASSA is present. Ammonia, if present, will have been formed in the prelimi- nary examination. E. Analysis of the portion insoluble in water and in acids. It may contain LEAD, BARYTA, STRONTIA, and LIME, as sul- phates ; LEAD and SILYEB, as chlorides ; it may also contain the insoluble modifications of the sesquioxides of ALUMINUM, IRON, and CHROMIUM ; silicates, aluminates, and fluorides ; and binoxide of TIN, and certain phosphates, arseniates, and antimo- niates. A small portion is first submitted to a careful pre- liminary examination by the blowpipe, whereby a general knowledge of its nature is obtained. It is then fused with three or four times its weight of a mixture of carbonate of pot- assa and carbonate of soda : the fusion may be made in a plati- num crucible, unless the preliminary examination baa indicated SYSTEMATIC QUALITATIVE ANALYSIS. 205 the presence of some one or more of the heavy metals, in which case a porcelain crucible must be used. The fused mass, when cold, is boiled for some time with distilled water, thrown on a filter, and the insoluble portion thoroughly washed. I. Treatment of the portion insoluble in boiling water. It may contain the bases mentioned above, some as oxides, and some as carbonates ; and silicic acid. It is redissolved in hydrochloric acid (unless the preliminary examination has in- dicated the presence of lead or silver, in which case nitric acid must be used). The ac'd solution is examined systematically, according to the instructions already given. II. Treatment of the portion soluble in boiling -water. It may contain certain inorganic acids, possibly also some oxide O/"TIN, oxide of ANTIMONY, oxide 0/iiioN, and LIME. The solution is divided into two parts (a, /?). a. Tin, antimony, iron, and alumina are specially tested for. ft. The inorganic acids are tested for. II. Preliminary Examination for Acids. 143. In the general preliminary examination of the com- pound under analysis, the following acids will (if the experi- ments have been carefully conducted) probably have been de- tected, viz. hydrosulphuric acid, sulphurous acid, carbonic acid, chromic acid, arsenious and arsenic acids, silicic acid ; the acids of chlorine, bromine, iodine, and fluorine ; the acids of cyanogen and of nitrogen ; oxalic, acetic, and some other of the organic acids. Nevertheless, when several of these acids exist toge- ther, it is quite possible that some may be missed even by the most experienced analyst. A careful special testing should always, therefore, be had recourse to, and the opera- tor must consider, by reference to the nature of the bases already discovered, what acid can possibly be in combination with them. I. Treat a portion of the substance in a test-tube with con- centrated sulphuric acid. Should a greenish-yellow gas, having the odour of chlorine, be evolved, it indicates the presence of one of the oxygen compounds of chlorine. This gas is highly ex- plosive, and therefore the application of heat must be avoided ; in the absence of this gas, apply heat. A. It becomes black, evolving gases (amongst which sul- 206 QUALITATIVE ANALYSIS. phurous acid can be recognized), and a strong odour of burnt sugar. Tartaric and malic acids must be specially sought for. B. It is decomposed without changing colour, evolving gas, but no odour of burnt sugar. a. The gas burns with a Hue flame (carbonic oxide) and renders lime-water turbid (carbonic acid). The presence of oxalic acid is indicated. /?. The gas burns with a blue flame, and restores the blue colour of reddened litmus-paper (ammonia). The acids of cyanogen must be specially tested for. y. A gas or vapour is evolved which is not inflammable, and the substance is not blackened. Carbonic, sulphurous, hydrosulphuric, hydrochloric, hydrofluoric, nitric, acetic, benzoic, and succinic acids may be present. 8. A volatile vapour is evolved having an agreeable odour, and burning with a clear white flame (acetone). Acetic acid is clearly indicated. e. A vapour is evolved which corrodes glass. Hydro- fluoric acid is present. II. Heat a portion of the substance with sulphuric acid and binoxide of manganese. a. Chlorine is evolved. (3. Violet vapours are evolved which colour starch-paste blue. Iodine is indicated. y. Dense red vapours are evolved which colour starch- paste orange. Bromine is indicated. III. Heat a portion of the substance with sulphuric acid and copper filings, brown-red fumes are evolved. One or more of the acids of nitrogen are present. IV. Heat a portion of the substance with dilute hydrochloric acid. a. A colourless gas is evolved which renders lime water milky. Carbonic acid is present. /8. A colourless gas is evolved which blackens lead-paper, and which possesses the odour of rotten eggs. Hydrosul- phuric acid is present. y. A colourless gas is evolved, possessing the smell of burning sulphur. Sulphurous acid is present. 8. Red vapours are evolved. Nitric acid is present. SYSTEMATIC QUALITATIVE ANALYSIS. 207 General Rules for the Preparation of the Solution to be tested for Acids. 144. Since many bases would be precipitated by the reagents employed to detect the presence of acids in the wet way, as a feneral rule, every metallic oxide, but those included in Group . (the alkalies) is removed from a solution before proceeding to test it for acids. This may nearly always be effected by boiling with carbonate of soda (1) in slight excess. In some cases, it may be necessary to remove the oxides of the heavier metals by hydrosulphuric acid and sulphide of ammonium. (2) Chromic acid, if present, should be reduced by hydro- sulphuric acid, and the oxide separated by ammonia. (8) Silicic acid should be separated by evaporating to dry- ness, with excess of hydrochloric acid, redissolving the residue in dilute hydrochloric acid, and neutralizing by ammonia. NOTE. In this case the volatile acids viz. hydrocyanic, hydrobromic, hydriodic, acetic, formic, succinic, and benzoic acids cannot be found in the solution. Acetic acid may be tested for specially with sulphuric acid and alcohol. The pre- sence of silicic acid does not interfere with the testing for suc- cinic and benzoic acids by sesquichloride of iron, nor with that of hydrocyanic acid, hydrobromic acid and hydriodic acid by nitrate of silver. (4) If the substance be insoluble in water, but soluble in acids, and if it has been found by the preliminary examination to contain an organic acid, or hydrocyanic, hydrobromic, or hy- driodic acid, it should be boiled with carbonate of soda ; the base remains in the residue either free, or combined with car- bonic acid, and the acid is dissolved in combination with soda. The solution should be exactly neutralized by nitric acid pre- vious to testing for acids. (5) If the solution to be tested for acids contain ammoniacal salts, boracic, tartaric, and citric acids cannot be recognized by the usual reagents. In such a case, the boracic acid may be tested for specially ; or the ammoniacal salts may be removed by long-continued boiling with potassa, and the solution after- wards neutralized by nitric acid. (6) If the substance be insoluble in water, but soluble in acids, and not a compound of an organic acid, or of hydrocy- anic, hydrobromic, or hydriodic acids, a part may be dissolved in hydrochloric acid, and tested for all acids but hydrochloric, 208 QUALITATIVE ANALYSIS. and another part in nitric acid, and tested for hydrochloric acid by nitrate of silver. (7) Carbonic, hydrostdphuric, and sulphurous acids should be separated, if present, by a strong acid, and the solution neutralized, one part with potassa, in order to test for boracic, tartaric, and citric acids, the other part with ammonia, to test for the other acids. Carbonic acid may be recognized by the precipitate caused by transmitting the gas through baryta water. Hydrosulphuric acid is recognized by its odour and by its reaction with a salt of lead, and sulphurous acid is also recognized by.|its odour, and by the white precipitate occa- sioned by passing the gas through a solution of sulphuretted hydrogen. The solution filtered from the precipitate occasioned by car- bonate of soda, (1) is divided into four parts. One part (A) is rendered slightly acid by hydrochloric acid; another part (B) is acidified by nitric acid ; a third part (C) by acetic acid ; and the fourth part (D), after having been acidified by nitric acid added in slight excess, is again made slightly alkaline by the addition of a few drops of caustic potassa. I. Examination of the Solution acidified by Hydrochloric Acid (A). Chloride of barium produces a white precipitate insoluble by boiling with great excess of acid. SULPIIUEIC ACID is pre- sent. NOTE. It being possible that the carbonate of soda used in preparing the solution may contain sulphate of soda, a small portion of the hydrochloric solution of the substance prepared for the examination for bases, may advantageously be employed for the detection of sulphuric acid. Sesquichloride of iron produces a deep-blue precipitate, and protosulphate of iron a pale-blue precipitate. HYDEOFEEEO- OYANIC ACID is indicated. Protosulphate of iron produces a deep-blue precipitate ; ses- quichloride of iron no precipitate, but a brown coloration. HT- DEOFERRIDCYANIC ACID is indicated. NOTE. If a dark-blue precipitate has been produced proto- sulphate of iron, and a dark-blue precipitate also by sesqui- chloride of iron, it is probable that hydroferrocyanic and hydro- ferridcyanic acids are both present. Proof of the presence of the latter acid is obtained by the solubility of the red-brown precipitate, which it forms with nitrate of silver ir^ ammonia ; SYSTEMATIC QUALITATIVE ANALYSIS. 209 the corresponding salt of hydroferrocyanic acid being insoluble in ammonia. Sesquichloride of iron produces a blood-red colour, which is destroyed by solution of chloride of mercury. STJLPHOCYANIC ACID is indicated. NOTE. It must be borne in mind that a similar coloration is produced with acetic acid; in this case however the colour is destroyed by hydrochloric acid. It must be remembered, more- over, that when solution of sesquichloride of iron alone is heated, its colour becomes considerably heightened. II. Examination of the Solution acidified by Nitric Acid (B). Nitrate of silver produces a white precipitate, insoluble even by boiling in strong nitric acid, but very soluble in ammonia. HYDROCHLORIC ACID is present. !N OTE. Hydrochloric acid is, however, best sought for in a nitric acid solution of a portion of the original substance, as it is almost impossible to obtain carbonate of soda quite free from chloride of sodium. Nitrate of silver produces a yellowish-white precipitate, so- luble in ammonia, but not so readily so as chloride of silver. HYDROBROMIC ACID is indicated. Confirmation. Add chlorine water to a portion of the original solution, and then agitate with ether, which will take up the bromine, and acquire a more or less deep yellow tinge. Nitrate of silver produces a yellowish precipitate, soluble with great difficulty in ammonia. HYDRIODIC ACID is indicated. Confirm by the blue colour imparted to starch paste. Nitrate of silver produces a white curdy precipitate, soluble in ammonia and in alkaline cyanides. Confirm by the peach-coloured flame, produced when the gas evolved by heating the dried precipitate in a tube is ignited. HYDROCYANIC ACID is indicated. IN OTE. As chloride of silver and cyanide of silver are both readily soluble in ammonia, a further test is required to dis- tinguish hydrochloric acid in the presence of hydrocyanic acid. The precipitate produced by nitrate of silver leaves after ignition, in the case of pure cyanide of silver, only metallic silver, which may be dissolved by nitric acid; any residue remaining is chloride of silver, which is insoluble in nitric acid. PART i. P 210 QUALITATIVE ANALYSIS. III. Examination of the Solution acidified by Acetic Acid (C). Sesquichloride of iron produces a yellowish-white precipitate. PHOSPHORIC ACID is present. Confirm by molybdate of ammonia. Sulphate of lime produces a white precipitate, which, after ignition, is soluble with effervescence in hydrochloric acid. OXALIC ACID is present. Chloride of calcium produces a ivhite gelatinous precipitate. HYDROFLUORIC ACID is probably present. Confirm by action on glass. IV. Examination of the Solution rendered nearly neutral by Caustic Fotassa (D). Chloride of calcium produces a precipitate, soluble in chloride of ammonium, and in caustic potassa in the cold. TARTARIC ACID is present. Chloride of calcium produces no precipitate till ammonia or lime water is added and the whole boiled. CITRIC ACID is present. NOTE. The recognition of tartaric and citric acids, when both exist together in a solution, requires considerable care. If a precipitate has been produced by chloride of calcium, it must be filtered off ; and the filtrate tested for citric acid by adding ammonia and boiling. When tartrate of silver, formed by precipitating a solution containing tartaric acid with nitrate of silver, is gently heated with ammonia in a test-tube, a shin- ing mirror of metallic silver is deposited ; a similar mirror is formed with citric acid, but only after long boiling. Sesquichloride of iron produces a yellowish-brown precipitate. BENZOIC ACID is probably present. NOTE. Indications of this acid will have been obtained in the preliminary examination. Sesquichloride of iron produces a reddish-brown precipitate. SUCCINIC ACID is probably present. NOTE. If the solution contain at the same time benzoic and succinic acids, the latter may be recognized thus : Digest the precipitate produced by sesquichloride of iron with am- monia ; filter off any benzoate of iron which may be formed, add to the filtrate chloride of barium, alcohol, and ammonia, by which succinic acid is precipitated, succinate of the alkaline earths being insoluble in alcohol. SYSTEMATIC QUALITATIVE ANALYSIS. 211 Sesqidchloride of iron produces a bluish-black precipitate. GALLIC or TANNIC ACIDS are indicated. NOTE. Tannic acid produces, with solution of gelatine, an insoluble curdy precipitate ; gallic acid is not precipitated by gelatine. BOBACIC ACID is recognized by its action on turmeric paper, which it reddens, like an alkali. If a solution, containing a borate, be treated with slight excess of hydrochloric acid, and a piece of turmeric paper be immersed in the liquid and then dried at 212, it will be found to have acquired a brownish-red colour. END OP PART I. JOHW EDWARD TAYLOR, PRINTER, LITTLE QUEEN BTKBET LINCOLN'S INN FIELDS, MANUAL OF CHEMICAL ANALYSIS. PART II. QUANTITATIVE. CHAPTER VII. ON QUANTITATIVE OPEEATIONS. 1. WEIGHING. 145. The Balance. Of the various implements required by the analytical chemist in the prosecution of his labours, the ba- lance is perhaps the most important ; for of what value would the most elaborate and exact experiment be, without a means of ascertaining the quantitative result ? The balance therefore demands our first attention. The process of weighing may be performed in a variety of ways, and with various kinds of machines ; but the instrument invariably employed by the chemist for determining the weights of the substances on which he is engaged, is what is generally known as the common balance, which is a lever of the first kind, with equal arms. A detailed investigation of the circum- stances which regulate the sensibility of a balance, and a ma- thematical development of the principles on which it is con- structed, would be out of place here : it may not, however, be amiss to describe briefly some of the most important points connected with its construction and properties. The philosophical balance may be described as a uniform inflexible lever or beam, Fig. 41, made as light as is consistent PART II. Q * 214 QUANTITATIVE ANALYSIS. with a proper degree of strength, and having three axes, one, the fulcrum, or centre of motion, c, on which it turns, and the other two, A and B, situated at equal distances from the fulcrum near to the extremities of the beam, and from which the scales Kg. 41. or pans depend : the beam thus supports the scales, and is itself supported by means of fine edges of hard steel working on steel, agate, or garnet, in order that the motion may be free and the distances of the points precisely defined. The scales hanging from fixed points in the beam, act on them always in the direc- tion of gravity ; and the effect is the same as if the whole weight were concentrated in those points. The requisites for a good balance are : (1) That the centre of gravity, G, shall be immediately below the fulcrum or centre of motion, c. (2) That the fulcrum or centre of motion shall be in the same right line with the points of suspension of the scales ; so that a straight line drawn from one extremity of the beam to the other, shall be exactly perpendicular to the straight line, c D, joining the centre of gravity with the centre of motion ; and (3) That the arms shall be of equal length. i"or if, in the first place, the centre of gravity and the centre of motion be coincident, the beam will rest in any indifferent position, the scales being equally loaded ; but if the centre of gravity be above the centre of motion, then the slightest im- pulse will cause the beam to upset: if, however, the centre of gravity be immediately under the fulcrum, then the line joining it with the latter will always settle itself so as to be in a vertical direction ; this line has the properties of a pendulum, the shorter it is, the greater the angle formed by its vibration from a given impulse, consequently the nearer the centre of gravity to the centre of motion, the greater the effect of a given weight added to the scale ; or, in other words, the more delicate the balance. The nearer also the centre of gravity of a balance is to its ON QUANTITATIVE OPERATIONS. 215 fulcrum, the slower will be the oscillations of the beam. The number of oscillations, therefore, made by the beam in a given time, affords the most accurate method of judging of the sensi- bility of the instrument, which will be the greater as the oscilla- tions are fewer. In the second place, if the points of suspension of the pans be situated below the centre of motion of the beam, the effect of each addition of weight to the scales will be to lower more and more the centre of gravity ; and this we have just seen diminishes the sensibility of the balance. If, on the other- hand, the points of suspension of the pans are above the ful- crum, the effect of the addition of weights will be to raise the centre of gravity, thus increasing the sensibility of the instru- ment (c&teris paribus), until at length the centre of gravity becomes itself the centre of motion, when the beam will rest indifferently in any position ; and finally, by a further addition of weights, the centre of gravity becomes raised above the ful- crum, and the balance upsets from the smallest disturbance. In the third place, with respect to the equality of the arms, this condition is of course essential to a good balance ; it is nevertheless possible to weigh as accurately with a balance with unequal arms as with one of the same workmanship with equal arms ; for this purpose the substance to be weighed is put into either scale and counterpoised by sand, shot, or any other ma- terial, with the greatest accuracy ; the substance is then re- moved and the balance once more brought into a state of equi- librium by substituting weights, which, it is perfectly obvious, must, under these circumstances, absolutely represent the weight of the substance under examination. It is not unfrequent to meet with commercial balances constructed with unequal arms for fraudulent purposes, the substance to be sold being always put into the scale depending from the longer arm ; the fraud is detected by simply transposing the substance and the weights after equilibrium has been established between them. Sensibility and stability are two properties essential .to a good balance. The sensibility of a balance is estimated by observing the angle through which a very small weight inclines the beam ; thus supposing we wish to compare the sensibility of two balances, and that a grain weight put into a scale of each in- clines the beam of the first 4, and that of the second only 2, then the first is twice as sensible as the second. Now, as the 216 QUANTITATIVE ANALYSIS. force which acts in turning the beam is proportional to the weight multiplied into the length of the lever at the extremity of which it acts, it is evident that, for a given weight, the sensi- bility of the balance (cceteris paribus) is proportional to the length of the beam. The stability of a balance is the force with which the beam endeavours to recover its equilibrium, and oscillate about its position of rest after it has been disturbed. This force is made up of two parts, the first of which is proportional te the weight of the learn multiplied into the length of the lever on which it acts (c a, Fig. 41) ; and the second is proportional to the load multiplied into its length of lever ; the whole restoring force then, and that which the predominating weight has to overcome in turning the scale, is proportional to the weight of the beam multiplied into the length of the lever on which it acts, added to the load multiplied into its length of lever. It is evident, then, that sensibility and stability are two properties in some degree opposed to each other, and that whatever tends to in- crease the one diminishes the other. The best construction is to make the lever through which the load acts equal to nothing, and this is done by placing the three points of action, ABC, Pig. 41, in the same straight line as has already been explained, and by keeping the centre of gravity a little below that line : the sensibility of the balance is thus rendered independent of the load, and this important property is still further increased by making the beam as light as possible. The conditions of a perfect balance, as determined by theory, are the guides to the artist in the construction of a good instru- ment ; and although an absolutely perfect balance is practically unattainable, nevertheless, by attending carefully to the prin- ciples above explained, the sensibility may be carried to a sur- prising extent. There is, in the possession of the Royal Society, a balance made by Earnsden, which is capable of weighing ten pounds, and is said to turn with the ten-millionth part of that load, or the thousandth part of a grain. The balances con- structed by the late Mr. Robinson have long enjoyed a just celebrity. They are thus described by Captain Kater (' Lard- ner's Cyclopaedia' Treatise on Mechanics), who himself con- structed an instrument for verifying the national standard bushel, which is probably the most sensible that has yet been made (Phil. Trans. 1826). " The beam of Robinson's balance is only 10 inches long. It is a frame of bell-metal, in the ON QUANTITATIVE OPERATIONS. 2 1/ form of a rhombus. The fulcrum is an equilateral triangular prism of steel, 1 inch in length ; but the edge on which the beam vibrates is formed to an angle of 120, in order to prevent any injury from the weight with which it may be loaded. The chief peculiarity in this balance consists in the knife-edge which forms the fulcrum bearing upon an agate plane throughout its whole length. The supports for the scales are knife-edges, each T 6 Q-ths of an inch long. These are furnished each with two pressing-screws, by means of which they may be made parallel to the central knife-edge. Each edge of the beam is sprung obliquely upwards and towards the middle, so as to form a spring, through which a pushing screw passes, and serves to vary the distance of the point of support from the fulcrum, and at the same time, by its oblique action, to raise or depress it, so as to furnish a means of bringing the points of support and the fulcrum into a right line. A piece of wire, 4 inches long, on which a screw is cut, proceeds from the middle of the beam downwards. This is pointed to serve as an index, and a small brass ball moves on the screw, by changing the situation of which the place of the centre of gravity may be varied at plea- sure. The fulcrum, as before remarked, rests upon an agate plane throughout its whole length, and the scale-pans are at- tached to planes of agate which rest upon the knife-edges, forming the points of support. This method of supporting the scale-pans is believed to be due to Mr. Cavendish. Upon the lower half of the pillar, to which the agate plane is fixed, a tube slides up and down by means of a lever which passes to the outside of the case. From the top of this tube arms pro- ceed obliquely towards the ends of the balance, serving to sup- port a horizontal piece, carrying at each extremity two sets of T's, one a little above the other. The upper Y's are destined to receive agate planes, to which the scale-pans are attached, and thus to relieve the knife-edges from their pressure; the lower to receive the knife-edges which form the points of sup- port : consequently these latter T's, when in action, sustain the whole beam. " When the lever is freed from a notch in which it is lodged, a spring is allowed to act upon the tube we have mentioned, and to elevate it. The upper Y's first meet the agate planes carrying the scale-pans and free them from the knife-edge. The lower Y's then come into action and raise the whole beam, ele- vating the central knife-edge above the agate plane. This is 218 QUANTITATIVE ANALYSIS. the usual state of the balance when not in use ; when it is brought into action the reverse of what we have described takes place. On pressing down the lever, the central knife-edge first meets the agate plane, and afterwards the two agate planes carrying the scale-pans are deposited upon their supporting knife-edges." A balance of this kind was employed by Captain Kater in adjusting the national standard pound. "With a pound troy in each scale, the addition of y-g-^th of a grain caused the index to vary one division, equal to y^th of an inch ; and Mr. Kobin- son adjusted these balances so that, with 1000 grains in each scale, the index varied perceptibly on the addition of 10 1 00 of a grain, or of one-millionth part of the weight to be deter- mined. Fig. 42. Pig. 42 represents a useful chemical balance, constructed by Messrs. Ladd and Oertling, of Moorgate Street, London, in- tended to carry from 700 to 1000 grains, which will, when loaded, indicate the T -oV^ ^ a g p ain. It is represented in the ON QUANTITATIVE OPERATIONS. 219 figure with a Liebig's potash apparatus attached to one of the arms, the object being to show the facility which a long beam gives to the operator, for weighing these awkward pieces of ap- paratus. The beam of this balance is 16 inches long. The balance, as seen in the figure, is at rest, so that a weight placed in one of the pans will not affect the equilibrium of the beam. By turning the knob in the front part of the case between the two drawers from right to left, the arm supporting the beam will descend, and the balance will be free ; the oscillations of the beam will then be observed by the index moving over an ivory scale in front of the stand. The fulcrum, or centre knife- edge of the beam, works upon agate, with which, however, it is only in contact when the balance is free. The pans are sus- pended by fine platinum wire, and are provided with steel hooks, which hang in rings attached to the ends of the beam ; these rings are ground out conically from each side, so as to form a perfect edge. The apparatus fixed on either side, above the beam, serves to move small weights along the arm, which is accurately divided into ten equal parts. This move- ment is effected by means of small knobs outside the case. The weight used for this purpose is made of fine gold 1 wire, with a loop at the top. It weighs -^th of a grain, conse- quently, when placed at the 1st, 2nd, 3rd, etc., division of the beam, it represents one, two, or three hundredth s placed in the pan. By subdividing each division the weight may be ascer- tained with great accuracy, and all weights less than T Vth of a grain may be dispensed with. The knob seen above the centre of the beam can be raised or lowered, and serves to adjust the centre of gravity, which ought to be a little below the fulcrum, or centre of motion. The beam has also adjustments for bring- ing the ends and the centre knife-edge into the same straight line, and also for equalizing the lengths of the arms ; these contrivances could not, however, be well represented in the figure. Fig. 43 exhibits a balance of a larger description. Its beam is 18 inches long, and it is intended to carry from 2 to 3 pounds in each pan ; and when loaded with such a weight it indicates distinctly -g^th of a grain. In order to obtain such a degree of delicacy with so large a weight, it is necessary that the best construction should be adopted. The beam is furnished with a straight knife-edge at each end, upon which the pans are suspended by agate planes ; by this construction not only 220 QUANTITATIVE ANALYSIS. a greater delicacy and durability, but also a greater constancy is obtained than could be arrived at with any other kind of ba- lance. The instrument is represented in the figure as set free Fig. 43. for oscillation, and on turning the knob in front from left to right, the straight arms close under the beam are raised, and receive the end pieces to which the pans are suspended by the small cylinders indicated in the figure ; therefore when the ba- lance is at rest, the centre knife-edge of the beam is taken off the agate plane of the stand, and the agate planes with the pans are taken off the knife-edges of the beam. The oscillations of this balance are also observed by an index moving over an ivory scale, in front of the stand. The pans in this balance are made of brass, and are 6 inches in diameter ; they are supported by stout brass wire, and the distance from the centre of the pans to the small brass plate to which the supporting wires are attached is 10J inches ; evaporating dishes and flasks of considerable size may therefore be weighed, and mineral waters and filtrates contained in large beakers may be estimated by weight instead of by measure, a ON .QUANTITATIVE OPEKATIONS. 221 method always to be preferred. Both balances are enclosed in glass cases, having convenient doors at the front and sides, and the smaller one is provided with a contrivance for steadying the pans during the operation of weighing. In performing the process of weighing, the student will do well to observe the following rules : 1. Always to use the same pan for the same purpose, for although in a good balance it is obviously a matter of indifference which pan receives the weights, and which the substance, nevertheless, an attention to the prac- tice of always placing the substance to be weighed in the same scale, serves in a great measure to obviate any error which may arise from inequality in the length of the arms ; as, if the weights are always put into the same pan, the substance weighed in the other will be in the same proportion as the weights, though not exactly equal to them ; and, in chemical experiments, propor- tional quantities are in general of far greater importance than real weights. 2. Never to place the substance to be weighed directly on the scale, but upon some interposed substance, as a watch glass, or, occasionally, on a piece of hot-pressed wove paper, counterpoised of course in the other scale by an equal- sized watch glass, or piece of paper, previous to putting in the weights. A substance should never be weighed while hot, in consequence of the effect produced by the ascending current of hot air, and the rushing in of a stream of cold air, which tend to give an upward motion to the pan, making it seem lighter than it really is. Substances which attract moisture from the air should be weighed in a covered crucible, or in a watch glass, the edge of which is ground so that it may be covered air-tight with another equal-sized watch glass. Another method of weighing certain substances which change in air is recommended by Faraday (' Chemical Manipulation' p. 41), and may sometimes be very conveniently adopted. It consists in transferring them, not to the pan of the balance, but to a por- tion of water, alcohol, etc., as the case may be, of known weight, and ascertaining the increase of weight so occasioned; the ana- lysis or experiment is then made with the solution instead of the solid body. A third important rule to be observed in weigh- ing, is never to make any adjustment of weights or substance in the scales while the beam of the balance is free, but always to take it off its support previous to interfering in any way with the contents of the pans. The neglect of this precaution may soon entail serious injury on a delicate balance. 222 QUANTITATIVE ANALYSIS. The following method of examining a newly purchased balance, recommended by Captain Kater, may be usefully in- serted here, by way of conclusion to the few observations we have thought it advisable to make on the subject of this im- portant instrument. 1. To ascertain whether the points of suspension and the fulcrum are in a right line, make the vibrations of the balance very slow by moving the weight which influences the centre of gravity, and bring the beam into a horizontal position by means of small bits of paper thrown into the scales. Then load the scales with nearly the greatest weight the beam is fitted to carry. If the vibra- tions are performed in the same time as before, no further adjustment is necessary ; but if the beam vibrates quicker, or if it oversets, cause it to vibrate the same time as at first, by moving the adjusting weight, and note the distance through which the weight has passed ; move the weight then in the con- trary direction through double this distance, and then produce the former slow motion by means of the screw acting vertically on the point of support. Eepeat this operation till the adjust- ment is perfect. 2. To make the arms of the beam of an equal length. Put weights in the scales as before ; bring the beam as nearly as possible to a horizontal position, and note the division at which the index stands ; unhook the scales, and transfer them with their weights to the other ends of the beam, when, if the index points to the same division, the arms are of equal length, but if not, bring the index to the division which had been noted, by placing small weights in one or the other scale. Take away half these weights, and bring the index again to the observed division by the adjusting screw, which acts hori- zontally on the point of support. If the scale-pans are known to be of the same weight, it will not be necessary to change the scales, but merely to transfer the weights from one scale-pan to the other. 146. Weights. The unit of weight adopted in this country is the grain, and in France and Germany the gramme. It must be regarded as an unfortunate circumstance, that we have in England two different sets of weights expressed by the same names. The pound and the ounce troy, and avoirdupois, have different values. The avoirdupois pound is greater than the troy pound, the former weighing 7000 grains, and the latter 5760 ; but as there are in the avoirdupois pound 16 ounces, while in the troy pound there are only 12, it follows, that the troy ON QUANTITATIVE OPERATIONS. 223 ounce is greater than the avoirdupois ounce, the latter weighing 437| grains, and the former 480. In order to avoid fractions, and as there are no other grains than troy grains, troy weight is generally used for philosophical purposes. The standard to which troy weight is referred, is the weight of one cubic inch of distilled water, which at 62 F. and 30 inches barometer weighs 252'458 grains. The troy weight table is gr. dwt. oz. Ib. 24 = I 480 = 20 = 1 5760 = 240 = 12 = 1 These denominations are never however employed in che- mistry, the weight of substance being (when the English system is used) always reckoned in grains, and in the decimal divisions of the grain, viz. O'l, 0'2, 0'3, etc., O'Ol, 0'02, 0'03, etc. The standard to which the French gramme is referred, is the weight of y^Q-th part of a cubic metre* of distilled water, at the temperature of melting ice. A gramme is equal to 15'434 grains troy, whence the following comparative table of French with troy weight. Gramme. Troy grains. Milligramme = "001 = -01543 Centigramme = *01 '15434 Decigramme -1 1'5434 Gramme = 1 = 15-434 Decagramme = 10 = 154'34 Hectogramme = 100 = 1543-4 Kilogramme = 1000 = 15434 Myriagramme = 10,000 = 154340 The kilogramme is equal to 2 Ib. 3 oz. 4'428 drachms avoirdu- pois, and to 2'679 Ib. troy weight. The student should assure himself once for all, on purchasing a set of weights, that they are correct among themselves, and if he has an opportunity of comparing them with standard weights, he should do so : for the purpose of adjustment, the small spherical knobs screwed into the centre of the cylindrical weights are provided with small cavities under the screws, to * A metre is the ten-millionth part of the length of the meridian arc between the pole and the equator = 39'37lO English inches, though the ac- curacy of this measurement has been disputed. 224 QUANTITATIVE ANALYSIS. receive portions of fine wire ; a very simple method of making small weights expressing the decimal parts of a grain, is to determine with great care the weight of three or four feet of fine silver, or platinum wire, and then to cut off such portions as are equal to the weights required. The weight box, as well as the balance, must be carefully kept out of the contact of acid or other vapours ; the brass weights may be gilded (pre- vious to adjusting them) to keep them bright and clean ; it is not, however, to be concluded that they necessarily become in- correct by getting tarnished, for Fresenius states, that he has examined many weights of this description, and has found them as exactly corresponding with each other in their relative pro- portions as when they were first used. As a general rule, how- ever, the weights should not be touched by the hand, to prevent the necessity for which the weight boxes are provided with con- venient brass forceps for the small weights, and with forks or tongs for the large ones. It is desirable to keep a shallow dish, containing quicklime, inside the glass case of the balances, in order to preserve the atmosphere in a state of dryness, and thus to prevent the oxidation of the fine steel edges and other me- tallic parts of the apparatus. 2. SPECIFIC GRAVITY. 147. It is very frequently an object of importance to com- pare the density of a substance, solid, liquid, or gaseous, with that of another substance assumed as a standard : this opera- tion is called the determination of its specific gravity. The density of solids and liquids is compared with that of pure dis- tilled water ; gaseous fluids are usually compared with atmo- spheric air. When great accuracy is required, the tempera- ture of the water is of importance, and in such cases it is con- venient to take the temperature of the maximum density of that fluid (viz. about 39 0> 4 F.), at which temperature it can be more easily maintained without variations than at 62, at which it is generally taken in this country, or at the tempera- ture of melting ice, 32, at which it is taken in France. Ge- nerally speaking, in the determination of specific gravities, it is sufficient to note the temperature, and to apply a correction depending on the known density of water at different degrees of the thermometer. (I.) Determination of the Specific Gravity of a Solid, a. Hea- vier than Water. This depends on the principle discovered by ON QUANTITATIVE OPEEATIONS. 225 Archimedes, that a body, when immersed in a fluid, loses just as much of its weight as is equal to the weight of an equal vo- lume of that fluid ; thus, supposing a piece of gold suspended from one of the pans of a balance by a fine thread or hair, to be accurately weighed in air, and then immersed in a jar of distilled water, it is evident that a quantity of water will be forced over the sides of the jar exactly equal in volume to the metal which displaces it : it is equally clear that the force with which the metal is pressed down in the water, is equal to the difference between its own weight and that of an equal bulk of \vater. The method proposed by Archimedes for solving the famous problem of the crown of Hero, king of Syracuse, was to immerse the solids to be compared, of known weights, in water contained in a cylindrical vessel of known area, and to note the relation between the heights at which the liquid stands, which gives the relation between their densities. The method now adopted, and w r hich is susceptible of far greater accuracy, is to weigh the substance first in air, and then in water, and to divide the weight in air by the loss it sustains when weighed in water, the quotient is the specific gravity of the substance. Let S = the specific gravity of the substance ; "W = its weight in air; W = its weight in water; then W = W W The method of performing the operation is shown in Pig. 43, p. 220. Thus, suppose a piece of gold to weigh in air 77 grains, and in water 73 grains, the loss in this case is 4 grains, and we have the proportion 4 : 77 : : 1 : 19'25 ; 19*25 is, therefore, the specific gravity or density of the gold. b. Lighter than Water. If the solid, the density of which is to be determined, be lighter than water, there are two ways of proceeding. By the first method the dimensions of that part of the solid immersed, while it is floating on the surface of water, is com- pared with its whole magnitude ; the two results bear the same proportion to each other as the specific gravity of the solid bears to that of water. The experiment is performed thus : A glass vessel, having perpendicular sides, and as narrow as the magni- tude of the solid will admit, is filled to a certain mark with dis- tilled water ; the substance is then set to float in it, and the point 226 QUANTITATIVE ANALYSIS. to which the surface of the water rises in the vessel accurately marked : the body is then totally submerged, and the point to which the surface of the water rises again observed. The ele- vations of the surface produced by the partial and total sub- mersion, indicate the portions of the solid in each case im- mersed, and are therefore in the ratio of the specific gravity of the solid to that of the liquid. This method is not, however, very often adopted. The following is more convenient : The body, the density of which is to be ascertained, is attached to another which is heavier than water, and of such a size that the united weights of the two will be greater than the weight of the water which they displace, and the whole consequently sinks when immersed. The weight of the united substances is then determined, first in air, and then in water ; the weight lost by immersion is equal to the weight of a quantity of water corresponding in bulk to the united bulks of the solids. The lighter solid is then detached, and the weight lost by the hea- vier by immersion in water ascertained ; this is the weight of a quantity of water equal in bulk to the heavier solid. This loss of weight being subtracted from the loss sustained by the com- bined masses, the remainder is the weight of a quantity of water equal in bulk to the lighter solid ; the proportion of the weight of the lighter solid to this will determine its specific gravity. This process may be still further simplified thus: Weigh the substance, the specific gravity of which it is desired to as- certain, first in air, then attach it loosely to a piece of heavy metal, the weight in water of which has been determined, and weigh the combined substances in water; the aggregate weight will be found less than that of the heavier body. Now subtract the weight of the lighter body from that of the heavier, and add the remainder to the weight of the former in air ; we thus ob- tain the weight of a quantity of water equal in bulk to the lighter body, and by dividing the weight of the lighter body in air by this last-mentioned sum, the quotient is the specific gravity required. Example. A piece of indigo covered with a thin coating of varnish (to prevent its absorbing water) weighed in air 105'1 grains: a piece of sheet lead which was used for ballast weighed in water 124'9 grains. The indigo and the lead were now tied together, not closely, but in such a manner that the water could have free access to any part of each substance, OK QUANTITATIVE OPEEATIONS. 227 and weighed in water, the aggregate weight was 108'5 grains; that is, 16'4 grains lighter than the weight of the lead alone in water. Now 16*4 + 105*1 = 121 - 5 = the weight of a quantity of water of equal bulk to the indigo, and -foT^ = 0*865 = the specific gravity of the indigo. c. Of a Powder. When the solid, the specific gravity of which is to be determined, is in the form of a powder, or in minute pieces, it may be placed in a cup which is counter- poised first in air, and then in water, in the usual manner ; or the same method may be pursued which is adopted in taking the specific gravity of a soil. A small bottle, containing, when filled and its stopper adjusted, a certain known quantity of water, is about half filled with distilled or rain water, and a weighed quantity of the dry powder introduced. It is then well shaken, and' after the solid matter has settled, it is com- pletely filled with water, the stopper put into its place, and then weighed. The weight obtained will be obviously less than that of the united weights of the bottle full of water and the substance added, and the difference is the weight of a quantity of water equal in bulk to the solid substance added, and by dividing the weight of the powder by this difference of weight we obtain the specific gravity required. Example. A bottle filled with distilled water and closed with its glass stopper weighed 750 grains. About half the water was poured out, and 500 grains of a certain powder introduced. The water and the powder were well mixed by agitation, and then allowed to settle. After awhile the bottle was completely filled up with water, closed with its stopper, and weighed. United weights of the bottle, water, and powder, (supposing the latter could have been introduced without displacing any of the water) .... 1250 grs. Actual weight of the bottle, water, and powder . . 1050 Difference, expressing the weight of a quantity of water equal in bulk to the powder 200 Therefore 500 2QQ- = 2*5 = specific gravity of the powder. d. Soluble in Water. When the solid, the specific gravity of which has to be determined, is soluble in water, it must be 228 QUANTITATIVE ANALYSIS. weighed in some liquid which exerts no action upon it; the density of the body is the product obtained by multiplying this number by the specific gravity of the liquid used. Thus, sup- pose the substance to have been weighed in oil, and its density with reference to that fluid to be 4'3, that of the oil being 09, then the specific gravity of the body will be 4'3 xO'9^3'87. (II.) Determination of the Specific Gravity of a Liquid. The specific gravities of liquids may be determined in two ways : 1st, by weighing them in vessels of known magnitude ; and 2nd, by weighing a certain solid in the liquid to be exa- mined, and comparing the loss of weight with that which it sustains by immersion in pure water. Bottles are sold by the philosophical instrument makers which hold, when full, and closed with their stoppers, a certain quantity, 1000, 500, or 200 grains of distilled water. The stoppers of these bottles are usually perforated, with the view of affording a free pas- sage for the excess of fluid when the stopper is inserted in its place ; there is, however, no occasion for this, and indeed it is very objectionable with corrosive fluids, such as sulphuric acid, and with volatile fluids, such as ether. For such substances M. Hegnault uses a flask with a narrow neck (Fig. 44). The Fig. 44. fluid is filled to the mark on the neck, and the stopper is inserted in its place ; evaporation is thus rendered impossible, and should the liquid expand during the operation of weigh- ing, no liquid can escape from the bottle in consequence of the enlargement of its neck. The bottles are accompanied with brass weights which exactly counterpoise them when empty and dry. The method of determining the specific gravity of a fluid by means of one of these bottles is very simple : we have ON QUANTITATIVE OPERATIONS. 229 only to fill the bottle with the liquid in question and weigh it ; its weight gives at once the specific gravity without any calcu- lation. Thus, supposing the bottle to hold 1000 grains of dis- tilled water at 62, and at the same temperature, 1845 grains of oil of vitriol, and 918 grains of spirit ; these numbers repre- sent the densities of the liquids, water being 1000 ; or, if water be taken as unity, then the specific gravity of oil of vitriol is 1'845 and that of spirit 0*918. In determining the specific gra- vity of liquids in this manner, due regard must be had to tem- perature, otherwise the results obtained will not admit of exact comparison. The bottle should be verified before it is used, and if it should be found incorrect, that is, not to hold exactly 1000, 500, etc., grains at 62, its capacity at that temperature must be rigorously determined once for all, and the weight of water found must be used as the divisor of the weight of any other fluid. Thus, suppose the bottle be found to hold exactly 995 grains of distilled water and 1836 grains of oil of vitriol, then O! =1-845= the specific gravity of the acid. "95 The determination of the specific gravity of a liquid by weighing in it a certain solid, depends on the proposition, that if a solid specifically heavier than water, and also specifically heavier than the liquid whose specific gravity is to be deter- mined, be successively immersed in water, and in that liquid, the losses of weight will be proportional to the specific gravities of water and the liquid : and if the number expressing the loss of weight in the liquid, be divided by the number expressing the loss of weight in the water, the quotient will be the specific gravity of the liquid. Example, Suppose a ball of glass, when weighed in water, to lose 200 grains, and when weighed in oil of vitriol, 396 grains, then '- =l'845=the specific gravity of the oil of ^00 vitriol ; but suppose the same ball, when weighed in spirit, to 183*8 lose 183-8 grains, then =0'918= specific gravity of the spirit. A series of hollow bulbs of different specific gravities, called " specific gravity bulbs," are also sold by philosophical instru- ment makers. Their use is almost entirely confined to practical chemical operations on the large scale, though they are some- 230 QUANTITATIVE ANALYSIS. times of great convenience in the laboratory for determining the degree of concentration of liquids. These bulbs are gene- rally numbered. They indicate the density of the fluid by sinking or floating in it, as the case may be. In cases where moderate accuracy only is required, the Jiy- drometer is extensively exployed for the determination of the specific gravities of liquids. Its use is principally confined to commercial purposes, where expedition is of greater conse- quence than great exactness. The indications of these instru- ments depend upon the fact, that a body when it floats in a liquid displaces a quantity of the liquid equal to its own weight ; and their accuracy depends upon giving them such a shape, that the part of them which meets the surface of the liquid in which they float ? is a narrow stem of which even a considerable length displaces but a very small weight of the liquid. Thus any error in observing the de- gree of immersion, entails upon the result an effect which is inconsiderable. The hydrometer shown in Pig. 45 is an ar- rangement of considerable delicacy, and is pe- culiarly useful for measuring the specific gra- vity of mineral waters ; many forms have been given to the instrument.* It consists of a ball of glass, three inches in diameter, with another joining it, arid opening into it, of one inch in diameter, B and c ; and a brass neck, d, into which is screwed a wire a 0, about 10 inches long and -^th of an inch in diameter, divided into inches and tenths of an inch. The whole weight of this instrument is 4000 grains when loaded with shot in the lower ball. It is found that when plunged into water in the jar, a grain laid upon the top a makes it sink one inch ; therefore the tenth of a grain sinks it the tenth of an inch. Now it will stand i n one ki nc [ O f wa ter a tenth of an inch lower than in another, which shows that the bulk of one kind of water equal to the bulk of the instrument, weighs one-tenth of * For further details respecting hydrometers, and specific gravity gene- rally, the reader may consult with advantage the Treatise on Hydrostatics, in the ' Library of Useful Knowledge,' from which the description of the hydrometer in the text is taken, and in ' Lardner's Cabinet Cyclopaedia.' ON QUANTITATIVE OPERATIONS. 231 a grain less than an equal bulk of the other kind of water ; so that a difference in specific gravity of 1 part in 40,000 is thus detected. This weight of 4000 grains is convenient for comparing water ; but the quantity of shot in the lower ball may be varied so as to make it lighter or heavier, and so adapt it to measure the specific gravities of lighter or heavier liquids. It will always be an accurate and very delicate mea- sure for liquids of nearly the same weight. Indeed its deli- cacy is so great that an impurity too slight to be de- tected by any ordinary test, or by the taste, will be discovered by this instrument. The most useful hydrometer for general purposes is that of Beaume. It contains for fluids heavier than wa- ter, a scale on which represents the depths to which it sinks in distilled water ; this is its highest point, and 15 is that to which it descends when immersed in a solution of 3 parts of lay-salt in 17 of water ; about 50 or 60 degrees are marked between these two extreme points. In the instrument used for fluids lighter than water, is marked at that point to which it sinks in a mixture of 1 part lay-salt and 9 water, and 10 at that where it stands in pure water. There are generally 50 degrees ascending from the lower of these points to the upper. frig. 46 shows the form of hydrometer ( TTrinometer) employed by Neubauer for determining the specific gravity of urine. It is graduated so as to allow the specific gravity to be ascertained within half a degree from I'OOO, the specific gravity of water, up to 1'040, about the highest specific gravity of human urine, and it encloses a small thermometer fixed in the floating portion of the instrument. In reading the indications of the scale of the hydrometer, the eye should be brought to a level with the under surface of the fluid : this level is attained as soon as the hinder border of the surface of the liquid ceases to be visible ; the scale is then read off at that level. If the position of the eye be not right ; if it be on too high or too low a level, the surface of the fluid will take the form of an ellipse, Fi the instrument is then pressed down a few degrees deeper into the liquid, allowed to rise freely in it, and the scale read off" a second time in order to correct mistakes. In T? 9 232 QUANTITATIVE ANALYSIS. this way with a good instrument very accurate results are ob- tained. (III.) Determination of the Specific Gravity of a Gas. This, though simple in principle, is in practice an operation of much delicacy, and requires considerable manipulatory skill on the part of the operator. The gas is weighed in a globe or balloon, the weight of which when empty, and when full of air, is known. The weights of the gas and of the air, are obtained by sub- tracting the weight of the exhausted globe from the weights of the globe filled respectively with air and gas. The quotient ob- tained by dividing the latter by the former, gives the specific gravity of the gas. Begnault employs balloons of the capacity of 10 litres (1 litre = 1'76 English pint), which he counterpoises on the ba- lance (which must be one of great delicacy) by another bal- loon of the same kind of glass, and of as near as possible the same bulk ; the volumes of air displaced by each balloon are accurately determined by filling each with water, and weighing it first in wafer of the same temperature, and then, after care- fully drying, in air, the difference in the two weighings giving of course the weight of the water displaced by the external volume of the balloon ; the difference found in the external vo- lumes of the two balloons is made up by attaching to the smaller a closed tube of glass, having the same external capacity in cubic centimetres as the weight of the water displaced is less than that displaced by the other balloon in grammes, one cubic centimetre of water weighing one gramme. In this way he avoids the numerous corrections which would otherwise be required for variations in the density, temperature, and hygrome- tric state of the atmosphere, during the continuance of an ex- periment, as well as that for the displacement of air, and the adherence of a film of moisture to the glass. The balloons are suspended beneath the scale-pans of the balance in a chamber closed with glass doors to prevent currents of air. The air having been removed from the balloon as completely as pos- sible by the air-pump, the gas the specific gravity of which is to be determined is allowed to enter ; the balloon is a second time exhausted, and again filled with the gas ; a third exhaus- tion as complete as possible is made, and the gas is allowed to enter the balloon slowly whilst it is kept surrounded with melting ice, it is then wiped with a damp cloth to prevent electrical excitation, and after the lapse of two hours weighed. ON QUANTITATIVE OPERATIONS. 233 The weight having been determined with the greatest possible care, the balloon is again exhausted whilst surrounded with ice, but as a perfect exhaustion is unattainable, it is necessary to ascertain the elastic force of the gas remaining, which is done by the use of the barometric manometer. This instrument con- sists of a barometer, and a tube of similar diameter and height, but open at both ends, standing in the same cistern of dry mer- cury. The cistern is partially divided into two unequal com- partments, the barometer standing in the smaller. The open end of the second tube is connected by means of a leaden pipe with the exhausted globe, mercury of course rises in the tube immediately the cock is turned, and the difference of level be- tween the column of mercury in this tube, and that in the barometer tube, is the measure of the elastic force of the gas. The temperature during the experiment is noted. The difference between the two weighings of the balloon re- presents the weight of the gas, which at C. (= 32 F.) fills the balloon under a pressure equal to the barometric pres- sure: Let "W = The weight of the balloon and gas. W = The weight of the exhausted balloon. B = The barometric pressure. B' = The elastic force of the residual gas. Then the formula Gives the weight of the gas at C., and under the normal pressure of 760 millimetres (= 29'922 inches). It may be useful to give here the corrections to be applied to an observed volume of gas for temperature and pressure ac- cording to the English standard. In actual analysis the more convenient French system is almost invariably adopted. The subject will be returned to, in a future chapter, where the me- thods of analysing nitrogenous organic substances are under consideration. a. Corrections for temperature. The experiments of Magnus and Eegnault give as the expansion of air from 32 to 212 -^Hhf or iiths of its volume at 32. The dilatation for each de- gree F. is according to the last observer 0*002036 or ?&? part. It follows therefore that air at 32 expands j^ of its bulk for every degree above, and contracts by the same amount for every degree below that temperature ; 491 volumes of gas at 32 be- 234 QUANTITATIVE ANALYSIS. come 459 volumes at 0; it may be stated therefore that gases expand 4^-gth part of their volume at F. for each degree ; that is, 459 cubic inches at become at the normal temperature, viz. 60, 519 volumes, and at 50, 509 volumes ; hence the follow- ing simple rules for determining the volume which a gas at any observed temperature would occupy at the normal temperature. Suppose 50 cubic inches of gas to have been observed when the thermometer was standing at 50, then Yolumes of air Volumes of air Observed vol. -p- -, . cn o , rr . . f*r\o f . K^O Volume at ol) . at 50. at 60 . of gas at 50 . As 509 : 519 : : 50 : 50'98 = the volume of the gas at 60. Suppose the same volume of gas to have been observed when the thermometer was standing at 70, then, as 459 volumes of air at become 529 volumes at 70, the proportions are Volume of air Volume of air Observed vol. * , Crt0 at 70. at 60. ofgasat70. Volume at 60 . As 529 : 519 : : 50 : 49'05 = the volume of the gas at 60. The reduction of the volume of a gas at any observed tem- perature to the volume it would occupy at 32 is attained by the use of the following formula : Let t = the observed temperature. t' = the required temperature (32 F.). x = the observed volume at t. x = the required volume at t'. Then, , _ (459 + Q x 459 + t b. Correction for pressure. This is very simple, for although there is reason to doubt the absolute accuracy of the la\v known as the " law of Marriotte," viz. that gases are expansible to an indefinite extent in proportion as the pressure upor them is diminished, and to be contractible under increased pres- sure exactly in proportion to the compressing force, yet the correctness at such pressures as occur naturally, and are indi- cated by the barometer, is generally admitted. The following is the rule for calculating the volume which a gas should possess at one pressure from its known volume at another pressure. . ON QUANTITATIVE OPERATIONS. 235 As the pressure to which the gas is to be reduced, is to the observed pressure or height of the barometer, so is the observed volume to the volume required. Thus, suppose the observed volume of a gas to be 120 cubic inches when the barometer is standing at 28*8 inches, we find its real volume at the normal pressure thus : Normal Observed Observed Kequired pressure. pressure. volume. volume. As 30 : 28-8 : : 120 \;0 ; x x = 115'2 cubic inces. Suppose the same quantity of gas to have been measured when the barometer was standing af 30'6 inches, then As 30 : 30-6 : : 120 : x x = 122'4 cubic inches. "When the correction of a gas is to be made both for tempera- ture and pressure, the reduction is first made for temperature, and the corrected volume is afterwards reduced according to the pressure. (IV.) DETERMINATION OF THE SPECIFIC GKAVITT OF A VAPOT7B. a. Gay-Lussac^s method. The liquid to be experimented upon is introduced in quantities of three or four grains, into one or two little glass bulbs. The manner of filling these bulbs is as follows : The bulb having been first accurately weighed, is warmed gently over the flame of a spirit lamp, and then allowed to cool, with its point immersed in the liquid with which it is to be filled ; as the bulb cools a certain portion of the fluid enters, this is then boiled in the bulb till all the air is expelled, and the bulb is filled with the pure vapour of the substance ; the point is then dipped under the surface of the liquid ; as the vapour condenses, the fluid rushes up to supply its place, and the whole becomes full ; the point of the bulb is then sealed by the blowpipe, and when cool it is weighed, the difference of weight gives, of course, the weight of the quantity of liquid which it contains. Two or three bulbs, being thus filled, are introduced under the edge of a narrow graduated jar, filled with, and standing over, a dish of mercury, placed upon a small furnace, or over a lamp. The jar is covered with a glass cylin- der, open at both ends, and secured on the mercury dish. This cylinder is filled, above the top of the jar, with colourless oil. All things being thus arranged, the furnace or lamp is lighted ; 236 QUANTITATIVE ANALYSIS. as the temperature of the mercury and oil rises, the liquor in the little bulbs forms steam, which at length bursts the bulbs and the level of the quicksilver in the jar immediately falls, the vapour occupying the space above it. "When the mercury ceases to descend, it is known that the whole of the liquid has been converted into vapour ; the temperature of the oil, necessarily the same as that of the vapour, is ascertained, and by the graduation on the jar, the volume occupied by the vapour is easily read off. The weight of the vapour is known, it being that of the liquid in the bulbs ; its volume at the temperature of the oil is thus found, from which its volume at the mean temperature may be calculated and compared with that of an equal weight of atmospheric air at the mean temperature and pressure. Fig. 47. b. Dumas' method. This is applicable to all temperatures be- low the melting-point of glass, and to the determination of the vapours of all substances vaporizable under that temperature, and which do not suffer decomposition by vaporization. Its discovery has therefore been of the greatest service to science, ON QUANTITATIVE OPERATIONS. 237 and has'led to results of the most important scientific character It is conducted as follows : A globe of difficultly fusible glass, and capable of holding from 15 to 30 cubic inches, is first thoroughly dried ; this is done by repeatedly exhausting and re- admitting air, dried by passing over long tubes filled with chlo- ride of calcium, the globe being buried in hot sand ; it is then drawn out to a long neck, having a capillary orifice ; this neck is bent, as shown in the figure. It is weighed while full of atmospheric air, the temperature and barometric pressure being noted. About 120 grains of the substance to be examined are poured into a glass, (if a solid, it must be liquefied by heat,) and the globe having been warmed, its point is immersed in the liquid, and 80 or 100 grains allowed to enter ; this part of the operation may be hastened by cooling the globe by dropping ether upon it. The globe is then fitted in a sort of wire cage, by which it is securely fixed in the centre of the bath in which it is to be heated, as seen in the figure. The bath is charged either with oil or with a saturated solution of chloride of cal- cium. The capillary beak of the tube must be long enough to project over the surface of the bath, as shown in the figure. The temperature of the bath is then increased to the desired point ; when the globe becomes sufficiently heated, the liquid boils, and its vapour in passing away carries off the air which had previously filled the globe. The force of the steam in- creases, at first, with the temperature of the bath, by degrees it diminishes, and finally ceases altogether ; the operator then knows that the excess of vapour has passed away, and that the globe is filled with the vapour of the substance, in a state of purity. Should any of the vapour have condensed in the neck of the globe, it must be chased away by a piece of red-hot char- coal. The orifice of the capillary neck is now closed air-tight by means of the blowpipe, the temperature of the bath being at the same moment carefully noted ; the globe is then removed, allowed to cool, cleaned, and weighed. The next point is to ascertain exactly the capacity of the globe, with which view the extremity of the neck is broken off under mercury ; this is done by first making on it a slight scratch with a file, and then plunging it underneath the surface of the fluid metal, and breaking it off by a slight tap against the side of the trough ; the mercury immediately rushes into the globe, filling up the vacuum caused by the condensation of the vapour. If the experiment has been well conducted, the whole of the globe 238 QUANTITATIVE ANALYSIS. becomes filled, but if there was a small quantity of air still re- maining at the time of sealing the neck, a bubble will remain in it. The globe full of quicksilver is then emptied into a gradu- ated jar, by which the quantity of the quicksilver being measured, the capacity of the globe is known, and all the elements for calculating the specific gravity of the vapour are thus ascer- tained. Thus, the volume of the globe being known, the weight of the atmospheric air it contained, is ascertained by calculation, and the globe having been weighed full of air, by deducting the weight of that volume of air. the weight of the empty globe is ascertained, and deducting this weight from that of the globe when filled with the vapour of the substance, or rather when containing the liquid resulting from the condensation of that vapour, we obtain the weight of the vapour of the substance at the temperature of the bath. The weight of the atmospheric air filling the globe at the temperature of the bath is now as- certained by calculation, by the rules already given, and this being known, the determination of the density of the vapour is made by the simple rule of proportion. Example. By way of illustrating this beautiful operation we quote an experiment made by the author of the process (Du- mas), the weights and measures being reduced to the English standard : we have omitted certain corrections applied by the Erench chemists, such as that for the expansion of the glass, and the reduction of the indications of the mercurial thermometer, to those of the air thermometer, according to the experiments of Magnus ; the results of these corrections not having a very significant bearing on the result. The experiment is the determination of the specific gravity of the vapour of camphor. The particulars are as follows : Temperature of the air 56'3 E. Height of the barometer . . . . 29 '2 in. Temperature of the bath at the time of sealing the balloon .... 471'2 E. Increase in the weight of the balloon 10'88 grains Capacity of the balloon as determined by the weight of the volume of mer- cury required to fill it . . . . 18*0 cub. inches. (I.) deduction of the volume of atmospheric air which the balloon contained at the temperature, and under the atmo- spheric pressure, at the time of weighing it, to 32 E. and 29' 9 in. barom. ON QUANTITATIVE OPERATIONS. 239 a. Correction for pressure : As 29-9 : 29-2 :: 18'0 : x x 17-57 /?. Correction for temperature according to the formula (459 + Q x * - 459 + * 459 + 32x17-57 ,__. ,. . , . f . = 16-74 cubic inches = the volume of air at 459 + 5b'3 32 F. and 29'9 in. barom. (II.) Weight of this Volume of Air : According to the experiments of Regnault, 100 cubic inches of atmospheric air at 32 E. and 29*9 in. barometer, weigh 32 '5684 grains, therefore Cubic Cubic Cubic inches. inches. inches. 100 : 16-74 : : 32'58684 : x x = 5*455 = the weight of the dry air with which the balloon was filled when it was weighed. Now the increase in the weight of the balloon with its charge of camphor vapour was 10*88 grains, if therefore we add this number to the calculated weight of the air, we shall obtain the actual weight of the vapour, thus 10-88 + 5*455 = 16-335 = the weight of the camphor vapour. (III.) Reduction of the Volume of Camphor Vapour (= 16*335 grains) to 32 F. and 29 -9 in. barometer : a. Correction for pressure As 29-9 : 29-2 : : 18'0 : 17'57 f3. Correction for temperature 459 + 32 x 17-57 = Q . 27 cubic incheg _ the volume of the vapour at 29'9 in. barometer and 32 F., and which consequently weighs 16'335 grains ; therefore 100 cubic inches weigh 176'2 grains. Now, as 100 cubic inches of air weigh at 32, 32-58684 grains, the specific gravity of camphor vapour is found by the propor- tion As 32-58684 : 176-2 : : 1 : x x = 5'40 = the specific gravity of camphor vapour. Dumas' calculation, with the corrections for the expansion of the glass globe and the reduction of the degrees of the mer- curial to those of the air thermometer, gives as the density 5-52. 240 QUANTITATIVE ANALYSIS. 3. MEASURING. 148. In the quantitative examination of liquids, recourse is generally, when practicable, had to weighing, in preference to measuring. For this purpose the balance, Fig. 43, is well adapted. When, however, it is deemed expedient to resort to the latter, the imperial pint and its subdivisions is a convenient standard for large quantities, aud the cubic inch and its sub- divisions for small quantities of liquid. The measure of the imperial pint is 34*65925 cubic inches, and it contains 8750 grains of distilled water at 62 F., and at 30 in. barometer. It is divided into 20 parts, each part having a capacity of 1 "7329625 cubic inches, and containing 1 ounce avoirdupois, or 437*5 grains of distilled water. The cubic inch consists of 252*456 grains of distilled water at the standard temperature and pressure, and of 3425*35 grains of mercury. Lipped jars of the imperial pint, half-pint, and quarter-pint capacity, graduated into ounces and half-ounces, are articles of commerce, and are easily ob- tained. The chemist never, however, takes it for granted that these gradations are correct, but verifies each jar for him- self by weighing into it successive quantities of water or mer- cury, and making a mark or scratch with a file or diamond at the correct division, should it not coincide with that previously made on the jar. Bottles having an exact cubic inch capacity, and holding, therefore, 252*456 grains of distilled water, and 3425*35 grains of mercury when the stopper is in its place, are also to be obtained at the philosophical instrument makers, and when correct, they are exceedingly useful, and save a great deal of time. For the measurement of gases various sized jars are required, Figs. 48, 49, 50, 51 ; they are always in this country graduated into cubic inches, tenths, and hundredths ; the laboratory should be furnished with a good supply of them, and some of them, as Fig. 49, should be sufficiently stout to bear, without risk, their weight when filled with mercury during the manipulations at the mercurio-pneumatic trough. As a general rule, all jars, whether for gases or for liquids, should be carefully verified before they are used for quantitative experiments; tubes in particular that are intended to be employed in gaseous analysis, must be rigorously tested; and it is better that the chemist should graduate them himself than trust to the accuracy of those met with in commerce. For this purpose, having selected ON QUANTITATIVE OPERATIONS. 241 a tube of the desired size and thickness, and of clear faultless glass, it is placed in a perfectly vertical position, with its closed Fig. 48. Fig. 49. Fig. 50. Fig. 51. end on a flat surface. A small piece of glass tube, Fig 52, of known capacity, the'edge of which is ground so that it may be closed air-tight with the piece of ground glass, serves as a measure or gauge ; this is filled with mercury, and all bubbles of air having been expelled by passing a rod of glass down the internal sides of the tube, its contents are transferred to the jar or tube to be graduated ; the sur- face of the fluid metal is marked on the outside of the Fig. 52. glass with a pencil. In doing this, the operator must be aware of the cohesive attraction existing between the mer- cury and the glass, the tendency of which is to depress that part of the surface of the metal which verges towards the sides of the tube ; hence a mark made to correspond with that sur- face would include a space in the tube somewhat larger than that occupied by the metal. Faraday (' Chemical Manipulation,' p. 75) proposes to get rid of the difficulty by using mercury which is not quite clean, but which, from containing a little of some other metal, as tin, lead, etc., has a film formed on its sur- face which gives it a flat surface in the tube. A second, and a third, etc., measure of mercury is introduced and marked, till the tube is filled. It is then emptied of its contents, and covered 242 QUANTITATIVE ANALYSIS. with a thin coat of transparent engraver's varnish, by means of a camel's-hair pencil. When dry, the pencil-marks, which are distinctly seen through the varnish, are laid bare by a sharp steel dividing-instrument, every fifth division being made somewhat longer than the others by way of distinction. The figures at each tenth division are cut through the varnish by the point of a steel pen, and the whole is then covered with hydrofluoric acid, which, attacking the glass where it is laid bare at the divisions and figures, effectually engraves them. If a relative division of the tube only is required, it is a matter of no consequence what the capacity of the little gauge maybe; but when the ob- ject is to graduate tubes with reference to a standard measure, as, for instance, the cubic inch, then it will be found most convenient to weigh into it successive por- tions of mercury : thus, to divide the tube into T Vths of a cubic inch, successive ad- ditions of 342*535 grains of mercury must be made, to divide it into T ^-oths the Fig. 53. weight of the metal must be 34'25 grains. In making the adjustments of the mercury in the pan of the balance, the small graduated pipette, Fig. 53, provided with a stop-cock and air screw, will be found very useful. The Burette and Pipette. Eor delivering small and definite quantities of liquids, as in the ope- rations of volumetric analysis, the burette is an indispensable instrument. Various forms have been given to it, some of which are shown in Pigs. 54, 55, 56. Fig. 54 represents Gay-Lus- sac's burette, which is very delicate, though rather liable to break, and not therefore so well suited as the others for the general work of the laboratory ; if however pieces of cork be inserted between the two tubes, and then string bound round both, as shown in the figure, the danger of fracture will be almost entirely obviated. Being round at the bottom, it requires a support, which may be a block of wood perforated by an oval hole. Pig. 55 is known as Binks's burette ; the Fig. 54. ON QUANTITATIVE OPERATIONS. 243 drop tube is here at the top, instead of at the bottom of the tube. The use of this instrument requires a steady hand, but after a little practice, a liquid may be delivered from it with great pre- Fig. 55. Fig. 56. cision. Eig. 56 is Mohr's burette, which is applicable to all cases in which the contact of organic matter, such as india-rubber, is not injurious to the liquid. It consists of a graduated tube 244 QUANTITATIVE ANALYSIS. drawn out at the end, on which is slipped a small piece of vulcanized india-rubber tubing, in the other end of which is in- serted a very small glass tube as a mouthpiece. The india-rub- ber is confined by a brass spring, or pinchcock, (Fig. 57.) " The advantages possessedby this instrument," observes Mr. Sutton (' Handbook of Volumetric Analy- sis'), "are that its constant up- right position enables the operator at once to read off' the number of degrees of test- solution used for any analysis. The quantity of fluid to be delivered can be regulated to the greatest nicety by the pressure of the thumb and finger on the spring clip or pinchcock, and the instrument not being held in the hand, there is no chance of increasing the bulk of the fluid by the heat of the body, and thus lead- ing to incorrect measurement, as is the case with Binks's or Gay-Lussac's form of instrument." Burettes are graduated either in accordance with the French decimal system, or in English grains. The great advantage possessed by the former is its uniformity throughout. Thus the cubic centimetre is the exact measure of a gramme of distilled water at 4 C., or 39 F. The kilogramme is the weight of a cube of distilled water at the same tempera- ture, whose side measures one decimetre ; it is 1000 grammes, and occupies the volume of 1 litre, or 1000 cubic centimetres. The Binks burette should be constructed to hold 50 cubic cen- timetres, and of a length of about 2 feet. For the purpose of preparing standard solutions, it is convenient to have cylinders of a litre capacity, accurately graduated into 100 parts ; flasks, also graduated for 100, 200, 250, 300, and 500 cubic centi- metres, will also be found very convenient. "When the Eng- lish system is used, the burette should be constructed to hold 1000 grains, and be divided into 100 equal parts, and Mr. Sutton proposes that one of these divisions, or 10 fluid grains, should be called a decem, a term corresponding to the cubic centimetre, bearing the same proportion to the 10,000 grain measure as the cubic centimetre does to the litre, namely, the one-thousandth part ; the same term has however been used by Mr. Ackland to represent the tenth part of a gallon. An- other decimal system, constructed on the gallon measure 70,000 grains, has been proposed by Mr. Griffin ; the unit is ON QUANTITATIVE OPERATIONS. 245 called the septem, equal to 7 grains, which bears therefore the same relation to the pound =7000 grains as the cubic centi- metre does to the litre, or as Mr. Button's decem to the 10,000 grains. In reading burettes, the lower part of the curve formed by the surface of the liquor is taken as the true level, the observation of which is facilitated by pasting a piece of black paper on a card, so as to cover one-half of it, and then placing this card behind the burette ; if this be moved until the line of separation of the two colours is about one-sixteenth of an inch below the bottom of the curve, the division may be easily read. Of pipettes, there are two kinds, those which deliver one certain quantity only ; and those which are graduated so as to measure small quantities of so- lutions. The delivery pipettes (Fig. 58, a and 5) have marks on their necks, indicating the amount they contain when full. It is con- venient to have several of each of these pipettes of different capaci- ties, from 100 to 10 cubic centi- metres. The graduated pipette c is used, like the burette, for mea- suring small quantities of solu- tions. The tops of all pipettes should be circular, and narrowed slightly, so as to be more easily closed by the finger, which should be slightly moistened, in order to produce an air-tight closure. For volumetric analysis, the following graduated instruments are recom- mended as a useful set by Mr. Scott (' Handbook of Volumetric Analysis ') : c. c. c. c. One Gay-Lussac's burette, containing 50, graduated to % Two Mohr's burettes . . 100, graduated to J PAET II. 58 - 246 QUANTITATIVE ANALYSIS. C. C. C. C. One Mohr's burette . . containing 60, graduated to i One Mohr's burette . . 20, graduated to -^ One graduated pipette . , 20, graduated to i Two graduated pipettes, containing 10 c.c., graduated to T Vc.c. Delivery pipettes for 1, 5, 10, 25, 100, c. c. respectively. A mixing cylinder, to contain 1 litre, graduated to 10 c. c. Two flasks, 1 and J- litre respectively. 4. DESICCATION. 149. Before the quantitative analysis of any substance can be proceeded with, it is absolutely necessary that its relations to water should be accurately ascertained. In some cases, it is merely necessary to obtain the body free from moisture pre- vious to weighing it for analysis ; in other cases, the amount of water present has to be ascertained. Bodies differ very much in their relative powers of absorbing and retaining water, but it is necessary carefully to distinguish between the water which a substance may contain in accidental admixture, and that which belongs to it essentially, being a part of its constitution ; in the preliminary operation of drying, this latter water must not be interfered with. There are some salts which cannot be exposed to a dry air at common temperatures without losing the whole of their water of crystallization ; others lose a part only, and others again absorb water from a moist atmosphere, eventually running into a liquid. Sulphate and carbonate of soda are examples of salts possessing the first property, viz. that of efflorescence, in a remarkable degree. The former, which crystallizes in six-sided prisms, contains ten equivalents, or 56 per cent, of water; by exposure to the air, it loses the whole of this water, and falls into a white powder. Carbonate of soda, which crystallizes in flat, oblique, rhomboidal prisms, also contains ten atoms of water of crystallization, the whole of which it loses when exposed to the air, falling, like the sul- phate, into a white powder. Common phosphate of soda affords an example of a salt possessing the second property, viz. that of losing a part of its water of crystallization only, when ex- posed to the air. In its crystalline state, that of oblique rhombic prisms, it contains twenty-five equivalents, or about 62'7l per cent, of water; of these, it loses ten atoms by ex- posure to the air, and. fourteen more at the temperature of boiling water ; the remaining atom of water must exist in the ON QUANTITATIVE OPERATIONS. 247 salt in a different state of combination from the others, as it is only expelled by a red-heat, and by its loss the nature of the salt is completely changed. Thus, on evaporating the aqueous solution of the salt, which has been dried at 212, it is repro- duced in its original state, viz. with twenty-five equivalents of water of crystallization ; but, on evaporating a solution of the salt after it has been ignited, we obtain a salt which contains only ten .equivalents of water of crystallization, and which pro- duces, with a neutral solution of nitrate of silver, a white pre- cipitate, whereas the precipitate formed by the original salt is yellow. As an illustration of a salt which gains water by ex- posure to the atmosphere, carbonate of potassa may be quoted. This salt may be obtained in the form of oblique rhombic octa- hedra, containing two equivalents of water ; but, by exposure to a moist atmosphere, it speedily loses its crystalline form, and becomes liquid. Chloride of calcium is also a salt eminently remarkable for its power of absorbing water, a property which renders it very valuable for drying gases for experimental pur- poses, and for removing water from liquids, as in the rectifica- tion of alcohol. The preparation of efflorescent and deliquescent crystals in a state fit for analysis is rather difficult ; it is obviously inadmis- sible to expose them to the air ; they must, therefore, be de- Pig. 59. prived of their water of admixture by pressing them in a finely pulverized state between folds of bibulous paper until there are s 2 248 QUANTITATIVE ANALYSIS. no longer signs of moisture on the latter. There are many substances, particularly organic, which, though they do not lose their water of admixture in a dry atmosphere at common tem- peratures, cannot nevertheless be exposed to the temperature of 212 without undergoing decomposition. The exsiccation of these substances is effectually accomplished by placing them in convenient vessels underneath the receiver of an air-pump, in close proximity to some substance having a powerful affi- nity for moisture, and exhausting the air. Fig. 59 shows the arrangement; the receiver, the edge of which, ground and greased, stands on the plate of the air-pump ; the substance to be dried is placed in a watch-glass or a shallow evaporating dish, supported over a shallow basin containing concentrated sulphuric acid. On exhausting the air, a liberation of aqueous vapour takes place from the substance to be dried, which is ra- pidly condensed upon coming into contact with the sulphuric acid, and its place supplied by fresh moisture from the sub- stance, which is in its turn condensed as before, and thus the process goes on, the aqueous vapour travelling from the sub- stance to the acid till it is completely dried. The desiccation of these substances may likewise be effected without the em- ployment of the air-pump. Thus, the body may be suspended in a watch-glass from a small support over the surface of sul- phuric acid contained in a beaker, as shown in Fig. 60 ; the edge of the beaker is ground so that it may be closed accurately by a ground glass plate b, in the centre of which there is a hole, through which the cork c is inserted, and i'rom which the support hangs, bearing the substance to be dried in the watch-glass d, immediately over the surface of the sulphuric acid. Another method which may be adopted, is to place the acid and the substance to be dried on a ground-glass plate of sufficient size, and smeared with grease ; the whole Fig. 60. i s covered with a well-ground re- ceiver, and then, by raising the latter and introducing for a moment the flame of a spirit-lamp, and suddenly replacing the receiver, an imperfect exhaustion is obtained, by which the process of exsiccation is considerably facilitated. ON QUANTITATIVE OPERATIONS. 249 Substances which will bear the heat of boiling water may be dried in the water oven, B, Fig. 61. It consist of a copper box, Fig. 61. about 6 inches square, provided with a water-tight jacket or coating, the edges being soldered with brass ; it is furnished with a door in which is cut a small ventilating aperture, which can be opened or closed at pleasure by means of a handle ; one of the chimneys on the top of the box communicates with the jacket, through this the charge of water is given to the ap- paratus, it serves also for the introduction of a thermometer to enable the operator to regulate the temperature of the bath ; the other chimney communicates with the interior chamber, in which the substance to be dried is placed on a watch-glass, the edge of which is ground so that it can be covered air- tight with another similar watch-glass held tightly up- on it during the operation of weighing, by a clasp, as shown in Fig. 62 ; by this arran gement a current of air is determined through the chamber, which greatly expedites the drying of the substance. If a higher temperature than boiling water be required, the bath may be filled with a fixed - 62 ' 250 QUANTITATIVE ANALYSIS. oil, such as olive oil, the temperature being regulated by the thermometer. The student must be particular not to take the weight before the glasses have got cold. The water-bath, A, Fig. 61, is an apparatus indispensable in the chemical labo- ratory, it serves to complete the desiccation of large quantities of liquid at an unvarying temperature, it is likewise exceedingly useful for digestions, concentrations, and numerous other ope- rations in which the heat of boiling water is required. Substances which will bear a tolerably high temperature with- out decomposition may be dried in the air-bath. Fig. 63 represents the arrangement known as Kam- melsberg's air-bath ; it consists of a small cylindrical box of stout cop- per, provided with a stage about midway between the cover and the bottom for supporting the platinum crucible or dish containing the sub- stance to be dried; the cover is provided with two apertures, one is left open to serve as a chimney for the escape of the vapour, the other is closed with a cork, through which a thermometer passes to govern the temperature ; the bath is heated over a small spirit or gas lamp on the ring of the retort stand, and by a judicious manage- ment of the flame the temperature may be kept for any length of time perfectly uniform. Fig. 64 represents Taylor's air- bath, and is a much more conve- nient apparatus than the one just described. It consists of a cylin- drical box of copper or japanned tinned iron, provided with a jacket or double casing, and a false bottom pierced with holes for the more equable diffusion of air through the interior. It is provided with a thermometer for regulating the temperature^ and .the great advantage which it possessed over Eammels- berg's bath, besides its size, consists in its being furnished with a tall chimney, the effect of which is to determine a powerful Fig. 63. ON QUANTITATIVE OPERATIONS. 251 current of air through the chamber, thereby greatly increasing the rapidity of the desiccation. If deemed necessary, the cover of the box being closed, a tube of chloride of calcium may be adapted to the aperture seen at the bottom of the back of the box, and thus a current of dry air swept through the appa- ratus. The substances to be dried are placed in dishes or watch-glasses on a trellis-work of iron or copper wire, supported about an inch above the false bottom. Fig. 65 represents an arrange- ment for drying substances in a current of dry air produced by the efflux of water. For this purpose a known weight of the substance is introduced into the small bent- glass tube which has also been weighed, the body of this tube is plunged into a copper water- bath 6, charged with a saturated solution of common salt ; it is kept in its place by a cover furnished with two apertures for the arms of the drying tube ; the wider arm is united, by means of a bent tube and a caoutchouc connector con- taining fragments of chloride of calcium, with a tube, and the other end is connected with another chloride of calcium tube, bent at right angles so as to pass through the cork of the aspi- rator A, and reach down nearly to the bottom. The corks must fit the bottle perfectly air-tight, and all the joints and con- nections of the whole apparatus must be perfect. The bottle A is filled with water, and on exhausting the siphon it flows out in a small stream, which may be regulated by the cock $, its place being supplied by the air drawn through c, and which becomes dried during its passage through the chloride of cal- cium tube. The bath is charged with water, a saturated solu- tion of common salt, or of chloride of calcium, according to the degree of heat required, and it is kept boiling by means of 'a spirit or gas lamp placed underneath. Fig. 64. 252 QUANTITATIVE ANALYSIS. Fig. 66 shows the arrangement for drying at any required Fig. 65. temperature in vacuo. The substance under examination is introduced into a glass tube #, which passes through a copper collar into the Bammelsberg air-bath/; this tube is closed air-tight with a good sound cork, into which a bent glass tube is inserted, establishing a communication with the chloride of calcium tube d, and finally with the exhausting syringe b. By raising and depressing the piston a few times, a Fig. 66. tolerable vacuum is produced in the tube, and by means of the stopcocks a and (?, air may be admitted from time to time, ON QUANTITATIVE OPERATIONS. 253 which becomes deprived of all moisture previous to entering the tube containing the substance, by passing over the chloride of calcium tube d. In some cases it may be desirable to exclude atmospheric air altogether from contact with the substance ; in such cases car- bonic acid or hydrogen may be introduced into the tube e by connecting the stopcock a with a proper gasometer filled with the gas required. An apparatus contrived by Rochleder for this purpose is shown in Fig. 67. The TJ-shaped chloride of Fig. 67. calcium tube on the right of the figure is connected by means of an india-rubber tube with a caoutchouc bag or bladder containing carbonic acid ; the substance to be dried, contained in the wide glass tube, is placed in a flask immersed in an oil- bath, the temperature of which is regulated by a thermometer. The TJ-shaped chloride of calcium tube on the left of the figure is connected with an air-pump, and the cock being closed, the air in the flask is rarefied by a few strokes of the piston ; the cock of the air-pump is then closed, and the cock communicating with the gas opened, upon which dry carbonic gas enters the flask. The operation is repeated two or three times, till the appa- ratus is completely filled with dry carbonic acid. The oil-bath being now raised to the requisite temperature, the flask is ex- hausted by the air-pump, fresh carbonic acid being from time to time admitted. The operation lasts about an hour. In substances which will bear a red-heat without decompo- 254 QUANTITATIVE ANALYSIS. sition, the amount of water may be determined by heating them in a combustion tube with perfectly dry carbonate of lead, and collecting the water expelled in a properly arranged chlo- ride of calcium tube. The details of the operation will be un- derstood when the methods of conducting organic analyses are described. Tor the methods of conducting various other operations of quantitative analysis, such as solution, precipitation, collect- ing and washing precipitates, evaporation, distillation, etc., the student is referred to Chapter I. 255 CHAPTEE VIII. ON THE QUANTITATIVE ESTIMATION OF SUBSTANCES AND THEIE SEPAEATION FEOM EACH OTHEE. 150. The following Table exhibits the names, symbols, and chemical equivalents, or combining proportions by weight, of the elements, according to the latest authorities. (The names of rare substances are printed in italics.) 1 I Element. Equiva- lent Number H=l. Principal Compounds with Oxygen. Al Aluminum 13-75 AL)Oo, Alumina Sb AH Antimony (Stibium) Arsenic ...... 122-0 75-0 SbO 3 , SbO 45 SbO 5 As O 3 , As O 5 : arsenious acid, arsenic Rft 68-5 acid BaO, BaO 2 Bi E Bismuth . . . ^ . 210-0 10.9 Bi0 3 B O 3 : boracic acid "Rr Bromine 80-0 Br0 5 Cd Ci Cadmium Calcium .... . 56-0 20-0 CdO CaO c Carbon 6-0 CO, CO 2 Ce cn Cerium Chlorine . . . 46-0 35-5 Ce 2 3 CIO, C10 3 , C10 5 , C10 7 O Chromium 26-27 Cr O 3 Co Cobalt 29-5 CoO Cs 133-0 CsO Cu D E Copper (Cuprum) . . Didymium Erbium 31-75 48-0 C^O, CuO DO F a Au H Fluorine Glucinum Gold (Aurum) . . . Hydrogen .... 19-0 4-66 196-66 1-0 GO AuOg HO HO 2 11 T Ilmenium 127-0 10,, 10, 256 QUANTITATIVE ANALYSIS. I Element. Equiva- lent dumber. H=l. Principal Compounds with Oxygen. Tr 98-56 Ir 2 O 3 Fe La Pb Li Iron (Ferrum). . . . Lantanium Lead (Plumbum). . . 28-0 46-0 103-5 7-0 FeO, Fe 2 O 3 , FeO 3 LaO PbO, PbO 2 LiO Mg Mn Magnesium Manganese .... 12-16 27-5 MgO MnO, Mn,O,, Mn0 9 , Mn0 5 Mn0 7 S g INTi Mercury (Hydrargyrum) Molybdenum .... Nickel .... 100-0 48-0 29-5 Hg 2 0, HgO. MO,MO 2 ,MO, NiO TVb Niobium .... 48'8 Nb0 2 IV- 14-0 NO,NO 2 ,NO,, N0 4 , NO 5 OR 99-5 OsO, Os 2 O 3 , 6s 2 , OsO 3) Os O 5 o Oxygen . , 8-0 HO Prl Palladium . . . 53-24 Pd^jO, PdO, PdO 2 P Pt Phosphorus .... 31-0 98 '56 P 2 0, PO, P0 3 , P0 5 PtO, PtO 2 K Eo Potassium (Kalium) . . Rhodium , . , 39-0 52-16 KO, K0 3 EoO, Eo 2 O 3 Eu Eb Se Si Ruthenium Rubidium Selenium 52-11 85-36 39-75 14-0 EuO, Eu 2 O 3 , EuO 2 , EuO 3 Se0 2 , Se0 3 SiO 2 Ag Na Sr Ta Te Silver (Argentum) . . Sodium (Natrium) . . Strontium Tantalum or Columbium 108-0 23-0 43-75 68-80 64*50 Ag 2 0,A g O,Ag0 2 NaO SrO Ta0 2 TeO 2 , TeO, Tb ? TbO Tl Th Sn Ti W TT Thallium Thorinum Tin (Stannum) . . . Titanium Tungsten (Wolfram). . 202-96 59-5 59-0 25-0 92-0 60-0 T10 ThO SnO, SnO 2 TiO, TiO 2 WO,, WO, uoru 9 o, Y 68-5 YO, YO 91 YO, Y 32-0 YO 7m Zinc 32-75 ZnO Zr Zirconium 44-75 Zr0 2 In describing the methods to be pursued for bringing the several substances mentioned in the above Table into a state fit for estimation, and for separating them quantitatively from other substances, we shall follow, as nearly as possible, the same arrangements which we adopted in treating of the com- POTASSIUM. 257 portraent of substances with reagents, viz. arrange them into a series of groups. With respect to the bases there will be no difficulty in following out this method : in treating of the acids, a variation may probably appear desirable. I. BASES. GROUP I. The Metals of the Alkalies proper : Potassium, Sodium, Ammonium, Lithium. 151. POTASSIUM. This metal is quantitatively estimated as sulphate, as chloride, as nitrate, and as double chloride of platinum and potassium. Quantitative determination as Sulphate. Potassium is in general estimated in the form of this salt when no other bases are present ; the solution containing it must be carefully eva- porated to dryriess, and the residue dried for some time before it is transferred to a platinum crucible to be ignited ; this pre- caution is necessary, to guard against a loss from decrepitation. It fuses at a strong red-heat, and at a very high temperature it volatilizes. Its composition is One equivalent of KO . . . 47 . . 54-02 One equivalent of S0 3 . . . 40 . . 45'98 One equivalent of KO,S0 3 . 87 . .100-00 If the solution contain excess of sulphuric acid, the hydrated bisulphate is obtained by evaporation ; this salt must be reduced to the neutral sulphate, which is done by igniting it in a plati- num crucible containing a fragment of carbonate of ammonia, and loosely shut with its cover, the excess of sulphuric acid flies off in an atmosphere of carbonate of ammonia ; a gentle igni- tion must be continued till the salt assumes the solid state, the neutral sulphate being far less fusible than the acid sulphate. Quantitative estimation as Nitrate. "When the alkali exists in a solution in the form of nitrate it may be weighed as such ; it should not be fused, as it is thereby partially decomposed. It is anhydrous. Its composition is One equivalent of KO . . . 47 . . 46'54 One equivalent of NO 5 . . . _54 . . 53'46 One equivalent of KO,N0 5 . 101 . . 100-00 258 QUANTITATIVE ANALYSIS. Quantitative estimation as Chloride. Potassium may likewise be weighed in the form of this salt, when it exists as such, in a solution. It should not be heated above feeble redness, and the crucible should be loosely covered, or a loss may be sus- tained. Its composition is One equivalent of K ... 39 . . 52 '35 One equivalent of Cl . . . 35 '5 . . 47 '65 One equivalent of KC1 . . 74'5 10OOO Quantitative estimation as double Chloride of Platinum and Potassium. For this purpose the alkali must be in the state of chloride, the solution is evaporated to a small bulk, and excess of bichloride of platinum and free hydrochloric acid added ; the mixture is then evaporated to dryuess on the water bath, and the crystalline residue digested with spirits of wine, which re- moves the excess of bichloride ; it is then transferred to a weighed filter, washed with spirits of wine and dried at 212. It composition is One equivalent of K . 39 . 16'00 = 19'262 KO One equivalent of Ft . 98 '50 . 40'37 Three equivalents of Cl . 106'5 . 43-63 Oneequiv. KCl,PtCl 2 . 244'00 lOOOO If the operator be certain that no other base but potassa is present, and that it exists in the solution in the state of neutral sulphate, or as chloride of potassium, it may then be deter- mined in an indirect manner, by estimating the sulphuric acid as sulphate of baryta, or the chlorine as chloride of silver, and calculating the amount of alkali present from the quantity of sulphuric acid, or of chlorine, thus obtained. 152. SODIUM. This metal is quantitatively determined as sulphate, as carlo- nate, and as chloride. Quantitative estimation as Sulphate. The same method is followed as with sulphate of potassa. It does not decrepitate. Its composition is One equivalent of NaO . . 31'0 . . 43*66 One equivalent of S O 3 . . 40'0 . . 56'34 One eqvivalent of NaO,SO 3 7To . . 100-00 Quantitative estimation as Chloride. The salt must be well SODIUM. ' 259 dried before it is ignited, to prevent decrepitation ; the crucible should be loosely covered. Its composition is One equivalent of Na . 23'0 . 39'31 = 52'99 NaO One equivalent of Cl . 35 '5 . 60'69 One equivalent of Nad 58'5 10OOO Quantitative estimation as Carbonate. As carbonate of soda does not attract moisture from the air so rapidly as the corre- sponding potassa salt, sodium may be weighed as such. It may be fused without volatilization or decomposition. Its composi- tion is One equivalent of NaO . . 31'0 . . 58-49 One equivalent of C0 2 . . 22-Q . . 41-51 53-0 100-00 153. Separation of Soda from Potassa. The mixed alkalies are converted into chlorides ; in most cases this may be done by simply digesting and evaporating with hydrochloric acid ; if, however, the alkalies are combined with sulphuric acid, that acid must first be precipitated by chloride of barium, and the excess of baryta removed by carbonate of ammonia mixed with a little caustic ammonia ; the solution is filtered, and the filtrate is evaporated to dryness, and ignited, the residue is redissolved in hydrochloric acid, and the alkalies are thus brought to the state of chlorides. If the alkalies are combined with phosphoric acid, the solu- tion is evaporated to dryuess, ignited, redissolved in water, and precipitated by neutral nitrate of silver; the fluid is filtered from the precipitated phosphate of silver, and the excess of nitrate of silver is' removed by hydrochloric acid. If the alkalies are in combination with loracic acid, the best method of separating them is by decomposing the boracic acid into gaseous fluoride of boron ; for this purpose the dry com- pound is digested in a platinum crucible with 3 or 4 parts of pure powdered fluor-spar and concentrated sulphuric acid, and the heat is continued as long as fumes continue to be evolved, the alkalies are thus converted into sulphates, which are changed into chlorides as above directed. The mixed alkaline chlorides, being carefully weighed, are dissolved in a small quantity of water and mixed with excess of an aqueous solution of bichloride of platinum, the solution is evaporated to dryness on the water-bath, and the dry mass 260 QUANTITATIVE ANALYSIS. is treated with a mixture of ether and spirits of wine ; the po- tassium is estimated as above described; the spirits of wine poured on the evaporated residue must acquire a yellow colour, if it does not, it is a proof that sufficient bichloride of platinum has not been added, and the experiment must be repeated with a fresh portion. Berzelius recommends to mix the dry chlorides with 3f times their weight of the crystallized double chloride of platinum and sodium, a quantity just sufficient to convert the whole mass into double chloride of platinum and potassium, supposing it to consist entirely of chloride of potassium ; the process is conducted precisely as before. The soda is either calculated from the difference between the quantity of chloride of potassium found, and the original weight of the mixture analysed, or in a direct manner, as follows : the filtrate from the double potassium salt is mixed with solution of sal ammoniac in excess, and strong alcohol added ; all the platinum salt is thus removed, and the colourless filtrate being evaporated to dryness, the residue is gently ignited and weighed as chloride of sodium. Maumene's method. For industrial operations, an approxi- mate result may, according to Maumene, be attained by adding to the mixture of the new alkalies a salt of alumina, and cal- culating the amount of potash from the weight of the alum formed. Pagenstecher's method of estimating the amount of Soda in crude Potash. (Chem. Gaz., January, 1848.) This is founded on the property possessed by a saturated solution of sulphate of potassa of still dissolving a large amount of sulphate of soda. A certain weight of the sample to be examined is mixed with water, and sulphuric acid added to it until the liquid has an acid reaction ; it is then evaporated to dryness, and the residue ignited and weighed. The powdered saline mass is well agi- tated in a graduated cylinder with six times its weight of a concentrated solution of sulphate of potassa, the liquid drawn off the sediment with a siphon, and the same quantity of the solution of the sulphate of potassa again poured over the re- sidue. After some time the residue is brought upon a weighed filter, the funnel covered during filtration, the filter, when the liquid has drained off, weighed moist, and again, after being dried at 212, the difference is the evaporated water of the so- lution of sulphate of potassa, the degree of concentration of which is known. It is known, therefore, how much of the SODIUM FROM POTASSIUM. 261 salt was dissolved in the evaporated water ; this quantity is subtracted from the weight of the saline residue. If the potassa were free from soda, the weight of the sulphate of potassa now remaining must be the same as that first obtained ; but if it contained soda, this has been removed as sulphate of soda, and the weight of the first saline residue has been re- duced. From the loss, the amount of the soda present can be calculated ; if the loss = Y, the amount of the soda is calcu- lated thus : Eq. ofNaO,S0 3 . Eq. of NaO,C0 2 . 71 : 53 :: V : X x = amount of carbonate of soda present in the specimen. It should, however, be observed that when soda is used to adulterate potassa, a kind is employed that contains about 20 per cent, of sulphate of soda. Before making the weighings, it is well to take the specific gravity of the filtered solution of the sulphate of potassa ; if it be the same as before it can have removed nothing, and if it has taken up sulphate of soda its density has naturally been raised. Indirect separation of Potassa from Soda. This method is founded on the difference which exists between the combining numbers of potassium and sodium, and is applicable to both sulphates and chlorides. It is deduced from the following con- siderations : Suppose the alkalies to be in a state of sulphates, and 100 grains of the mixture to have yielded 153 grains of sulphate of baryta=52'6 grains of sulphuric acid. Let the quantity of sulphate of potassa present be called A, and that of sulphate of soda B : then A + B = 100, or 100 B= A. That is, the proportion of sulphate of potassa in the mixture, is the weight of the mixture analysed minus the proportion of sul- phate of soda. Now one part of sulphate of potassa contains '4598 sulphuric acid. And one part of sulphate of soda contains *5634 ditto. The 52'6 grains of sulphuric acid found in the mixture must therefore be (A x -4598) + B x '5634 ; that is, the number of units present of sulphate of potassa multiplied by the quan- tity of sulphuric acid in one unit, added to the number of units present of sulphate of soda multiplied by the quantity of sul- phuric acid in one unit : PART II. T QUANTITATIVE ANALYSIS. (A x -4598) + (B x -5634) =52-6 52-6 (Bx -5634) consequently A = - ^^ but A has been shown above to be = 100 B; substituting this value, therefore, for A, we have IOO _-R_ 52-6- Bx -5634 4598 or, 100 x -4598- (B x -4598) = 52 "6- (B x -5634) or, 45-98- (B x -4598) = 52-6 - (B x -5634) and, putting the two B's on the same side of the equation, we have (B x -5634) - (B x -4598) = 52'6-45'98 P_ 52-6-45-98_ 6'62_ 63 . 9 "5634- '4598~ -1036 The mixture contains, therefore, 63 '9 per cent, of sulphate of soda, and 36'1 per cent, of sulphate of potassa. Suppose the alkalies to be in the state of chlorides, and 133 grains of the mixture to have given 287 grains of chloride of of silver =71 grains of chlorine ; Then let the quantity of chloride of potassium be . . A And that of chloride of sodium ........ B then A + B = 133; and A = 133 B. Now one part of chloride of potassium contains ........... *4765 chlorine And one part of chloride of sodium contains '6069 ditto. Therefore, as before, 71 = (A x '4765) + (B x -6069), 71 - (B x -6069) consequently A = - \ ; or, 133 x -4765- (B x -4765) = 71- (B x '6069) ; or, 63-38 - (B x -4765) = 71- (B x -6069); and (B x -6069) -(B x -4765) = 71 - 63-38 ; 7163-38 7-62 = -6069 -4765 =W4 The mixture is therefore composed of 58*43 grains of chlo- ride of sodium, and 74'57 grains of chloride of potassium. ALKALIMETRY. 263 From the above calculations the following rule for the in- direct determination of soda and potassa, when together present in a mixture, in a state of sulphates, is derived : From the quantity of sulphuric acid subtract the product obtained by multiplying the weight of the mixture into the quan- tity of sulphuric acid in a unit of sulphate of potassa, and divide the remainder by the difference between the quantity of sulphuric acid in a unit of sulphate of soda, and the quantity in a unit of sulphate of potassa ; the quotient is the quantity of sulphate of soda in the mixture. Example. In 100 grains of a mixture of the two sulphates, 50 grains of sulphuric acid are found. 50 (100 x -4598) _ 50-45-98 _ 4-02 5634 --4598 '1036 ~ T 1036 ~ the sulphate of soda ; consequently, 100 -38- 8 = 61-2 = the sulphate of potassa. The same rule applies to the determination of the alkaline chlorides, substituting chlorine for sulphuric acid ; the results are tolerably accurate, but the determinations of sulphuric acid and chlorine must be made with the greatest care. 154. Alkalimetry. The commercial value of the several varie- ties of potassa and soda depends on the amount of carbonated or caustic alkali which they contain ; it is important therefore to the purchaser of these articles, to be in possession of a simple and expeditious method of determining with tolerable accuracy the amount of available alkali in the crude salt ; a complete chemical analysis of such a heterogeneous mixture requiring great skill and time. Two methods are adopted for this pur- pose : the first, that of Gay-Lussac, depends on the constant sa- turating power of sulphuric acid of a certain determinate strength, being founded in common with every other process involving chemical combination or decomposition, on the law of definite proportions ; the second, that of 1'resenius and Will, de- pends on the determination of the amount of carbonic acid evolved from a given weight of the salt during its decomposi- tion by sulphuric acid. (I.) The Volumetric process (modification of Gray-Lussac's method). (a.) Preparation of the standard acid. Ordinary oil of vitriol is diluted with ten or twelve parts by measure of water, and the mixture allowed to cool. A quantity of the best bicarbonate of T 2 264 QUANTITATIVE ANALYSIS. soda is washed on a filter with cold distilled water, till the fil- trate when neutralized with pure nitric acid ceases to give a precipitate either with nitrate of silver or chloride of barium ; it is then well dried, and heated to low redness for some time in a porcelain or platinum crucible ; when cold it is weighed, again heated for some time, and then again weighed ; the pro- cess is complete, that is, the second atom of carbonic acid has been completely expelled, when the weight remains constant. Of the purified carbonate of soda thus prepared, 530 grains are dissolved in exactly 10,000 grains of distilled water, a solu- tion is thus obtained, 1000 grain measures of which contain 53 grains, representing one equivalent of carbonate of soda. This quantity of the alkaline solution is poured into a small beaker, and a suflicient quantity of infusion of litmus is added to communicate to it a distinct blue colour. The burette is filled exactly to with the diluted acid, and the beaker being placed on a small sand-bath over a lamp, the acid is poured in until the blue colour is changed to a bright red ; the operator must be careful not to mistake the wine-red colour which the liquor assumes from the evolution of carbonic acid, for the distinct red produced by a very slight excess of the acid, and when the point of saturation is approached, the addition of the acid must be made with great care, the alkaline liquid being in a state of ebullition. The operation being finished, the quantity of acid which has been required, is observed, and the experiment is re- peated again and again, on fresh 1000 grain measures of the alkaline solutions, until perfectly concordant results are ob- tained. Suppose that 600 grain measures of acid have been required to effect an exact saturation of the alkaline solution, this then is the quantity of acid equivalent to 53 grains of carbonate of soda, and by adding 400 grain measures of water to every 600 grain measures of such acid, a solution is obtained, 1000 grain measures of which contain exactly 49 grains of H 0, S O 3 equi- valent therefore to 53 grains of carbonate of soda, to 40 grains of hydrate of soda, to 69 grains of carbonate of potassa, and to 56 grains of hydrate of potassa. Mohr prefers oxalic to sulphuric acid, for the following rea- sons : 1st, because it is easily obtained in a state of purity by recrystallization, and in its crystalline form its composition is perfectly definite (C 2 3 HO, + 2 HO); 2nd, because being a solid it is better adapted for weighing than a liquid, and is not ALKALIMETRY. 265 liable to deliquesce or effloresce ; and 3rd, because it is quite fixed in heated solutions. If oxalic acid be adopted as the base of the volumetric alka- limetrical system, 63 grains of the pure crystallized acid are dissolved in water, and the solution diluted to 1000 grain mea- sures at 60 F. This quantity of solution will then neutralize exactly 40 grains of hydrate of soda, or 56 grains of hydrate of potassa, and 53 grains of carbonate of soda, and 69 grains of carbonate of potassa. In order accurately to control this solu- tion, Mohr tests it with a solution of caustic potassa or soda ; he prefers the former, as it has less action on the glass. Scott prefers caustic soda, from the greater facility with which it may be obtained free from silicic and sulphuric acids. It is imma- terial which alkali is employed. The caustic solution is diluted with distilled water until it is exactly equal in strength to the normal acid. It is preserved in a bottle, through the stopper of which is inserted a tube containing a mixture of equal parts of Glauber's salt and quicklime, previously mixed together, dried, and ignited. (b.) Performance of the AlTcalimetricallassay. When a sam- ple of commercial soda has to be tested, 530 grains, or a sub- multiple of that quantity, should be weighed out from the thoroughly powdered and mixed sample ; after being dried, it should be gently ignited in a porcelain or platinum crucible, and allowed to cool without exposure to the air, or under the receiver of the air pump, as represented in Eig. 59 ; when cold, it is again weighed ; the loss indicates the amount of moisture. It is then washed into a beaker, in which it is dissolved ; should any insoluble residue remain, it is filtered off, dried, and weighed ; the clear filtrate is made up to exactly 10,000 grain measures, sup- posing 530 grains to have been taken, or to 5000 grain mea- sures, supposing half that weight to have been used. The so- lution is well mixed together, and from it 1000 grain measures are taken, transferred to a beaker, the solution rendered blue with litmus, heated to boiling, and then tested with the normal acid in the manner just described, until the neutral point is reached ; the process may be repeated several times, if necessary, to be certain of the accuracy of the analysis. In order how- ever to avoid all ambiguity arising from the carbonic acid, a sufficient quantity of acid may be added to render the liquid very decidedly red, and then the normal caustic alkali added drop by drop, until the liquid suddenly changes to violet-blue ; 266 QUANTITATIVE ANALYSIS. the number of divisions of the burette that have been required to effect this, must be deducted from the quantity of acid origi- nally used; by this "backward" or "residual" method very sharp results are obtained. Suppose 850 burette divisions of the normal acid have been required, the following calculation gives the amount of real carbonated alkali in the sample : As 1000 : 850 :: 53 as x = 45, the amount of carbonate of soda in 53 grains of the sample. The carbonate of soda of commerce sometimes contains two or three salts arising from imperfect decomposition, which inter- fere with the accuracy of the alkalimetrical assay. These salts sue sulphide of sodium (NaS) ; hyposulphite of soda (NaO,S 2 2 ); and sulphite of soda (NaO,S0 2 ) ; the presence of the former is indicated by the solution blackening salts of lead ; all error arising from the presence of these salts is removed by eva- porating the solution to dryness after the addition of a few crystals of chlorate of potassa, and heating the residue ; the sulphur is hereby converted into sulphuric acid, which is harmless. It is true that if hyposulphites are present another error is introduced, S 2 O 2 becoming by oxidation 2 S 8 , so that one equi- valent of sulphuric acid is free to neutralize one equivalent of alkali, but as it is very seldom that this oxygen compound of sulphur is present in quantity, it is of little practical impor- tance. But the sample may contain caustic soda, as is the case with the crude soda of commerce, and the alkaline leys used in soda, paper, and other manufactures. Now as the above alkalimetrical process makes no distinction between carbonated and caustic alkalies, both exerting the same action on litmus, and both being converted into sulphates by sul- phuric acid, it is necessary in the examination of such mixtures to determine the amount of carbonic acid by an independent experiment. This object is accomplished by adding to a hot solution of the salt, chloride of barium as long as a precipitate takes place, this precipitate contains in the form of carbonate of baryta all the carbonic acid which existed in the sample in the form of carbonated alkali ; it is collected on a filter, washed rapidly without exposure to the air, and the amount of car- bonic acid determined by one of the processes recommended ALKALIMETRY. 267 under " Carbonic acid." The amount of carbonic acid being thus determined, also by the volumetric assay of another por- tion of the sample, the total amount of alkali; the relative proportions of carbonated and caustic alkali are easily ascer- tained. In the examination of commercial pearlashes, the mode of operating is precisely the same, but as the equivalent of carbo- nate of potassa is 69, the weight of the sample to be operated upon to make in solution 10,000 grain measures, will be 690 grains. It may sometimes be convenient to employ a normal sulphuric acid, 1000 grain measures of which shall be equiva- lent to precisely 100 grains of the anhydrous caustic alkali ; for this purpose it is obvious that different standard acids will be required for soda arid for potassa, that for soda must be of such a strength that 1000 grain measures shall saturate exactly 171 grains of pure carbonate of soda, and that for potassa must be precisely equivalent to 146' 8 grains of pure carbonate of potassa, the advantage of the standard acid above described is its equivalency both to potassa and soda. Mr. Scott ('Hand- book of Volumetric Analysis') gives the following directions for preparing the infusion of litmus: " Treat 1 Ib. of litmus with about 7 or 8 Ibs. of water, raise the whole to the boiling-point, and then after hav- ing allowed to cool slightly, throw on a filter in order to re- move the mineral and other impurities al- ways present in the commercial article. The solution thus prepared should be kept in an open bot- tle, as if the air be Fig. 68. 268 QUANTITATIVE ANALYSIS. not allowed access to it, it is very liable to undergo decom- position, the addition of a little spirit improves its stability. A fixed quantity for use in each analysis should be taken out by means of a small graduated pipette, in order to ensure the same intensity of colour in each analysis. The infusion of litmus is slightly alkaline, but it may be rendered perfectly neutral, by boiling with a small quantity of chloride of am- monium. Those who are engaged in the frequent practice of alkalime- trical assays, will find it very convenient to prepare large quan- tities of normal solutions of acid, and caustic alkali, and to keep them in large Wolfe's bottles, mounted on a convenient stand, as in Fig. 68. The acid should be coloured red by the addition of a few drops of infusion of litmus, and the alkali blue. The manner of arranging the pipettes so that they may be filled without detaching them from the apparatus, will be understood from a simple inspection of the figure. The air-tube of the bottle containing the normal caustic alkali, may be replaced by a wider tube containing a mixture of Glauber's salt and quick- lime previously ignited. (II.) Method of Fresenius and Will. The alkalimetrical pro- cess above described seeks its object by determining the amount of alkali, calculating it from the measure of acid which it requires for its neutralization. In the method now to be described the result is obtained by determining the quantity of carbonic acid with which the al- kali was combined. The process is conducted as follows (Fig. 69) : a and b are two flasks. Wide- mouthed medicine bottles even may be employed, b should have a capacity of from 2^ to 2 ounces of water: it is advisable that a should be somewhat smaller, say of a capacity of about 1^ to 2 ounces. Both flasks are closed by means of doubly perforated corks : these perforations serve for the reception of the tubes, Fig. 69. c, J, and e. All these tubes are ALKALIMETRY. 269 open at both ends ; when operating, the end of the tube d is closed by means of a small piece of wax. The substance to be examined is weighed and projected into the flask b, into which water is then poured to the extent of one-third of its capacity ; a is filled with common English sulphuric acid to about half its capacity. Both flasks are then corked, and the apparatus is weighed. The air in the whole apparatus is then rarefied by applying suction to the tube c ; the consequence is, that the sulphuric acid contained in a ascends into the tube e, and thus a portion of it flows over into b ; immediately upon its coming into contact with the carbonate contained in this flask, the evolution of carbonic acid begins briskly. The peculiar con- struction of the apparatus compels the carbonic acid evolved to pass through the sulphuric acid contained in a before it is per- mitted to escape through the tube c, this being the only aper- ture of the apparatus. It is obvious that during this transmission through sulphuric acid all the moisture with which the carbonic acid may be charged will be retained more completely than could be done in any other manner. Upon the influx of the sulphu- ric acid, the fluid in b becomes heated, and expands together with the air contained in the flask : upon cooling, both acquire their original volume again, owing to which a new portion of sulphuric acid flows over into b as soon as the evolution of gas has ceased : the process is, however, expedited by applying suc- tion to the tube c every time the evolution of gas ceases, and in this way the operation may be accomplished in a few minutes. "When the carbonate is completely decomposed (which is ascertained by no further evolution of gas taking place upon the influx of fresh portions of sulphuric acid into 5), a some- what larger quantity of the sulphuric acid contained in a is, by means of renewed suction, made to pass over into b : the fluid in this flask becomes hereupon heated to such an ex- tent as to expel all the carbonic acid which it had absorbed in the course of the operation. As soon as the evolution of gas has completely ceased, the aperture of the tube d is opened by taking on the piece of wax, and suction applied to the tube c, until all the carbonic acid still contained within the apparatus is replaced by air. The apparatus is then allowed to cool, wiped dry, and weighed. The loss of weight indicates the amount of carbonic acid which was contained in the test spe- cimen. A Common apothecary's balance, that will turn with one-sixth of a grain, is sufficiently delicate for weighing not 270 QUANTITATIVE ANALYSIS. only the whole apparatus, but the test specimen also ; and it is not one of the least of the advantages of this method, that it enables the operator to experiment on a much larger scale than is possible in an ordinary analysis. Previous to submitting the sample to analysis, it must be thoroughly dried by expo- sure to heat over a lamp for about five minutes in a Berlin cru- cible, and allowed to cool with the cover on : it must also be as- certained whether any insoluble earthy carbonates are present, which is done by dissolving the specimen in water, filtering and well washing the residue : if the sample contain sulphide, sul- phite, or hyposulphite of the alkali, a teaspoonful of yellow chromate of potass a is added to the solution in the flask, which decomposes both the sulphurous acid and the sulphuretted hy- drogen at the moment of their liberation, all the products of the decomposition, viz. sulphate of chromium, water, and sul- phur, remaining in the apparatus. The amount of caustic soda and potassa present in the specimen, the determination of which is of great importance, is found by comparing the quan- tity of carbonic acid evolved from a given weight of the alkali in its ordinary state, with that evolved from a similar quantity after it has been mixed in a moist state with carbonate of am- monia, and dried at a high temperature, by which means all the caustic alkali becomes carbonated. Should any bicarbonates be present, they are converted into neutral carbonates, by the application of a gentle red-heat: in order to ascertain their presence the solution to be examined is mixed with solution of chloride of calcium in excess, filtered, and ammonia added to the filtrate, a turbidity indicates the presence of bicarbonate. The presence of free soda in the commercial article is de- tected by the alkaline reaction which the solution of the sam- ple exhibits after the addition of chloride of barium in excess. Presenius and Will apply the same method to the analysis of carbonates, the bases of which form insoluble compounds with sulphuric acid. The apparatus is, however, somewhat mo- dified, in order to allow of the introduction of nitric acid in the place of sulphuric acid in the bottle b. For this purpose the tube d is expanded to a bulb in its upper part, and drawn out to a fine point at its lower end ; it must be adjusted into the cork of b in such a manner as to allow of its being raised or depressed, still, however, preserving the bottle air-tight. It is filled with dilute nitric acid, and a wax stopper having been inserted into its upper aperture it is introduced into the cork, so that its AMMONIUM. 271 point just reaches the surface of the water in 5, through which the carbonate to be analysed is diffused. The nitric acid is prevented from escaping from the tube by the wax stopper. The whole apparatus is weighed, the tube d is then cautiously depressed, so that its point nearly reaches the bottom of the flask, and by removing the wax stopper the nitric acid gradually escapes into b, occasioning the decomposition of the carbonate, the liberated carbonic acid escaping through the tube c, and becoming deprived of moisture by the sulphuric acid in a, pre- vious to escaping through e. When the decomposition is com- plete, the carbonic acid which has been absorbed by the water in b is expelled by plunging the apparatus into hot water, air having been previously drawn through the flasks by suction at the tube e. As soon as the whole is cool the flasks are wiped dry and weighed. The loss indicates the amount of carbonic acid. It is scarcely necessary to say that nitric acid is here employed, in consequence of its forming soluble compounds with the bases of such carbonates as may require to be analysed in this apparatus, viz. those of lime, strontia, and baryta. 155, AMMONIUM. Ammonia is estimated as chloride of 'ammonium , and as double chloride of platinum and ammonium. It is also estimated by the volumetric process. Quantitative estimation as Chloride. When the alkali exists in a solution, either in an uncombined state, or as a carbonate, or combined with a weak volatile acid, or as chloride, it may be weighed in the form of the latter salt, for which purpose slight excess of hydrochloric acid is added to the solution, which is evaporated to dryness on the water bath, the residue being heated thereon till it ceases to lose weight ; at this temperature the loss from volatilization is almost inappreciable. Its com- position is One equivalent of NH 4 . . 18'0 . . 33'65 One equivalent of Cl . . . 35'5 . . 66'35 One equivalent of NH 4 01 . !ftl> . . lOO'OO Quantitative estimation as double Chloride of Platinum and Ammonium. Ammoniacal salts may be analysed by converting the ammonium into a double salt, with bichloride of platinum ; for this purpose the solution is supersaturated with hydro- chloric acid, and an aqueous solution of bichloride of platinum 272 QUANTITATIVE ANALYSIS. added ; the mixture is evaporated to dryness, and the residue treated precisely as the corresponding potassium salt; it is washed on a weighed filter, with alcohol, and dried at 212; its composition is One equivalent of NH 4 . 18-00 . 8'07=7'62 NH 3 =6'278 IS One equivalent of Pt . 98-50 . 44-17 Three equivalents of Cl . 106-50 . 47'76 Oneeq.ofNH 4 Cl,PtCl 2 223-00 100*00 When this double salt is heated to redness it is entirely de- composed, metallic platinum in a fine spongy form alone re- maining ; this operation may therefore be performed upon it in order to control the previous determination ; the ignition must be effected carefully in a thin Berlin crucible, a gentle heat being first applied, and gradually increased until the organic matter of the filter is entirely destroyed. The crucible should be covered at first, but the cover must be removed towards the close of the operation, the crucible being then placed obliquely, in order to favour the access of air. Ammoniacal salts may also be analysed by igniting them with a mixture of hydrate of soda and hydrate of lime ;* the ammonia is liberated from its previous combination, and is received into, and condensed in hydrochloric acid ; it is subse- quently estimated as double chloride of platinum and ammo- nium, in the manner already described. The process is con- ducted as follows : the salt to be analysed having been thoroughly dried in the water-bath, is weighed, and intimately mixed with a sufficient quantity of soda-lime to half fill a combustion tube of hard German glass, about 14 or 16 inches long and half an inch internal diameter ; in the operation of mixing, forcible pressure must be carefully avoided, or the ammoniacal salt will undergo partial decomposition, even in the cold; the combustion tube is drawn out to a point, and bent obliquely upward at its closed end : soda-lime is first introduced so as to occupy about an inch of the end of the tube, this is followed by the mixture ; the remainder of the tube is filled to within an inch of the top with soda-lime that has served to rinse out * Prepared by slaking a weighed amount of the best caustic lime, with solution of soda of such a strength that there shall be about* one part of hy- drated soda to two parts of anhydrous caustic lime ; the mixture is heated to feeble redness in a Hessian crucible, pulverized, and kept in a well-closed phial. AMMONIUM. 273 the mortar, and, finally, a stopper of recently ignited asbestos is inserted ; the tube is now laid in a horizontal position on the table, and a few smart taps given to it, in order to open a channel above the mixture for the passage of ammoniacal gas. The condensing apparatus, containing a small quantity of hydrochloric acid, is now attached to the tube by means of a perforated cork, the combustion tube is placed in the furnace, and the whole apparatus having been proved to be air-tight, the tube is gradually heated with red-hot charcoal, commencing at the anterior portion, and proceeding gradually towards the closed end, until the tube is red-hot throughout its whole length ; when the evolution of ammonia has ceased, the point of the combustion tube is quickly broken off, and air drawn through the apparatus, so as to bring the whole of the ammonia into the hydrochloric acid. The arrangement of the apparatus is shown in Fig. 70. The only source of error to be Fig. 70. apprehended in conducting this operation arises from the powerful affinity existing between hydrochloric acid and am- monia, in consequence of which the acid is apt to rush back with violence, and enter the combustion tube, thus spoiling the whole analysis. The inventors of the method, Will and Var- rentrapp, recommend mixing the substance analysed with an equal amount of sugar, which gives rise to the evolution of other and more permanent gases which serve to dilute the ammonia. The accident may be also prevented by employing a capacious condensing apparatus, and using a moderate quantity of hydro- chloric acid. The operation being over, the contents of the 274 QUANTITATIVE ANALYSIS. condensing apparatus are transferred to an evaporating basin, and the apparatus repeatedly rinsed out with water ; bichloride of platinum in excess is then added, and the remainder of the process conducted as has been already directed. Volumetric estimation of Ammonia and of Nitrogen. Instead of weighing the ammonia condensed by the hydrochloric acid in the above process as double chloride of ammonium and platinum, it may be estimated by the method recommended by Peligot, (' Comptes Eendus,' March 29, 1847.) The bulbed receiver is partly filled with a fixed weight or volume of sulphuric acid of known strength. Now, as the ammonia which combines with the acid lowers its strength ; by determining after combustion the composition of this liquid, and comparing it with that it pre- viously possessed, it is easy to ascertain the quantity of ammonia that has been condensed ; and consequently, the amount of ni- trogen yielded by the substance submitted to analysis. This operation is executed with as much rapidity as precision, by means of an alkaline solution, likewise of known strength. Peligot prefers a solution of lime in syrup. By triturating slaked lime with a solution of sugar, it dissolves in far greater proportion than in pure water ; and the compound formed pos- sesses the same alkaline reaction as if the base which it contains were free. It may be preserved for a considerable time with- out alteration, in bottles protected from contact with the car- bonic acid of the atmosphere : nevertheless, as the insoluble carbonate of lime renders the liquid turbid, it need only be filtered to serve again to determine the strength of the sul- phuric acid. The operation is conducted as follows : A certain quantity of sulphuric acid of known strength is accurately measured into the condensing apparatus. The acid employed by Peligot contains 61*250 grammes of (HO,S0 3 ) to 1 litre of water ; consequently, 100 cubic centimetres corre- spond to 2*12 grammes of ammonia, or 1'75 gramme of ni- trogen. When the combustion is finished, the acid which has served to condense the ammonia is poured into a cylindrical glass ; the apparatus carefully washed, and to this liquid, diluted with much water, a few drops of litmus are added, to give it a red colour. By means of the solution of lime in sugar, which is contained in a burette graduated in cubic centimetres and in tenths of a cubic centimetre, the acid liquid is accurately satu- rated, the operator being guided by the blue colour which the AMMONIUM. 275 liquor suddenly acquires when the point of saturation is at- tained. The quantity of alkaline liquor required to produce this effect is read off on the burette ; and as the amount of the lime compound which saturates 10 cubic centimetres of the normal solution of sulphuric acid has been determined by a previous experiment ; by subtracting from this quantity, that found for the acid which has absorbed the ammonia from the nitrogenous substance, we find the volume of the acid solution which has been saturated by the ammonia, and, consequently, the weight of the nitrogen which the body contained. ScUlcesing's method of determining the amount of Ammonia when accompanied by nitrogenous substances. (Ann. de Chim., February, 1851.) A known weight of the substance is dissolved in water and placed in a shallow vessel, over which is supported, on a glass tripod, a shallow saucer containing sulphuric acid of known strength. The ammoniacal solution is mixed with caus- tic potassa or with milk of lime, and covered with a bell glass, the edges of which rest on mercury contained in a large shal- low dish. After a period varying from 24 to 28 hours, accord- ing to the bulk of the ammoniacal solution, the whole of the ammonia is disengaged and is absorbed by the sulphuric acid, the strength of which is again determined by means of sac- cbarate of lime according to Peligot's method. This process does not answer so well with insoluble salts of ammonia, which the author directs to 'be dissolved in nitric acid ; the operation then goes on very slowly, requiring sometimes 72 hours for the complete absorption. Walker (Chem. News, Nov. 24, 1860) condenses the am- monia produced by the decomposition of nitrogenous organic substances, manures, etc., by igniting them with soda-lime, in a dilute solution of chloride of zinc (sp. gr. 1*025) using about ten fluid ounces for each analysis ; oxide of zinc is precipitated, every 40 parts of which correspond to 17 of ammonia. The results he states to be very accurate. Eor determining the amount of ammonia in rain-water, Bous- singault submits to distillation 1 litre (about 85^ oz.) of water, to which a little caustic potassa has been added, and collects and tests by the alkalimetrical method two or more successive fifths or tenths of the products. For this purpose he uses a test acid only -^th of the strength of that employed by Peligot, and a test alkaline solution not quite a third of the strength, 276 QUANTITATIVE ANALYSIS. volume for volume, of this test acid. In practice Boussiugault finds that the limit of error in the use of this dilute alkaline so- lution is about O g 2 cub. cent, measures of it, equal to about O033 milligramme, equal to 0*0005 grain of ammonia. Messrs. Lawes and Gilbert operate upon several litres of water in the first instance, reducing the bulk by successive distillations to one-half; it is thus brought to a convenient amount for final distillation and subsequent testing of measured proportional amounts of the distillate, according to Boussingault's method. The amount of ammonia in manures, the ammoniacal liquor of gas-works, etc., may be determined by distilling known weights or volumes in a flask, into which caustic potassa is allowed to flow from a suitable apparatus furnished with a clip, the product being received in another large flask containing a known volume of normal acid, the strength of which after the operation, is again tested in the usual manner. The distillation should be conducted cautiously, and the flask should be suffi- ciently large to prevent the liquid from frothing over; the end of the delivery tube should be cut slanting, and just reach the surface of the acid in the receiver ; the boiling should continue for fifteen or twenty minutes. Separation of Ammonia from Potassa and Soda. "When all three alkalies are present in combination with the same vo- latile acid, the ammoniacal salt may be expelled from a weighed portion of the mixture by careful ignition, and the amount of ammonia calculated from the loss of weight. If the mixture contain ammonia in the form of chloride or carbonate, it may likewise be expelled by heat from the salts of the other alka- lies, but if the ammonia be present in combination with a non- volatile acid, or if the mixture cannot be dried at a temperature insuflicient to expel ammonia, it must be determined by com- bustion with soda lime, and the other alkalies must be estimated from a separate portion, which has been previously gently ig- nited it to expel all the ammonia. If no potassa be pre- sent, ammonia may be separated from soda by bichloride of platinum. BARIUM. 277 GROUP II. The Metals of the Alkaline Earths : Barium, Strontium, Calcium, Magnesium. 156. BARIUM. Oxide of Barium, or Baryta, is weighed as carbonate and as sulphate, most frequently as the latter, which is completely in- soluble in water, and extremely sparingly so in all acids. Quantitative estimation as Sulphate. To the solution con- taining the .earth, heated to 212, dilute sulphuric acid is added in excess ; the fluid is well agitated, poured into a beaker, and allowed to stand till the precipitate has settled, and the super- natant fluid become quite clear: it is then transferred to a filter, the amount of ash yielded by which is known, and washed with hot distilled water until the wash-water no longer produces any precipitate with chloride of barium, it is then dried and ignited, the heat being continued until the organic matter of filter is completely destroyed, and the contents of the crucible perfectly white ; when quite cold, it is weighed. Its composition is One equivalent of BaO . . 76'5 . . 65'66 One ditto of SO 3 .... 4Q-Q . . 34'34 One ditto of BaO, S0 3 . . 116'5 lOO'OO Quantitative determination as Carbonate. In certain cases the precipitation of baryta by sulphuric acid is inadmissible; it is then thrown down in the form of carbonate, by carbonate of ammonia, mixed with a little caustic ammonia ; the precipi- tate is allowed to settle in a warm place, and is washed on the filter with water, rendered slightly alkaline by ammonia. It may be heated to redness, without losing carbonic acid. Its composition is One equivalent of BaO . . 76*5 . . 77'C6 One ditto of C0 2 .... 22'0 . . 22 "34 One ditto of BaO, C0 3 . . 98'5 100-00 Baryta is estimated as carbonate when it exists in combi- nation with an organic acid : the salt is carefully ignited in a platinum crucible, first covered, and afterwards, with free ac- cess of air ; the heat is continued till the residue is perfectly white ; it is then allowed to cool, moistened with carbonate of ammonia, and again gently ignited, PART TI. U 278 QUANTITATIVE ANALYSIS. Separation of Baryta from the Alkalies. The compound is dissolved in water or in hydrochloric acid, if necessary; the baryta is precipitated as sulphate, the alkalies being in this case obtained also in the state of sulphates by evaporating the filtered solution ; or as carbonate, in which case the alka- lies may be obtained as chlorides, which is the most convenient form if they have subsequently to be separated from each other. 157. STRONTIUM. Oxide of Strontium or Strontia is also weighed as sulphate and as carbonate ; it is precipitated in the form of both salts precisely as baryta ; but, as sulphate of strontia is not alto- gether insoluble in water, spirits of wine are added to the fluid, to dimmish its solubility : when this is inadmissible, it is better to precipitate the earth as carbonate, washing the salt on the filter with water containing ammonia and carbonate of ammonia. The composition of sulphate of strontia is One equivalent of SrO. . . 51-75. . 56-44 One ditto ofSO 3 . . . 40-00. . 43-56 One ditto ofSrO,S0 3 9175 100-00 The composition of carbonate of strontia is One equivalent of SrO . . 51-75 . . 70*20 One ditto of CO,. ; 22-00 . 29-80 One ditto ofSrO,C0 2 . 73-75 . . 100-00 Separation of Strontia from Baryta. The hydrochloric so- lution of the earths is mixed with excess of hydrofluosilicic acid, and allowed to remain at rest for some hours : the crys- talline precipitate of silico-fluoride of barium is collected on a weighed filter, washed and dried at 212 : the strontia in the filtrate is estimated either as sulphate or as carbonate. The composition of silico-fluoride of barium is One equivalent of Ba . . 68'5 . 49-50 = 54'84 BaO One ditto of Si . . 14-0 . 10'63 Three ditto of F . . 57'0 . 40-87 One ditto of BaF,SiF 2 . fslJKS 100-00* * According to Eose (Fogg. Ann., Nov. 6, 1850), silico-fluoride of barium is not entirely insoluble in water, and therefore perfectly accurate re- sults are only obtained when the precipitation is effected from a spirituous solution. A small quantity of alcohol need only be added to render the precipitation complete. CALCIUM, 279 Separation of Strontia from the Alkalies. This is effected in the same manner as the separation of baryta. 158. CALCIUM. Oxide of Calcium, or Lime, is estimated as sulphate and as carbonate. As sulphate of lime is soluble to a considerable ex- tent in water, it is necessary to add to the solution about to be precipitated by sulphuric acid, twice its volume of alcohol, and to wash the sulphate of lime on the filter with spirit : it is ignited previous to weighing. The composition of sulphate of lime is One equivalent of CaO . . 28 . . 41-17 One ditto ofSO 3 . . .40 . . 58-83 One ditto ofCaO,S0 3 . 68 . . 100 "00 Quantitative determination as Carbonate. If the lime salt be soluble in water, and if the solution be perfectly neutral, oxalate of ammonia is added as long as a precipitate is pro- duced ; the oxalate of lime is allowed to settle completely, which usually requires some hours, the vessel (a beaker, or a Phillips' s precipitating-jar) being covered, and placed in a warm situation ; it is then filtered, and the salt, having been thoroughly washed with hot water, is dried and exposed to a dull red-heat for twenty minutes ; the oxalate of lime is by this means converted into carbonate ; but, as it is possible that it may have lost carbonic acid during ignition, it is safer, be- fore weighing, to moisten it with a few drops of solution o carbonate of ammonia, to evaporate to dryness on the water- bath, and again expose it to a very gentle ignition. The composition of carbonate of lime is One equivalent of CaO ... 28 .. 56 One ditto of C0 2 ... 22 . . 44 One ditto ofCaO,C0 2 . . 50 . .100 If, however, the lime salt can only be held in solution by a free mineral acid, or if it will not bear the addition of excess of ammonia without a precipitation taking place, as is the case with phosphate of lime for example, the lime is best esti- mated as sulphate, because, oxalate of lime being soluble in mineral acids, it would be impossible to precipitate it from a solution containing free nitric or hydrochloric acid ; oxalate of lime is, however, insoluble, or nearly so in acetic acid, and like- u 2 280 QUANTITATIVE ANALYSIS. wise in oxalic acid. Very accurate results may therefore be obtained by adding to the hydrochloric solution of the lime salt sufficient ammonia to occasion a slight precipitate, then a drop or two of hydrochloric acid to redissolve this precipitate, then oxalate of ammonia in excess, and finally, acetate ofpotassa, a portion of which becomes decomposed by the free hydro- chloric acid, liberating a corresponding quantity of acetic or oxalic acid, in which, as before observed, oxalate of lime is almost entirely insoluble. Separation of Lime from Baryta, Strontia, and the Alkalies. (a.) From Baryta. This is best effected by adding to the acid solution of the two earths very diluted sulphuric acid, as long as precipitation occurs ; if the sulphuric acid be suffi- ciently diluted, no lime will be precipitated. (b.) from Strontia. This is effected with some difficulty ; the only good method is that recommended by Stromeyer, which is based on the complete solubility of nitrate of lime in absolute alcohol, and the insolubility of nitrate of strontia in the same menstruum. The two earths are converted into ni- trates, excess of nitric acid being carefully avoided, the solu- tion is evaporated to dryness in a flask that can be closed, and the residue digested for several hours with absolute alcohol, being frequently shaken ; the mixture is then filtered, and the undissolved residue washed with alcohol ; both earths are then estimated as sulphates. (c.) From ^Baryta and Strontia. When all three earths are together in a solution, the baryta is first precipitated by hydro- fluosilicic acid ; the lime and the strontia are obtained from the solution filtered from the silico-fluoride of barium in the form of sulphates, by the addition of sulphuric acid and evaporation to dryness ; the mixed sulphates are fused in a platinum crucible with three times their weight of mixed carbonates of potassa and soda ; the fused mass is extracted with water, and the earths are thus obtained in the state of carbonates. They are next converted into nitrates, and separated by alcohol, as above directed. (d.) From tlie Alkalies. The process for effecting this is very simple ; the lime is separated by oxalate of ammonia, with the usual precautions ; the filtered solution is evaporated to dryness, and ignited to expel the ammoniacal salts ; the re- sidue* is dissolved in water, and the alkalies determined ac- cording to the method already described. MAGNESIUM. 281 159. MAGNESIUM. Oxide of Magnesium, or Magnesia, is weighed as sulphate, or as pyrophosphate, and sometimes as pure magnesia. Quantitative estimation as Sulphate. The earth is determined in the form of this salt when no other fixed constituent is pre- sent ; the solution containing it, is mixed with excess of sul- phuric acid, evaporated to dryness, and ignited in a platinum crucible ; the residue is again treated with dilute sulphuric acid, evaporated to dryness, and again gently ignited ; the residue is pure anhydrous sulphate of magnesia, the composition of which is One equivalent of MgO . 20-16 . . 33'51 One ditto of S0 3 . . . 4Q-QQ . . 66'49 One ditto of Mg 0, S 3 . . 60'16 . . 100-00 Quantitative estimation as Pyrophosphate. Chloride of am- monium is added to the solution, then ammonia in excess ; should a precipitate hereupon occur, a fresh quantity of chloride of ammonium is added, by which the precipitate is redissolved ; solution of phosphate of soda is then dropped into the mixture as long as precipitation takes place. The whole is well agitated, and allowed to repose for several hours: the precipitate is collected on a filter, the amount of ash furnished by which is known, and washed with water containing about one-eighth of ammonia, in which the basic phosphate of magnesia and ammonia (2MgO,NH4O,P0 5 + 24aq.), is scarcely sensibly soluble ; the washed salt is dried, and carefully ignited, together with the filter; the latter requires considerable time for incineration, which is best effected by cutting in strips and burning it on the lid of the crucible. The composition of the ignited salt is Two equivalents of MgO . . 40-32 . . 36-22 One equivalent of P0 5 . . . 71*00 . . 6378 One ditto of 2 MgO, P0 5 . .111-32 . .10000 Separation of Magnesia from Baryta, Strontia, Lime, and the Alkalies : (a.) From Baryta and Strontia. The two latter earths are precipitated by carbonate of ammonia, a sufficient quantity of chloride of ammonium having previously been added to the solution to prevent the precipitation of the magnesia : the pve- QUANTITATIVE ANALYSIS. cipitated carbonates of baryta and strontia are dissolved in hy- drochloric acid, and the baryta separated from the strontia by hydrofluo-silicic acid ; the magnesia is determined in the filtrate as pyrophosphate ; if no strontia be present, the baryta may be precipitated as sulphate. (b.) From Lime. There are several methods of effecting this. 1st, by oxalate of ammonia : Chloride of ammonium is added to the solution of the two earths, then slight excess of ammonia ; the lime is then precipitated by oxalate of ammonia, with the proper precautions, and the magnesia in the filtrate determined as pyrophosphate. 2nd, by sulphate of lime : For this purpose the two earths must be in the state of sulphates ; the mixture having been ignited, is digested with a saturated solution of sulphate of lime, and the insoluble residue washed on a filter with the same salt ; the whole of the sulphate of magnesia is thus removed, the sulphate of lime, being quite insoluble in a saturated solution of sulphate of lime, remaining on the filter ; it is heated to redness and weighed, and from the difference in weight, before and after the operation, the amount of sulphate of magnesia is calculated. 3rd, by chlorate of potassa : The hydrochloric solution of the two earths is evaporated to dryness, and ignited in a platinum crucible, chlorate of potassa is then added in small quantities at a time, until the evolution of chlorine ceases. The mass on cooling is extracted with water, which dissolves the chloride of calcium, and leaves a residuum of pure magnesia, the chloride of magnesium having, under the influence of heat, aided by the chlorate of potassa, been com- pletely decomposed : the lime is determined in the aqueous solution by oxalate of ammonia. (c.) From the Alkalies : I. Berzelius's method. The bases are brought to the state of chlorides, to a concentrated solution of which, finely powdered and perfectly pure red oxide of mercury is added in excess ; mutual decomposition of the chloride of magnesium and of the oxide of mercury takes place, chloride of mercury being formed, which combines with the alkaline chlorides, forming a soluble double salt, and magnesia, which remains undissolved, on ex- tracting the evaporated and ignited mass with water. The solution, filtered from the magnesia, is evaporated to dryness, and heated, to expel the chloride of mercury: the alkaline chlorides alone remain, which are separated from each other as directed (page 259) ; the magnesia may be contaminated with MAGNESIA FROM THE ALKALIES. 283 undecomposed oxide of mercury, from which, however, it is readily freed by heat. This method is well adapted for the the analysis of mineral waters, soils, etc. II. Liebig's method. This also gives accurate results ; baryta- water is added to the hydrochloric solution to alkaline reaction, magnesia is thereby precipitated, baryta being a stronger base than magnesia, in consequence of which it deprives it of its electro-negative element ; to the solution, filtered from the pre- cipitated magnesia, carbonate of ammonia, mixed with caustic ammonia, is added, the excess of baryta is thereby removed, and is filtered off ; the alkalies are contained in the filtrate in the form of chlorides, and are separated from each other in the usual manner. in. Sonneschein? s method (Poggendorff 's ' Annalen/ 74, p. 313). The compounds to be separated are converted in the usual manner into chlorides, which are evaporated to dryness, and then heated to faint redness, when any ammoniacal salt con- tained in the solution, as well as a portion of the hydrochloric acid combined with the magnesium, escapes. When the dry mass is cool, water is poured over it, it is then boiled with carbonate of silver until the liquid has a strong alkaline reaction. The boiling is continued for about ten minutes, stirring the whole time, when the decomposition is complete. The solution is then filtered as hot as possible, and the precipitate washed with hot water. The filtered liquid now contains only the alkalies, and a trace of the silver salt, which is removed by hydrochloric acid, and the alkalies are then determined in the usual manner. The residue left upon the filter is digested with hydrochloric acid, and, after removing the chloride of silver, the magnesia is pre- cipitated in the usual manner with phosphate of soda and am- monia. The carbonate of silver is best prepared by carefully precipitating nitrate of silver with carbonate of ammonia ; after subsidence, the precipitate is easily freed from the ammo- niacal salt by the frequent addition of water, and decanta- tion. It is not necessary to filter it, as the moist precipitate is more easily decomposed. IT. Heintz's method (Poggendorff's c Annalen,' 73, p. 119). The solution containing the three bases is supersaturated with ammonia, and if it contain no chlorides a few drops of chloride of ammonium are added. Should a troubled appear- ance arise in the liquid, it is treated with solution of chloride df ammonium until this vanishes. The magnesia is then precipi- 284 QUANTITATIVE ANALYSIS. tated by phosphate of ammonia, the precipitate washed with ammoniacal water, dried, heated to redness, and weighed. The free ammonia in the filtrate having been partly removed by boiling, the phosphoric acid is precipitated at the boiling tem- perature by nitrate or acetate of lead. When an excess of the lead salt has been used, a solution of carbonate of ammonia, containing free ammonia, is added to the hot liquid, which is then allowed to stand for a few minutes ; it is subsequently filtered, and the amount of potash and soda contained in the filtrate is ascertained by the usual methods. It is essential that some chloride should be present in the solution in order that a compound of chloride and phosphate of lead may be produced, from which ammonia cannot extract any phosphoric acid. v. Watts' 's method. This consists in precipitating the mag- nesia by a known weight of carbonate of soda, using consider- able excess, then boiling and filtering, treating the filtrate with a slight excess of acid, evaporating to dryness, and igniting the residue to render it neutral, weighing the neutral salt thus ob- tained, and making the proper correction for the quantity of soda-salt introduced. To ensure accuracy, the solution, after the addition of the carbonate of soda, must be well boiled for at least half an hour, in order to decompose a difficultly soluble double carbonate of soda and magnesia, which is formed on the first addition of the alkaline carbonate. The carbonate of soda must likewise be added in considerable excess, otherwise the precipitation will not be complete. The precipitate of carbo- nate of magnesia must be washed with boiling water, and the washing not too long continued, the carbonate not being com- pletely insoluble. The washing should be discontinued as soon as the wash-water ceases to give a distinct alkaline reaction : when this takes place, the water begins to dissolve the carbo- nate of magnesia. (Quart. Journ. Chem. Soc.) YI. Ebelmen's method. To the solution of the mixed sul- phates, recently prepared carbonate of baryta is added, and then a stream of carbonic acid passed through the liquid till ifc contains a portion of the baryta in solution in the form of bi- carbonate, which period is known by filtering a few drops of the liquid, and adding one drop of exceedingly dilute sulphuric acid, if a cloudiness is produced, the liquid can contain no sul- phate in solution. It is then filtered, and the filtrate evaporated to dryness at a pretty high temperature, in order to reduce the ALUMINUM. 285 whole to the state of neutral carbonates ; the residue is treated with a small quantity of boiling water, which dissolves only the alkaline carbonates. (Ann. de Chim. et Phys. Nov. 1859.) vn. Beynoso's method. (* Comptes Eendus,' Ivi. No. 18.) From the liquid acidulated with nitric acid, the lime (if pre- sent) is precipitated by the addition of ammonia and oxalate of ammonia. To the filtered liquid, phosphate of ammonia or phosphoric acid alone is added, and the ammonio-magnesian phosphate collected. The filtrate is evaporated to dryness and calcined to get rid of the ammoniacal salts, and the potassa and soda remain united with phosphoric acid only. To ensure ac- curacy, the calcined residue should be again treated with nitric acid and again evaporated to dryness and ignited ; it is then treated in a flask with a large excess of tin and nitric acid ; the phosphoric acid is hereby eliminated ; it is filtered, and the fil- trate concentrated; the residuum composed of nitrates of potassa and soda is calcined till completely decomposed, and as soon as the capsule is cold, the caustic alkalies are weighed, or trans- formed into carbonates, after which they are converted into chlorides and separated by bichloride of platinum. GROUP III. Aluminum, Glucinum, Yttrium, Thorinum, Zirconium, Chromium. 160. ALUMINUM. Sesquioxide of Aluminum, or Alumina, is precipitated from its solutions by ammonia, or carbonate of ammonia, chloride of ammonium having been previously added and heat applied. The precipitate, which is very bulky, requires long-continued wash- ing with hot water ; its ignition must be carefully performed in order to avoid loss by spirting. It shrinks very much in dry- ing. Its composition is Two equivalents of Al . ; 27'4 . . 53'32 Three ditto of O .... 24-Q . . 46*68 One ditto of A1 2 3 . . . 51-4. .100-00 It is never weighed in any other form, than as pure alumina. The washing of precipitated alumina is greatly facilitated by allowing it to dry slightly on the filter before commencing washing. 286 QUANTITATIVE ANALYSIS. Separation of Alumina from the Alkaline Earths and Al- kalies. To the hydrochloric solution of the mixture, chloride of ammonium is added, then ammonia quite free from carbonic acid ; the alumina is precipitated, carrying with it a little mag- nesia ; it must be separated by nitration as rapidly as possible, the funnel being covered with a glass plate to prevent the ac- cess of carbonic acid, which would determine the precipitation of the earthy carbonates ; the precipitate on the filter is well washed with hot water, it is then, while still moist, dissolved in hydrochloric acid, and boiled with excess of pure potassa ; the small quantity of magnesia remains undissolved, and, hav- ing been separated from the alkaline ley by filtration, it is well washed, dissolved in a small quantity of hydrochloric acid, and added to the solution containing the alkaline earths and the rest of the magnesia. The potassa ley contains the whole of the alumina, which is precipitated by adding excess of hydro- chloric acid, and, finally, supersaturating with ammonia ; the alkaline earths and alkalies are separated from each other in the manner directed in the last section. The success of this process depends on the freedom of the ammonia from carbonic acid, and on the rapid filtering and careful washing of the precipi- tated alumina. If alumina has to be separated from baryta only, the latter earth may effectually be removed by sulphuric acid. Separation of Alumina from Lime only. The alumina is pre- cipitated by ammonia free from carbonic acid, with the precau- tions just prescribed ; and the lime in the filtrate is precipitated by oxalate of ammonia. Separation of Alumina from Magnesia alone. This may be effected by either of the following methods : If the quantity of magnesia be small, the mixture of the two earths may be dissolved in hydrochloric acid, excess being avoided, and the solution boiled with excess of caustic potassa, by which the alumina is dissolved, magnesia remaining behind. If the quan- tity of magnesia be considerable, chloride of ammonium is added to the hydrochloric solution of the two earths, and the alumina is precipitated by ammonia ; but, as it carries with it a small quantity of magnesia, it must be redissolved in hydro- chloric acid and treated with caustic potassa as above ; it is not safe to treat at once the hydrochloric solution of the earths with excess of caustic potassa when the quantity of magnesia is considerable, it being in this case very dif- cult to separate them by an alkaline ley. The best method GLUCINUM. 287 of separating alumina from magnesia is probably to preci- pitate the former by bicarbonate of potassa, and to estimate the magnesia in the filtrate as pyropnospliate of magnesia; the alumina precipitated in this manner carries with it a portion of potassa, which it is almost impossible to remove by washing ; it must therefore be redissolved in hydrochloric acid and reprecipitated by carbonate of ammonia. Separation of Alumina from Lime and Magnesia. (Rose). Instead of being careful to employ ammonia free from carbonic acid and avoiding the presence of this gas, the liquid in which the alumina has been precipitated by excess of ammonia is heated to ebullition. When the evolution of ammonia ceases, all the alumina is in the precipitate, and may be separated by filtration without requiring any special precaution, for the simple reason that in the presence of ammoniacal salts the car- bonate of lime is decomposed, the lime entering into solution ; a little chloride of ammonium may be added if there is a chance of there not being sufficient to favour this decomposition. When the lime is present in small quantities only, tartaric acid may be added, and the solution then supersaturated with am- monia. The lime is precipitated in the form of tartrate if there be only a little alumina present, otherwise it remains in solu- tion, but can be perfectly separated in the form of oxalate. When magnesia is present, the tartaric acid process may also be used, the lime being precipitated first by oxalic acid, and then the magnesia in the state of ammonio-magnesian phos- phate, but traces of alumina are carried down with the mag- nesia. 161. GLUCINUM. Oxide of Glucinum, or G-lucina, like alumina, is only weighed in its pure form, as precipitated from its solution by caustic ammonia. Its composition is One equivalent of Gl . . . 4 '66 . . 36 '9 One ditto of O 8'0 . . 63*1 One ditto of G1O . . . . 12 '66 . . 10OO Separation of Glucina from Alumina. Three methods have been proposed. The first depends on the solubility ofglucina in carbonate of ammonia, and the insolubility of alumina in that reagent. The solution containing the two earths is mixed in a flask with a very considerable excess of a concentrated so- 288 QUANTITATIVE ANALYSIS. lution of carbonate of ammonia, the flask is closed, and occa- sionally shaken; when it is observed that the precipitate ceases to diminish in bulk, the alumina is separated by filtration, and the filtrate evaporated to dryness, and ignited, to expel the am- inoniacal salts ; the residue, provided no other base or fixed acid be present, is pure glucina : or both the earths may be to- gether precipitated by ammonia, and the precipitate digested with carbonate of ammonia till the glucina is entirely dissolved. The second method is to dissolve both earths in a hot and concentrated solution of caustic potassa, to allow the ley to cool, and then to dilute it considerably with water, and again boil ; the glucina is in this manner precipitated, while the alu- mina remains in solution. The third method is that proposed by Berthier. The earths are dissolved in sulphuric acid, the solution concentrated, and sulphate of ammonia added, which causes the separation of the greater portion of the alumina in the form of an alum : to the decanted and diluted liquid, sulphite of ammonia is added, and it is boiled until sulphurous acid ceases to be liberated ; the alumina is entirely precipitated, and the glucina remains in so- lution, and may afterwards be precipitated by ammonia : or both earths may be together precipitated by ammonia, and, while moist, treated with sulphurous acid, which redissolves both, but on boiling the solution, the alumina is completely precipitated. The separation of glucina from lime, magnesia, and the alka- lies, is effected in the same manner as the separation of alu- mina. 162. TTTEIUM. Oxide of Yttrium, or Yttria, is weighed as the pure earth, in which state it is precipitated by caustic alkalies ; it must, how- ever, be observed, that when it is dissolved in nitric or sulphu- ric acid, potassa must be the precipitant employed ; in the lat- ter case it is, according to Wohler, almost impossible to procure it free from sulphate of potassa. Its composition is One equivalent of Y . . . . 32 . . 80'0 One ditto of 8 .... . 2OO One ditto of TO 40 . . lOO'O Separation of Yttria from Alumina and Glucina. This is easily accomplished by digesting the mixture of the three eartha in caustic potassa in which yttria is insoluble. THOKINUM. ZIRCONIUM. 289 Separation of Yttria from Magnesia and the Alkalies. From magnesia, yttria is separated by caustic ammonia, chloride of ammonium having been previously added to the solution ; from the alkalies it is separated precisely in the same manner as alumina. 163. THOKINTJM. Of Oxide of Thorimm, or ZTiorina, little is known : from alu- mina it is distinguished by its insolubility in caustic potassa, and as it is completely precipitated from its solutions by am- monia, even in the presence of chloride of ammonium, it may thus be separated from magnesia and lime. Tnorina forms, with sulphate of potassa a double salt, quite insoluble in sul- phate of potassa ; this salt may, therefore, be employed to sepa- rate it from most other substances, but it must be concen- trated and in excess ; the double salt, after being washed, is dissolved in boiling water, and the tJiorina precipitated by cau- stic potassa. Its composition is One equivalent of Th . . . 59'5 . . 88'15 One ditto of 8'0 . . 11-85 One ditto of ThO .... 67'5 . .10000 164. ZIBCONIUM. Oxide of Zirconium, or Zirconia, is precipitated from its solu- tion by ammonia and by caustic potassa ; the latter is the best precipitant, the former often throwing down subsalts instead of the pure earth. Sulphate of potassa added in crystals and in sufficient quantity to saturate the liquid, which must be first completely neutralized by potassa, throws down the whole of the earth in the form of a double salt ; it must be washed with water containing ammonia, and then boiled with caustic potassa, which leaves hydrate of zirconia in a pure state. Its compo- sition is One equivalent of Zr . 44'75 . . . 73'66 Two equivalents of . . 16'00 . . . 26'34 One equivalent of ZrO 2 6075 . . . 10000 The solubility of zirconia in bicarbonate of potassa affords a means of separating it from alumina, but an accurate method of separating ifc from yttria and glucina remains to be dis- covered. 290 QUANTITATIVE ANALYSIS. 165. CHROMIUM. Sesquioxide of Chromium is usually estimated quantitatively in its pure state ; the solution containing it, is heated to the boiling temperature, and, ammonia being added in slight excess, it is boiled for about half an hour, or until the solution is per- fectly colourless ; the precipitate is collected on a filter, washed with boiling water, dried, and ignited. Its composition is Two equivalents of Or . 52'6 . . . 68'67 Three ditto of O .; . . 24'0 . . . 31-33 One equivalent of Cr 2 3 76'6 . . . lOOOO The ignition of this oxide must be performed with care, as at a particular temperature it suddenly becomes incandescent with a sort of explosion, whereby a portion may be projected from the crucible. The crucible should be closed with its cover. When chromium exists in a solution in the form of chromic acid, it may be estimated as chromate of baryta or as chromate of lead, by adding respectively nitrate of baryta or nitrate of lead to the solution ; it may also be precipitated by subnitrate of mercury, the resulting chromate being decomposed by ignition into mercury, oxygen, and oxide of chromium ; from the weight of the latter the quantity of chromic acid may be calculated ; it is better, however, to reduce the chromic acid to oxide of chromium in the solution previous to precipitating it, which is done by concentrating the solution and boiling it with excess of hydrochloric acid, mixed with alcohol, till the liquid as- sumes a pure green colour. The reduced oxide of chromium is precipitated by ammonia, after the excess of alcohol has been expelled by a gentle heat. Estimation of Chromium in its compounds : I. VohVs method (Liebig's 'Annalen,' Sept. 1847.) This method is applicable to the determination of the metal, whether occurring in the form of oxide, or as chromic acid, but when the metal exists as oxide of chromium, it must in the first place be converted into chromic acid. The process depends on the con- version of oxalic into carbonic acid, by the oxygen furnished by the reduction of chromic acid into oxide of chromium ; thus 2(Cr0 3 )+3(C 2 3 ) = Cr 2 3 + 6Cp 2 For each equivalent of chromic acid, three equivalents of car- bonic acid are formed, or 66 parts by weight for 5O3 parts of chromic acid. To determine the amount of carbonic acid, the CHROMIUM. 291 author employs the alkalimetrical apparatus of Eresenius and Will (page 268). If merely the chromium has to be determined, any oxalate may be taken, but, if the alkalies are to be deter- mined in the residuary liquid, oxalate of ammonia or baryta is employed. The analysis is very simple : when the chromium exists in the compound in the form of acid, the salt is taken just as it is, and the process is precisely the same as in the analysis of manganese ores (to be described further on). If the salt is a chloro-chromate, before allowing the sulphuric acid to pass over, oxide of mercury must be mixed with the salt, to prevent the elimination of chlorine or hydrochloric acid : after the operation, the amount of chlorine can be determined from the chloride of mercury by nitrate of silver, and the quantity of chloro-chromic acid contained in the salt calculated from it. The determination is less simple when the salt contains oxide of chromium. In the first place, the oxide must be con- verted into chromic acid, and this is best effected in the follow- ing manner : The salt to be examined is dissolved in water, and caustic potassa added to it until the whole of the hydrated oxide of chromium has redissolved ; upon which, keeping the solution cold, chlorine is passed into it until the green colour is changed into a yellowish-red one : an excess of potassa is now added to this liquid, which is evaporated in the water- bath, and heated to faint redness in a platinum crucible. The whole of the chlorate of potassa is decomposed, chromate of potassa and chloride of potassium remaining ; these are dis- solved, transferred with oxide of mercury into the apparatus, and the operation conducted as with chromates. The amount of oxide of chromium can be easily calculated from the quantity of carbonic acid which has escaped ; 6 equivalents of carbonic acid are set free for each equivalent of oxide of chromium : Cr0 3 + 3 =2CrO 3 823 88 3 In analysing a salt in which both chromic acid and oxide of chromium occur, two determinations must be made. In the first place, the carbonic acid which the salt yields as it is, is deter- mined, and the chromic acid calculated from the amount ; upon which the liquid is treated as a salt of the oxide, the quantity of carbonic acid first obtained subtracted from that last obtained, and the amount of oxide of chromium calculated from the dif- ference. This method will, according to M. Vohl, render it possible 292 QUANTITATIVE ANALYSIS. for every one to submit to analysis those compounds of chro- mium which occur so frequently adulterated in commerce. II. Bunsen's method. When a chromate, e.g. bichromate of potassa, is boiled with excess of fuming hydrochloric acid, every 2 at. of chromic acid eliminate 3 at. of chlorine : KO,2Cr0 3 + 7HC1 = Cr 3 01 8 + 7HO + KC1 + C1 8 . The decomposition is rapid and complete, and on it Bunsen has founded a volumetric process for analysing chromates. The decomposition of the salt is effected in a small flask, having a gas delivery tube, which conducts the evolved chlorine into a solution of iodide of potassium, 3 equivalents of iodine are set free, which are estimated by means of a standard solution of sulphurous acid. (* Journal of the Chemical Society,' vol. viii. p. 227.) in. Schwartz's method. The whole of the chromium is con- verted into chromic acid, and when oxide of chromium is pre- sent it is oxidized by fusion with hydrate and chlorate of potash. The chromic acid is then reduced to sesquioxide of chromium by a known quantity of protoxide of iron ; the quantity which remains unoxidized after the operation being determined accord- ing to Marguerite's method (see IRON) by a standard solution of permanganate of potassa ; the difference between the protoxide of iron left, and that consumed, gives the amount of protoxide oxidized by the chromic acid, thus 6FeO + 2Cr0 3 =:3Fe 2 3 + Cr0 3 . From this formula the quantities of chromium, sesquioxide of chromium and chromic acid may be determined : TOGO iron corresponds to O3143 chromium,'to 4571 sesquioxide, and to 0'600 chromic acid. Analysis of Chrome Ores. Eight or ten grains of the ore reduced to an impalpable powder are put into a platinum crucible, and covered with ten or twelve times the weight of bi- sulphate of potassa ; the crucible is carefully heated for about fifteen minutes to about the temperature of the fusing-point of the bisulphate, then raised to low redness, at which tempera- ture it is kept for another quarter of an hour ; the mass enters into quiet fusion, and vapours of sulphuric acid are evolved freely. The temperature is gradually raised until the whole is in perfect fusion. To the fused mass about 50 grains of pure carbonate of soda are added, and the mixture is again fused at as low a temperature as possible, pure nitre being added by degrees in quantity equal to that of the carbonate of soda. CHROMIUM. 293 The crucible is kept at a low red-beat for about an hour, arid then raised to a bright red-heat, at which it is kept for a quarter of an hour longer. When cold the fused mass is dissolved in boiling water, filtered whilst boiling hot, and the residue on the filter washed with boiling water. If this residue should con- tain any undecomposed ore, it must be again fused as before with bisulphate of potash, carbonate of soda, and nitre. The filtrate contains the whole of the chromium as chrotnate of potash, together with small quantities of silicic acid, alumina, and some- times titanic and manganic acids. Excess of nitrate of ammonia is added, and the solution is evaporated nearly to dryness on the water-bath, until all the liberated ammonia has been ex- pelled, water is now added, and the precipitate, if any, is filtered off: the filtrate is made strongly acid with sulphurous acid, carefully heated to boiling, precipitated with slight excess of ammonia boiled for a few minutes and filtered. The thorough washing of the chromic oxide is not easy. Genth (Chein. News, No. 137) finds it succeed best in the following way : After the precipitate has settled, the clear liquid is passed through the filter, then boiling water is added to the precipitate, and after settling, the supernatant liquid is filtered, the preci- pitate is then thrown on the filter, and washed twice or three times with boiling water; it is then washed back again into the dish, and boiled with water, until the little lumps which clog together are completely broken up ; it is then filtered, and this operation is repeated until the wash-waters do not show the presence of any sulphates by chloride of barium. The precipi- tate is then dried and burned, after which it is again boiled with water to which a few drops of sulphurous acid have been added, then ammonia added, and the precipitate washed, dried, ignited, and weighed. By this mode of analysis, Dr. Grenth states that the chromic oxide is obtained quite pure, and that two analyses of the same sample will not, if the process be care- fully conducted, vary more than O25 per cent. Separation of Oxide of Chromium from Alumina, the Alkaline Earths, and the Alkalies : (a.) From Alumina. The usual method is to fuse the mixture of the two oxides with twice its weight of carbonate of soda, and twice and a half of its weight of nitre, the oxide of chromium be- comes converted into alkaline chromate, which is extracted with water, and the alumina which remains undissolved is freed from alkali by dissolving in hydrochloric acid, and precipitating by PART II. X 294 QUANTITATIVE ANALYSTS. ammonia. According to Dr. Schaffhaeutl, however, a portion of alumina in this process always dissolves along with the alkaline chromate ; he therefore recommends to convey the precipitate obtained by adding ammonia to the solution containing the two oxides 'into a hot concentrated solution of caustic potassa, and to boil the whole down till near solidification ; when quite cold, water is added, and the whole of the alumina dissolves without carrying with it a trace of oxide of chromium. (b.) From the Alkaline Earths. This is also effected by fusing the mixture with carbonate of soda and nitre, the earths are thus obtained in the form of carbonates, and are separated from each other according to the instructions given (page 280). (c.) From the Alkalies. The oxide of chromium is precipi- tated by ammonia, sal-ammoniac having previously been added to the solution. The alkalies are determined in the filtrate ac- cording to the directions given (pp. 259 et seq.). GROUP IV. Zinc, Nickel, Cobalt, Manganese, Iron. 166. ZINC. Oxide of Zinc is precipitated from its solutions, for the pur- pose of quantitative estimation, by a fixed alkaline carbonate, or by sulphide of ammonium ; in either case it is subsequently brought to the state of oxide, in order to be weighed. Precipitation as Basic Carbonate. Carbonate of potassa or soda is added in excess to the solution, which is then boiled for some time ; if, however, ammoniacal salts be present, a con- siderable excess of carbonate must be added, and the solution must be evaporated to dryness at the boiling temperature, in order to decompose the ammoniacal salts. The dry mass is well washed with hot water, and the residue, which is basic carbonate of zinc (3HO,ZnO + 2ZnO,C0 2 ), is strongly ig- nited, by which it becomes converted into oxide. Precipitation as Sulphide. The solution, if it be acid, is first supersaturated with ammonia, the alkali being added in suffi- cient quantity to redissolve the oxide of zinc, which first pre- cipitates ; sulphide of ammonium is then added till it no longer occasions a precipitate, and the sulphide of zinc, which is white, and very voluminous, is allowed completely to subside before ZINC. 295 filtration ; the sulphide is washed with water containing sul- phide of ammonium, and then digested with concentrated hy- drochloric acid until the solution ceases to smell of sulphuretted hydrogen ; the resulting chloride of zinc is then precipitated as basic carbonate, and subsequently converted into oxide in the manner and with the precautions just described : it is to be observed, that it must not be neglected to examine whether the filtrate from the carbonate of zinc is free from oxide ofzinc y which is done by adding to it a few drops of sulphide of ammo- nium ; the formation of a white bulky precipitate indicates the presence of dissolved oxide of zinc, which must be collected and treated as above. The composition of oxide of zinc is One equivalent of Zn . . 3275 . . 80-27 One ditto of O . . {\' ( . . 8 "00 . . 19'73 One ditto of ZnO . . . 40'75 . . lOO'OO Separation of Oxide of Zinc from Oxide of Chromium, and from the Oxides of the Third, Second, and First Groups : (a.) From Oxide of Chromium. The solution is mixed with tartaric acid, excess of potassa is then added, and the clear so- lution is precipitated with colourless sulphide of potassium ; the sulphide of zinc is treated as above, the oxide of chromium is contained in the filtrate, and is obtained by evaporating it to dryness, and fusing the ignited residue with nitre and carbo- nate of soda, the alkaline chromate thus obtained is treated as directed (p. 293) ; when both the oxides are combined with acids that form soluble salts with baryta, their separation can, accord- ing to Eresenius, be accomplished by digesting the acid solu- tion for several hours with excess of artificially prepared car- bonate of baryta : the whole of the oxide of chromium is removed, and is mixed with the excess of carbonate of baryta, from which it is separated by dissolving in hydrochloric acid, and adding excess of sulphuric acid; the oxide of zinc remains in solution, and is precipitated by carbonate of potassa. (b.) From Alumina. The solution containing the two oxides is supersaturated with ammonia; alumina is precipitated, and oxide of zinc remains in solution. If, to a solution of the mixed oxides in sulphuric acid, excess of cyanide of potassium be added, heat being avoided, alumina is precipitated, and oxide of zinc retained in solution. If, to a solution of the two oxides in sulphuric acid, excess of acetate of baryta be added, and then sulphuretted hydrogen passed into the solution, sulphide of 296 QUANTITATIVE ANALYSIS. zinc is precipitated, while alumina remains dissolved ; the sul- phide of zinc is separated from the sulphate of baryta by di- gesting the mixture in hydrochloric acid, the oxide of zinc is then precipitated from the solution as basic carbonate, and the alumina is precipitated by ammonia. (c.) From Magnesia. Chloride of ammonium is added to the solution, and then sufficient ammonia to retain both oxides in solution : the zinc is precipitated from the ammoniacal so- lution by sulphide of ammonium, and the magnesia in the fil- trate determined in the usual manner. Another plan which has been proposed is to precipitate both oxides by carbonate of potassa ; and then, having added a sufficient quantity of cyanide of potassium to dissolve the zinc, the whole is evaporated to dryness at a boiling temperature, a little more carbonate of potassa having first been added : on treating the dry mass with water, the magnesia remains undissolved, and the zinco -cyanide of potassium is held in solution. (d.) From the Alkalies and Alkaline Earths. The bases are all brought into the state of acetates by adding acetate of baryta to the sulphuric acid solution as above described : the zinc is precipitated by sulphide of ammonium. To separate oxide of zinc from lime, baryta, and strontia, it has been recommended to treat the mixed solution with carbonate of potassa, until it ac- quires an alkaline reaction, then to add excess of cyanide of potassium, and apply heat ; the earthy carbonates remain un- dissolved, while that of zinc is taken up. The solution is boiled with hydrochloric and nitric acids, until all hydrocyanic acid is expelled, and the oxide of zinc is then precipitated with carbonate of soda, those precautions being taken which are ne- cessary when a salt of ammonia is present. Volumetric Determination of the amount of Zinc in ores and metallurgical products. (Max Schaffuer.) The principle of the method is as follows : Oxide of zinc dissolves very readily in a mixture of ammonia and carbonate of ammonia; from this ammoniacal solution the zinc is com- pletely precipitated by sulphide of sodium. Now if exactly as much sulphide of sodium is added as is necessary for the pre- cipitation of the zinc, the amount of zinc may be calculated from that of the sulphide of sodium employed. It is impossi- ble, however, to observe the moment exactly, when the neces- sary quantity of sulphide of sodium has been added, for which reason it is necessary to call in the aid of another reaction. ZINC. 297 which shows the completion of the operation with certainty. This object is very simply and perfectly attained when a few drops of percTiloride of iron are added to the ammoniacal solu- tion, by which a voluminous red precipitate of hydrated oxide of iron is produced. If sulphide of sodium be then added, this red precipitate remains unchanged as long as the solution still contains zinc ; but as soon as all the zinc is thrown down as hydrated sulphide of zinc the red suspended precipitate of hydrated oxide of iron becomes black, sulphide of iron being formed, and this is the sign that the operation is completed. The valuation of the sulphide of sodium is effected in the fol- lowing way. About 0'2 grm. of chemically pure zinc are weighed and dissolved in hydrochloric acid, the ammoniacal fluid is then added, and afterwards the same number of drops of per- chloride of iron which will be used in the analysis ; sulphide of sodium is then poured in from the burette until the red preci- pitate appears black, the number of cubic centimetres employed is noted, the quantity serving for the blackening of the oxide of iron is deducted, and from this is calculated how much zinc is represented by 1 cubic centimetre of sulphide of sodium. The dilution of the sulphide of sodium is about right when 17 or 18 cub. cent, represent about 0*2 grm. of zinc. To deter- mine the quantity of sulphide of sodium required to convert oxide of iron into sulphide, the following is the process adopted. Into a glass flask is put about as much water containing am- monia as there is fluid obtained in the analysis ; to this is added the same number of drops of perchloride of iron as is employed in the analysis, and after this, the solution of sulphide of sodium, until the red colour of the precipitate has become converted into black. When well corked, the solution of sul- phide of sodium remains for some time without alteration ; it is however better to test it again when it has been standing for some days, especially as the new testing does not occupy more than 8 or 10 minutes, 1 gramme of substance is usually quite sufficient for investigation. With very poor substances, it is better to take 2 grammes ; with very rich substances half a gramme is sufficient. In some rare cases nickel and cobalt are present in zinc ores, in which case caustic potassa is employed as a solvent instead of ammonia. This method of assaying zinc ores is not given by the author as rigidly accurate, but he asserts that it is sufficiently so for practical purposes, provided care be taken that the ammoniacal 298 QUANTITATIVE ANALYSIS. solution of the zinc is not too concentrated, in fact, if the so- lution be diluted, and that of the perchloride of iron be very concentrated, the separation is very complete and distinct. 167. NICKEL. Oxide of nickel is precipitated from its solutions either as hydrated oxide, by caustic potassa, or as sulphide, by sulphide of ammonium ; in either case it is converted into anhydrous pro- toxide to be weighed. Precipitation as Hydrated Protoxide. Solution of pure caustic potassa is added, and the solution heated to boiling ; the volu- minous apple-green precipitate which is formed requires pro- tracted washing with hot water, after which it is dried and ignited ; ammonia cal salts do not interfere with the precipita- tion of oxide of nickel by caustic potassa, but with carbonate of potassa it is not so complete. Precipitation Ity Sulphide of Ammonium. The complete preci- pitation of nickel in the state of sulphide, by sulphide of am- monium is not easy, in consequence of the partial solubility of sulphide of nickel in that reagent. To ensure success the sul- phide of ammonium must be perfectly saturated and colourless, and not added in great excess, the vessel must be covered, and placed in a warm situation ; the fluid above the precipitate should be free from colour. The sulphide of nickel is washed on the filter with water to which a few drops of sulphide of am- monium have been added ; it is then dried, transferred to a beaker, and digested at a gentle heat until completely dissolved, with concentrated aqua-regia ; the filter is burnt, and the ashes added to the solution, this is better than digesting the filter and precipitate together, since the presence of organic matter pre- vents the complete precipitation of oxide of nickel by caustic alkali. The solution, filtered from the precipitated sulphur, is precipitated by potassa, washed, dried, and ignited. The com- position of anhydrous protoxide of nickel is One equivalent of Ni . . 29*5 . . 78'37 One ditto of O ..... 8 . . 21'63 One ditto of Ni;O .... . . 100-00 Separation of Oxide of Nickel from Oxides of Zinc and Chro- mium, and from the Oxides of the Third and Second Groups : (a.) From Oxide of Zinc. Several methods have been pro- posed. NICKEL. 299 a. Berzelius's method. The greater part of the oxide of zinc is first extracted by caustic potassa; the residue, after being well washed and heated, is mixed with pure pulverized sugar, and carefully carbonized in a porcelain crucible. The crucible is then surrounded with lime in a larger Hessian crucible, and heated for an hour in a blast-furnace as strongly as possible. The oxides are reduced, and the zinc is driven off in vapour ; the nickel is dissolved in nitric acid, evaporated to dryness in a platinum crucible, and ignited. The loss of weight gives the quantity of oxide of zinc. The principal point to be attended to is to extract all the potassa from the mixed oxides : the sugar must be perfectly pure, with which view it should be crystallized from an alcoholic solution. ft. Ullgrens method, The oxides are precipitated by carbo- nate of soda, the whole evaporated, and the residue gently heated, so that they remain perfectly insoluble when the mass is treated with water. The oxides are washed and dried, and then being placed in a tube with a bulb, are reduced by hydro- gen at a low red-heat; the tube is allowed to cool, while a continuous current of hydrogen is passed through it; it is then closed at one end by fusion, filled with a concentrated solution of carbonate of ammonia, corked up, and placed in a warm situation for twenty-four hours ; the oxide of zinc, which is not reduced, is dissolved in the carbonate of ammonia ; the oxide of nickel is reduced, and the metal, after being well washed with carbonate of ammonia, is dried and weighed ; the oxide of zinc is obtained from the ammoniacal solution by eva- poration. The oxides must be finely pulverized before they are exposed to the action of hydrogen gas. y. Rose's method. The mixed oxides, after having been ignited, are placed in the bulb of a reduction tube, which com- municates on the one side with an apparatus, from which dry hydrochloric acid gas is liberated, and, on the other, with a flask containing a very dilute solution of ammonia. As soon as the air is expelled, the bulb is gradually heated to redness, the oxides are converted into chlorides, the chloride of nickel remains in the bulb, and the volatile chloride of zinc passes into the solution of ammonia, in which it dissolves. The oxides are then determined in the usual manner. This method, though tedious and somewhat complicated, yields very accurate results. <$. Smith's method. The oxides are converted into acetates ; and, excess of acetic acid being added, sulphuretted hydrogen 300 QUANTITATIVE ANALYSIS. is passed through the solution, by which the whole of the zinc is precipitated as sulphide, while the oxide of nickel remains in solution : this method, which is well adapted for the analysis of German silver, the constituents of which are copper, zinc, and nickel, has been found by M. Louyet to yield very accurate results. . Brunner's method. The solution of the two metals in hydrochloric or nitric acid is diluted with a large quantity of water, and almost completely neutralized with carbonate of soda. To effect this a slight excess of the latter is added, and the precipitate which ensues is redissolved in a few drops of acid, so as to introduce a very slight excess of the latter. A current of sulphuretted hydrogen is then made to pass through the liquid, and to complete the precipitation of the zinc, a few drops of a diluted solution of acetate of soda are added. Care must be taken not to introduce an excess of acetate, and not to heat the liquid. A fresh current of sulphuretted hydrogen is now allowed to pass, until the precipitate does not appear to increase. The vessel is then left to itself for ten or twelve hours, at the ordinary temperature, after which the precipi- tate is collected on a filter, and washed. To make sure that all the zinc has been thrown down, a drop of acetate of soda may be added to a little of the filtered liquid, which should remain clear, when a little sulphuretted hydrogen is intro- duced. The nickel is thrown down from the filtrate by po- tassa. If the original solution contain iron, it is essential to separate it first of all, by carbonate of baryta. Field prefers hypochlorite of soda as a precipitant for nickel ; the metal is thrown down as peroxide, which, after boiling the solution in which it is suspended, separates as a rather dense precipitate, which can be easily washed, whilst the bulky gela- tinous protoxide thrown down by potassa is washed with diffi- culty, the alkali adhering with remarkable pertinacity to the hydrated oxide. The precipitation should be effected in a beaker. The peroxide is heated to whiteness, and the nickel weighed as protoxide. (b.) Prom Oxide of Chromium. The mixed oxides are fused with nitre and carbonate of soda ; the oxide of chromium is thus converted into chromic acid, and is dissolved on boiling with water as alkaline ehromate. (c.) From Alumina. Excess of cyanide of potassium is added, heat being avoided ; alumina is precipitated, and oxide NICKEL. 301 of nickel retained in solution : according to Berthier, on add- ing sulphite of ammonia to a solution of the two oxides, the alumina only is precipitated. (d.) From Magnesia. This is effected in the same manner as the separation of oxide of zinc from the same earth. (e.) From Baryta., Strontia, and Lime. Cyanide of potas- sium is added in excess, and then carbonate of potassa, the whole is boiled, and the insoluble carbonates separated by fil- tration from the nickel cyanide. The solution is boiled with hydrochloric acid until all the hydrocyanic acid is expelled ; potassa in excess is then added, and the solution boiled till all the ammonia is liberated j the oxide of nickel is then precipi- tated by potassa. Lewis Thompson thinks that absolutely pure nickel has not hitherto been obtained in quantity. He prepared it by re- ducing granulated oxide of nickel by hydrogen gas in a red-hot porcelain tube. When fused, the button obtained was white and silvery-looking; its specific gravity was 8'575, and it was almost as soft as copper ; it could be rolled out nearly to the thinness of tinfoil ; its malleability was therefore very great. Thompson suggests, as a great improvement in the metallurgy of this metal, to reduce the common ore (the arsenio-sulphuret) after washing, by mixing it with half its weight of chalk, and strongly heating in a cubilo-furnace in full blast ; the use of the large quantity of arsenic on which the present practice depends is thus avoided, and there is no appreciable loss of metal. If we have in solution a mixture of sulphates of nickel, co- balt, zinc, manganese, iron, and copper, almost every particle of nickel and cobalt may, according to Thompson, be separated as a green crystalline powder by adding to the solution in a warm state as much sulphate of ammonia as it will dissolve ; this separation depends on the fact that the sulphates of nickel and cobalt form with sulphate of ammonia triple alums, which are absolutely insoluble in cold saturated solutions of sulphate of ammonia. Analysis of KupfernicJcel (Ebelmen). The purified mineral is treated with aqua-regia. The sulphuric acid is precipitated by chloride of barium, and the excess of barium being removed by sulphuric acid, the arsenic acid is converted into arsenious acid by ebullition with sulphurous acid, and then precipitated by sulphuretted hydrogen. The sulphide of arsenic obtained is after drying and weighing, decomposed by aqua-regia, to obtain 302 QUANTITATIVE ANALYSIS the sulphur. The liquor freed from sulphide of arsenic is con- centrated with the addition of nitric acid, and precipitated by excess of ammonia ; an abundant precipitate of peroxide of iron is formed, which generally retains a little nickel. It is re- dissolved on the filter by hydrochloric acid, and the liquor treated cold with carbonate of baryta. The peroxide of iron alone is precipitated ; the carbonate of baryta with which it was mixed is readily separated. The liquor containing the nickel is treated with sulphuric acid, and after filtration is added to the ammoniacal solution of the rest of the nickel : this is precipitated by excess of potassa, and, after drying and calcining, is weighed ; its quantity indicates that of the metallic nickel. The ammoniacal liquor, afterwards treated with hydro- sulphate of ammonia, sometimes yields a black precipitate, which, collected, calcined, and weighed, gives with borax the reaction of cobalt. (' Annales des Mines,' tome xi. p. 56.) 168. COBALT. This metal is precipitated from its solutions either by caustic potassa or by sulphide of ammonium. It is weighed either as oxide or in the metallic state; when great or even moderate ac- curacy is required, it must be estimated in the latter form, as the ignited hydrated protoxide is of indefinite composition, varying according to the degree of heat employed. Precipitation as Hydrated Protoxide. The solution is boiled with excess of caustic potassa ; the precipitate, which at first is bluish, becomes, after long boiling, of a dirty rose-red colour; if ammoniacal salts are present, they must be decomposed, and the ammonia discharged by long-continued boiling, with great excess of caustic alkali ; the precipitated hydrate requires long- continued washing with hot water ; on being dried and ignited it turns black : in this state it is weighed ; if the operator has determined on reducing it to reguline cobalt, he proceeds in the following manner : Reduction of Oxide of Cobalt to Heguline Cobalt. The small bulb-tube of hard glass attached to the U-shaped chloride of calcium tube b, Fig. 71, having been carefully weighed, a certain known quantity of the previously weighed oxide is introduced, and the weight of the bulb-tube again accurately noted ; a stream of hydrogen gas, dried by passing through the U-shaped chloride of calcium tube b, is then caused to flow through the apparatus from the flask a, and a flame applied to the bulb containing the COBALT. 303 oxide ; the heat must at first be gentle, but it must gradually be raised to full redness, this high temperature being necessary to prevent the reduced metal from acquiring a pyrophoric pro- perty, which would cause it to inflame on coming in contact with atmospheric air; the hydrogen reduces the metallic oxide, giving rise to a proportional amount of water, which escapes Fig. ft. in the form of steam. When it is seen that water no longer continues to form, the flame is removed from the bulb, and it is allowed to cool, while a stream of hydrogen still continues to pass over it ; the bulb-tube is then again weighed, and from the amount of reduced metal obtained the quantity contained in the whole of the protoxide precipitated, is calculated. If a sufficiently high temperature has not been employed, the re- duced cobalt oxidizes at common temperatures, but if a full red-heat has been employed it absorbs oxygen very slowly. Precipitation as Sulphide. Sal-ammoniac is added to the so- lution, then slight excess of ammonia, and finally sulphide of ammonium as long as precipitation takes place ; the resulting sulphide of cobalt is washed with water containing a little sul- phide of ammonium, then digested with aqua-regia, and preci- pitated by potassa, precisely as sulphide of nickel. The hy- drated oxide, after being washed, dried, and ignited, is reduced by hydrogen in the manner just described. Itose recommends that, in the presence of ammoniacal salts, cobalt should always be precipitated by sulphide of ammonium. Separation of Oxide of Cobalt from Oxides of Nickel, Zinc, 304 QUANTITATIVE ANALYSIS. and Chromium, and from the oxides of the third and second groups : (a.) From Nickel. Several methods have been proposed by different chemists, but until recently Phillips's process was the one generally preferred, which is as follows : a. Method of Phillips. Both oxides are dissolved in an acid, and the solution supersaturated with ammonia, having previ- ously added a sufficient quantity of sal-ammoniac to prevent any precipitation from taking place ; the solution, which has a sky-blue colour, is largely diluted with water, which should have been previously well boiled, to free it from atmospheric air : caustic potassa is added to the hot solution, and the vessel is closed with a cork ; oxide of nickel is precipitated, and oxide of cobalt remains in solution ; when the former has completely settled, the supernatant liquid, which should have a rose-red colour, is poured through a filter, and the oxide of nickel washed with hot water, ignited, and weighed ; the oxide of cobalt in the filtrate is precipitated by sulphide of ammonium. The rea- son why it is necessary to dilute the solution of the two oxides with water, free from atmospheric air, is, that oxide of cobalt in an ammoniacal solution is converted into peroxide of cobalt, which, precipitating as a black powder, would contaminate the oxide of nickel. The more dilute the solution is, the less easily does the oxide of cobalt become oxidized. When a large quan- tity of ammoniacal salt is present, the quantity of caustic potassa required to precipitate the oxide of nickel is very con- siderable. According to Presenius the separation by this me- thod is not complete, the cobalt invariably containing traces of nickel, and the precipitated nickel often, traces of cobalt. ft. Liebig's method. This is founded on the following con- siderations : When any salt of cobalt is warmed with cya- nide of potassium and an excess of hydrocyanic acid, it is con- verted into the percyanide of cobalt and potassium, or cobalticy- anide of potassium (Co 2 Cy 6 , K 3 ), the aqueous solution of w r hich does not undergo any decomposition by boiling with either of the mineral acids. On the other hand, the precipitate produced by cyanide of potassium in solutions of salts of nickel is redis- solved by cyanide of potassium, but the solution is decomposed by dilute sulphuric acid, cyanide of nickel being precipitated. When, therefore, a mixture of a cobalt and a nickel salt con- taining free acid is treated with excess of cyanide of potassium, and slightly warmed, we obtain in solution the double cyanide COBALT. 305 of nickel and potassium, and cobalticyanide of potassium, (NiCy + KCy) + (Co 2 Cy 6 K 3 ,)and, on adding dilute sulphuric acid in the cold, three cases present themselves. i. If tho cobalt and nickel exist in the solution in the pro- portion by weight of two cobalt to three nickel, we have in the solution, 3 (NiCy, KCy) -f- (Co 2 Cy 6 K 3 ), and the three equiva- lents of nickel replacing the three equivalents of potassium in the cobalticyanide of potassium, produce cobalticyanide of nickel (Co 2 Cy 6 Ni 3 ), which is precipitated of a bluish-white colour, leaving in the solution no trace of either cobalt or nickel. ii. If the solution contain less nickel than corresponds to the above proportions, a certain quantity of cobalticyanide of potassium remains in solution, while cobalticyanide o'f nickel is still precipitated. in. If the solution contain more nickel than corresponds to the above proportions, cobalticyanide of nickel is still precipi- tated, together with the excess of cyanide of nickel, which, by long boiling with hydrochloric acid, is completely converted into chloride of nickel, which remains in solution. Now cobalticyanide of nickel, though insoluble in hydro- chloric acid, is decomposed by boiling with caustic potassa into protoxide of nickel and cobalticyanide of potassium, thus, Co 3 Cy 6 Ni 3 + 3KO = Co 3 Cy 6 K 3 + 3NiO, and chloride of nickel is also decomposed by caustic potassa into protoxide of nickel and chloride of potassium. Hence, the following method of analysing mixtures of cobalt and nickel, which is applicable to all proportions. Hydrochloric acid is added to the solution of the metals, and then cyanide of potassium in such excess that the precipitate at first formed is redissolved ; the whole is boiled, adding from time to time hydrochloric acid, until hydrocyanic acid ceases to be evolved. Caustic potassa is then added in considerable excess, and the boiling continued until the hydrated protoxide of nickel is completely precipitated ; it is then filtered, the filtrate con- tains the whole of the cobalt in the form of cobalticyanide of potassium ; it is evaporated to dryness with excess of nitric acid, the residue fused, and treated with hot water : peroxide of cobalt remains, which is dissolved in hydrochloric acid, and the solution treated as already directed. "When a mixture of the two oxides is dissolved in hydrocy- anic acid and potassa, and the solution kept boiling in a flask, 306 QUANTITATIVE ANALYSIS. the cobalt becomes converted into cobalticyanide of potassium, and the nickel into protocyanide of nickel and potassium ; all the cyanogen is extracted from the latter by oxide of mercury, by which the nickel is precipitated in the form of oxide ; oxide of mercury produces no change in the cobalt compound. In- stead of adding oxide of mercury to the cold solution of the mixed cyanides, this may be supersaturated with chlorine, and the resulting protocyanide of nickel again dissolved by caustic potassa. The chlorine has no action on the cobalt compound, whilst the nickel compound is decomposed, and all the nickel is at last precipitated as black peroxide. The operation must not be performed with heat, as otherwise Co 2 3 is precipitated with the nickel, and care must be taken that during the intro- duction of the chlorine the fluid is kept strongly alkaline. In analysing ores of nickel, which contain small quantities only of cobalt, considerable excess of hydrochloric acid must be taken to precipitate the cyanides dissolved in cyanide of po- tassium, and the mixture must be continued in ebullition for a full hour. y. Rose's method. This is founded on the greater tendency in the protoxide of cobalt than in the protoxide of nickel to pass to a higher degree of oxidation. Both metals are dissolved in hydrochloric acid ; the solution must contain a sufficient ex- cess of free acid : it is then diluted with much water ; if 20 or 30 grains of the oxides are operated on, about 2 pints of water are added to the solution. As cobalt possesses a much higher colouring power than nickel, not only in fluxes but also in so- lutions, the diluted solution is of a rose colour, even when the quantity of nickel present greatly exceeds that of the cobalt. A current of chlorine is then passed through the solution for several hours ;* the fluid must be thoroughly saturated with it, and the upper part of the flask above the liquid must remain filled with the gas after the current has ceased. Carbonate of baryta in excess is then added, and the whole allowed to stand for 12 or 18 hours, and frequently agitated ; the precipitated peroxide of cobalt, and the excess of carbonate of baryta, are well washed with cold water, and dissolved in hot hydrochloric acid ; after the separation of the baryta by sulphuric acid, the cobalt is precipitated by hydrate of potassa, and, after being * Mr. Henry recommends a solution of bromine, which is to be added till the solution smells strongly of it ; he found this to answer equally well with ohlorine, and the process is rendered less tedious and unpleasant. COBALT. 307 washed and dried, is reduced by hydrogen gas, in the manner shown in Fig. 71. The fluid filtered from the oxide of cobalt is of a pure green colour ; it is free from any trace of cobalt. After the removal of the baryta by means of sulphuric acid, the oxide of nickel is precipitated by caustic potassa. To ensure accurate results, it is indispensably necessary to wait a con- siderable time, at least 12, or even better 18 hours after the ad- dition of the carbonate of baryta, as the oxide of cobalt is precipitated very slowly. 8. Another method. The two oxides are covered with hydro- cyanic acid, and then potassa added till a portion remains undis- solved. The solution is kept boiling for a quarter of an hour, moist hydrated oxide of mercury is then added till a portion remains undi&solved ; a green precipitate occurs containing all the nickel, with the excess of oxide of mercury. By ignition, pure oxide of nickel remains. Acetic acid is added to the fil- trate to aid reaction ; it is then precipitated with blue vitriol. The blue precipitate contains all the cobalt; this is dried, ignited, redissolved in hydrochloric acid, the copper, precipitated by sulphuretted hydrogen, and then from the filtrate, the cobalt by potassa. The method depends upon the fact, that nickelo- cyanide of potassium is decomposed by oxide of mercury, while cobalto-cyamde of potassium experiences no change. (b.) From Oxide of Zinc. The above method may, according to Eose, be employed for the separation of these two oxides ; and also for that of other oxides from oxide of cobalt, which are strongly basic, and which are not converted into superoxides : oxide of zinc may likewise be separated from oxide of cobalt by Liebig's process with cyanide of potassium, cobalticyanide of zinc is gradually dissolved in boiling hydrochloric acid, and a clear solution is obtained : on the addition of caustic potassa and boiling, both the cobalt and the zinc are retained in solu- tion, the former as cobalticyanide of potassium, and the latter as oxide, and from the solution, zinc is precipitated, by sulphu- retted hydrogen (Brunuer) : the methods of Berzelius, Ullgren, and for the separation of oxide of nickel from oxide of zinc may likewise be employed for the separation of oxide of cobalt. (c.) From Oxide of Chromium. This is effected in the same manner as the separation of oxide of nickel from oxide of chromium. (d.) From Alumina. Cyanide of potassium is added, heat being avoided ; the cobalt is dissolved as cobalticyanide of 308 QUANTITATIVE ANALYSIS. potassium, and the alumina precipitated ; according to Ber- thier, the separation may likewise be effected by sulphite of ammonia. (e.) From the Alkaline Earths. This is effected in the same manner as the separation of the oxide of nickel. Analysis of Cobalt Ores, containing Cobalt, Nickel, and Iron. (Liebig.) Add to the hot acid solution ammonia till slightly alkaline, then succinate or benzoate of ammonia to throw down the iron. Precipitate the filtrate by potassa, and filter. Wash the oxides of cobalt and nickel repeatedly ; dissolve them in pure cyanide of potassium, then boil with an excess of fresh pre- cipitated oxide of mercury. The whole of the nickel deposits with the excess of oxide of mercury. Filter, saturate with acetic acid : boil ; then add sulphate of copper, and reboil. The precipitate has the following formula: Cu 3 Co 2 Cy 6 . Collect it, and ignite, to destroy the cyanogen ; dissolve in aqua-regia, and separate the copper by sulphuretted hydrogen ; this precipitate must be redissolved, and the oxide of copper precipitated by potassa. Erom the amount of copper that of cobalt is calculated. Three equivalents of copper are equal to two of cobalt. Extraction of Cobalt and Nickel from the Ore. The following process is followed in a manufactory at Birmingham, the ore employed consisting principally of metallic sulpho-arseniurets, and containing generally about 6 per cent, of nickel and 3 per cent, of cobalt. The ore is mixed with a small quantity of carbonate of lime and fluor-spar, and the whole is heated to a white-red heat in a reverberatory furnace : the mass fuses at this high temperature, and a slag is obtained floating on the surface of a fluid mass, of metallic appearance ; the latter is let out of the furnace by a particular aperture, and watered in order that it may be broken into fragments with greater faci- lity. It has been ascertained from experience that when the slag is of a dull colour, iron is present, but if its surface be black and brilliant, it is free from that metal. The metallic mass is reduced to a very fine powder, which is calcined at a bright-red heat in a furnace, the temperature being graduated so as to avoid fusion, and constantly raked : a considerable quantity of arsenious acid is driven off. The air has free ac- cess to the mass, which becomes oxidized and diminished in weight. The calcination, which lasts for about twelve hours, is continued until no more white fumes are given off, and the COBALT. residue is treated with hydrochloric acid, which dissolves nearly the whole of it : the liquid is diluted with water, and milk of lime and chloride of lime added to precipitate the iron and the arsenic; the precipitate, after being well washed, is thrown away. A current of washed sulphuretted hydrogen, generated from sulphide of iron and dilute sulphuric acid, is passed into the liquor until it is saturated : it is discontinued, when some ammonia, added to a sample of the filtered liquor, gives a black precipitate ; if there was not an excess of sulphuretted hydro- gen, the precipitate produced by ammonia would be green : the precipitate is washed and then thrown away, a current of sulphuretted hydrogen being passed into the wash waters. The cobalt is then thrown down with a solution of hypochlo- rite of lime. The precipitate, washed, dried, and then heated to redness, is considered to be oxide of cobalt, and part is sent in this state into the market. Another portion is heated to a white-red ; by this treatment the oxide loses in weight, but increases in density ; it is sold as protoxide of cobalt. The liquid from which the cobalt has been precipitated is treated with milk of lime, which precipitates the nickel in the state of hydrate : this precipitate is washed, dried, and heated to red- ness ; it is then mixed with charcoal, and, by means of a strong heat, reduced to the state of a spongy nickel, which is em- ployed in the manufacture of German silver. The oxide of cobalt thus prepared is remarkably pure. 169. IEON. This metal is weighed as sesquioxide ; if it exist in the solu- tion in the form of a protosalt, it must be peroxidized by heating with nitric acid as long as fumes of nitrous vapour are discharged, or by boiling with chlorate of potassa and hydro- chloric acid : ammonia is the precipitant employed, as the oxide precipitated by potassa or soda is always contaminated by a certain quantity of the alkali, from which it is almost impossible to free it by washing : the hydrated sesquioxide shrinks greatly on drying, and, by ignition, loses all its water. Iron is sometimes precipitated as sulphide by sulphide of am- monium : the solution must not contain free acid : it must be well washed on the filter with water, to which a few drops of sulphide of ammonium have been added, the funnel being pro- tected from the air with a glass plate to prevent a portion of the sulphide from becoming oxidized into sulphate, which PART II. T 310 QUANTITATIVE ANALYSIS. would dissolve and be carried through the filter : the washed sulphide is, together with the filter, digested with dilute hydro- chloric acid, filtered, and the filtrate peroxidized by nitric acid and precipitated by ammonia. When iron has to be separated from other bases, it is sometimes precipitated by succinate of ammonia : the solution must be very exactly neutralized by ammonia, the alkali being added in drops in a very diluted state until the small quantity of sesquioxide of iron which it precipitates is not redissolved by applying a gentle heat, the supernatant liquid possessing a red colour ; the neutral succi- nate of ammonia is then added, upon which succinate of per- oxide of iron of a brown colour is precipitated ; it is filtered when quite cold, and washed first with cold water, and, finally, with a warm solution of ammonia, in order to remove a portion of the succinic acid : it is then dried, and ignited in a current of air, in order thoroughly to peroxidize the iron. The compo- sition of sesquioxide of iron is Two equivalents of Fe . . 56 . . 70 Three ditto of O . . 24 . . 30 One ditto of Fe 2 O 3 80 100 Separation of Peroxide of Iron from Protoxide of Iron. This is attended with difficulty, and can only be accomplished when the compound is soluble in acids. When no other base but iron is present, as, for example, in native magnetic iron ore, Rose directs that a weighed -quantity of the substance should be dissolved in hydrochloric acid, and having boiled with nitric acid to peroxidize the whole of the iron, it is to be precipitated by ammonia ; the increase of weight is owing to the acquisition of oxygen, which has combined with the pro- toxide of the compound, and is half as much in quantity as the oxygen previously existing in the protoxide, for the protoxide of iron, on being fully converted into peroxide, acquires one- half more oxygen than it already possessed. Thus then, finding first the quantity of oxygen gained by the substance operated on, we find next the quantity of oxy- gen belonging to the protoxide existing in the compound, and from this it is easy to calculate the quantity of the protoxide. When this is found, the quantity of peroxide contained in the substance is learned from the difference in weight between IRON. 311 the quantity of the compound submitted to analysis and the quantity of protoxide made out by calculation. It is easy, however, to see that as the proportion of protoxide of iron is generally small in comparison with that of peroxide, the greatest accuracy in experimenting is necessary, in order to arrive at correct results, a very trifling error becoming a very considerable one in the subsequent calculation of the quantity of protoxide. Another method, given by Rose, for determining the quan- tity of oxygen in a compound consisting merely of protoxide and peroxide of iron, is by converting the oxides into metallic iron by igniting them in a current of dry hydrogen gas, and determining not only the quantity of iron revived, but also the weight of the water formed. To determine experimentally the quantity of peroxide of iron in a soluble compound of peroxide and protoxide, a weighed quantity of the pulverized mixture is introduced into a flask, the whole of the atmospheric air from which is then expelled by a current of carbonic acid gas ; hydrochloric acid sufficient to dissolve the compound is then added, and the flask quickly and securely closed. Solution being effected, recently pre- pared perfectly clear sulphuretted hydrogen water is added in excess, and the flask again closed, and allowed to remain at rest for some days ; the peroxide of iron is reduced to pro- toxide by the sulphuretted hydrogen, and a proportional quan- tity of sulphur is deposited ; this is carefully collected on a small weighed filter, washed and dried at a gentle heat ; the filter must be protected from the atmosphere during the pro- cess of filtration ; from the weight of the sulphur the quantity of oxygen that has entered into combination with the hydro- gen of the decomposed sulphuretted hydrogen is found ; this oxygen was derived from the peroxide of iron ; and, by mul- tiplying it by three, the whole quantity of oxygen that was present in the substance in the form of sesquioxide of iron is found. Another method (Tresenius). A known quantity of the finely divided substance is introduced into the flask A (Fig. 72), which is then filled with carbonic acid through the tube d\ hydrochloric acid, not in great excess, is then added through the funnel c, and the solution of the compound assisted by heat, a steam of carbonic acid passing all the time through the apparatus; hot water is next added, and the solution T 2 812 QUANTITATIVE ANALYSIS. Fig. 72. boiled and allowed to cool; pure recently precipitated car- bonate of baryta is now mixed into a milky fluid, and poured into the flask through the funnel till it pre- dominates ; the whole mix- ture is then digested at a very gentle heat. The flask is filled with boiling water nearly up to the end of the tube 5, which, being de- pressed as far as necessary into the liquid, the clear fluid is drawn off; the tube is then raised, and the flask again filled with water; when the precipitate has settled, the clear fluid is again drawn off by the siphon b, the flask is then rinsed out with boiling water, and the pre- cipitate thrown on a filter, and well washed with boiling water, as much as possible out of access of air : the amount of peroxide of iron in the washed precipitate is then determined. The liquid is drawn off by the siphon, and the filtrate from the precipitate contains the whole of the dissolved protoxide of iron ; the solution is concentrated, the iron peroxidized, and, finally, precipitated, after the removal of the baryta. Indirect method (Rose). A weighed portion of the substance is dissolved in hydrochloric acid in a flask, which has pre- viously been filled with carbonic acid gas ; the solution being effected, solution of chloride of gold and sodium is added in excess, and the flask closed ; reduction takes place, and metallic gold is precipitated, which is collected, washed, ignited, and weighed ; from the quantity obtained, the quantity of oxygen, which was necessary to convert the protoxide of iron into per- oxide, is ascertained by calculation, thus 6Fe Cl + Au C1 3 = 3 Fe 2 C1 3 4- Au. One equivalent of precipitated gold corresponds to six equiva- lents of protochloride or protoxide of iron. Another method (Fuch). A weighed portion of the substance is dissolved, as in Rose's meiliod, in hydrochloric acid, in a flask which has previously been filled with carbonic acid gas ; IRON. 313 a weighed slip of clean copper is introduced, and the flask, having been filled nearly to the brim with boiled water, accu- rately closed ; the mixture is digested until the fluid becomes colourless, or nearly so ; the slip of copper is then removed from the flask, dried, and weighed ; the diminution in weight indicates the amount of chlorine consumed to convert the original protochloride of iron into perchloride, every equivalent of copper corresponding to an equivalent of chlorine, and every one equivalent of chlorine converting two equivalents of proto- chloride of iron into perchloride, 2FeCl + Cl = Ee 2 Cl 3 ; it fol- lows that every equivalent of dissolved copper corresponds to two equivalents of perchloride of iron in the solution, or, what amounts to the same, to two equivalents of peroxide of iron present in the analysed substance. The quantity of iron actually present in the specimen must be determined by per- oxidizing a solution of a weighed quantity, and precipitating by ammonia. The method is founded on the fact that when air is excluded, hydrochloric acid is incapable of dissolving copper ; but that on the addition of peroxide of iron, or when that substance is already present in the mixture, the acid dis- solves a quantity of copper corresponding thereto. Separation of Oxide of Iron from Oxides of Nickel, Cobalt, Manganese, Zinc, and from the Oxides of tlie Third and Second Groups. (a.) From the Oxides of Nickel and Cobalt (Field's method). The solution containing the oxides in the form of nitrates is evaporated to dryness, and after the addition of water, litharge is added, and the whole boiled for ten minutes or a quarter of an hour. The iron is entirely precipitated, the nitrates of nickel, cobalt, and lead remaining in solution. After nitration, which can be effected with great readiness, dilute sulphuric acid is added, and on standing for sixteen hours, the sulphate of lead is filtered off, and the nickel and cobalt precipitated and estimated. The filter containing the peroxide of iron and excess of litharge is digested in dilute sulphuric acid, filtered and washed ; the sesquioxide of iron in the filtrate is preci- pitated by ammonia. (b.) From Manganese. The iron in the solution is first brought to the state of sesquioxide, and is then precipitated by succinate or benzoate of ammonia, a sufficient quantity of sal- ammoniac having previously been added ; the precipitated suc- cinate is treated in the manner already described, and the 314 QUANTITATIVE ANALYSIS. manganese in tlie filtrate is precipitated by carbonate of soda. Another method of separation is by digesting the solution of the two metals with excess of recently precipitated carbonate of baryta, which throws down the iron as basic carbonate of peroxide ; the washed precipitate is dissolved in hydrochloric acid, the baryta separated by sulphuric acid, and the iron pre- cipitated by ammonia ; the filtrate from the basic carbonate of peroxide of iron contains the manganese, together with a solu- ble baryta salt, the latter is removed by sulphuric acid, and the manganese precipitated by carbonate of soda. Another method (Field). The solution containing the oxides is boiled with oxide of lead, by which the whole of the iron is precipitated ; from the filtrate, the oxide of lead is precipitated by dilute sulphuric acid, and the sulphate of lead being filtered off, the oxide of manganese is precipitated from the filtrate by potassa. The mixed oxides of lead and iron are redissolved in dilute nitric acid, and then digested with dilute sulphuric acid, filtered and washed, and the sesquioxide in the filtrate precipi- tated by ammonia. (c.) From Zinc. The iron in the solution is peroxidized by nitric acid or by chlorate of potassa, the solution is evaporated to dryness, and all excess of acid removed ; the residue is dis- solved in acetic acid, and the zinc precipitated by sulphuretted hydrogen : the precipitated sulphide of zinc should have a pure white colour. (d.) From Chromium (Eose). To the solution of the two metals a sufficient quantity of tartaric acid is added to prevent the precipitation of either of the metals by potassa, that alkali is then added, and the iron precipitated by sulphide of potas- sium ; the solution filtered from the sulphide of iron contains the oxide of chromium; it is evaporated to dryness, ignited, fused with carbonate of soda and nitre, and the chromium in the alkaline chrornate thus formed determined as directed (page 290). Another method (Berthier). The oxides are precipitated by ammonia or carbonate of ammonia, and, while still moist, di- gested with slight excess of sulphurous acid : the whole of the iron dissolves, and also a certain quantity of the oxide of chro- mium, while the remainder of this latter metal is converted into pure subsulphite. The solution is boiled until it is de- colorized, when it only contains iron. To precipitate this metal the sulphurous acid is either expelled by sulphuric acid, IRON. 315 or oxidized by aqua-regia, an alkali or an alkaline carbonate is theo added ; or the iron is precipitated by an alkaline sulphide without expelling the sulphurous acid. Another method (Liebig). The mixture of the two metals in solution is first saturated with sulphuretted hydrogen, to be certain that the iron is contained in the liquid as protoxide, (an addition of a few drops of sulphide of ammonium answers the purpose), and then thrown down by cyanide of potassium, and an excess of the latter added : the iron dissolves immediately as ferrocyanide of potassium, while the oxide of chromium re- mains behind, (e.) From Tttria (Scherer). To a neutral solution, oxalate of potassa is added, a white crystalline precipitate consisting of the double oxalate of yttria and potassa, is gradually formed, which, by ignition, is converted into yttria and carbonate of potassa; the mixture is dissolved in hydrochloric acid, diluted with much water, and the yttria precipitated by caustic ammo- nia, it must be well washed with boiling water, after which it may be ignited and weighed. Another method (Berthier). The moist hydrates are boiled with sulphurous acid, the yttria is deposited, and the iron remains in solution ; to prevent the formation of an ochreous deposit from the action of the air, the solution should be boiled in a flask with a long neck, and when no more sulphu- rous a-eid is disengaged, it should be filled with boiling water and corked ; when it has become cold, the liquid is decanted on to a filter replaced by boiling water, and finally filtered and edulcorated. (f.) From Alumina (Weeren). To the solution, containing the two oxides a sufficient quantity of tartaric acid is added to prevent the precipitation of the bases by ammonia, sulphide of ammonium is then added in excess, and the beaker covered with a glass plate till the whole of the sulphide of iron has pre- cipitated. The clear supernatant liquid is then poured. off, the precipitate thrown on a filter, and rapidly washed with water containing a little sulphide of ammonium. The sulphide of iron is dissolved off the filter by hydrochloric acid, and the iron having been peroxidized by nitric acid is precipitated by am- monia. The solution containing the whole of the alumina is evaporated to dryness and ignited strongly ; if the amount of charcoal be small^ it is to be completely burnt ; but if it be con- siderable, it must be treated with boiling hydrochloric acid, and 316 QUANTITATIVE ANALYSIS. the alumina precipitated from the filtered solution by sulphide of ammonium. When a very small quantity of alumina is present, Eeynolds recommends the addition to the solution of sufficient oxalic acid to prevent the precipitation of the earth, the solution is then poured into a mixture of ammonia and sulphide of ammonium, the precipitated sulphide of iron is treated as above, the solution containing the alumina is concentrated and then boiled with hydrochloric acid and chlorate of potassa, this will oxidize the oxalic acid and the sulphide of ammonium ; the solution is now rendered alkaline with carbonate of am- monia, and boiled for some time, when the alumina will be precipitated. ' "When alumina has to be separated from oxide of iron, which is accompanied by manganese, lime, and magnesia, Eichter fuses the compound with ten times its weight of carbonate of soda, and then, by means of water, with the addition of potassa, dissolves the alumina as aluminate of potassa and soda. Another method (Fresenius). The iron in the solution is; reduced to the minimum of oxidation by boiling with sulphite of soda ; it is then neutralized with carbonate of soda, an excess of caustic soda added, and the solution, after being well agi- tated, is boiled until it has become black and granular (a coil of platinum wire put into the flask prevents succussion) ; it is allowed to subside, the clear liquid passed through a close filter, and the precipitate washed with hot water, at first by decautation and then on the filter > the filtered solution treated with hydrochloric acid and chlorate of potassa furnishes, upon precipitation with ammonia, and several hours' standing, the whole of the alumina in a perfectly pure state. (g.) From Lime and Magnesia (Rose). Ammonia is added in excess, and the liquid boiled ; a little lime is carried down in the precipitate, but it is completely redissolved on ebulli- tion ; tbe lime in the filtrate is precipitated by oxalic acid, and the magnesia in the state of ammonia-phosphate of magnesia : or the hydrochloric solution of the oxides may be diluted and neutralized as far as practicable with ammonia, and the oxide of iron then precipitated by succinate or benzoate of ammonia. Dry assay of Iron ores. The object of the dry assay of an iron ore is to ascertain, by an experiment on a small scale, the amount of iron which the ore should yield when smelted on the large scale in the blast furnace. For this purpose the IRON. 317 metal must be deoxidized, and such a temperature produced as to melt the metal and the earths associated with it in the ore, so that the former may be obtained as a dense button at the bottom of the crucible, and the latter in a lighter glass or slag above it. Such a temperature can only be obtained in a wind furnace, connected with a chimney at least thirty feet high ; and when made expressly for assaying, the furnace is generally built of such a size that four assays may be made at the same time, namely about fourteen inches square and two feet in depth from the under side of the cover to the moveable bars of iron which form the grate. In order that the substances associated with the iron in the ore, should form a fusible com- pound it is usually requisite to add a flux, the nature of which will depend on the character of the ore under examination. Berthier divides iron ores into five classes : 1. The almost pure oxides, such as the magnetic oxide, ologistic iron, and the hce- matites. 2. Ores containing silica, but free, or nearly so, from any other admixture. 3. Ores containing silica and various bases, such as lime, magnesia, alumina, oxide of manganese, oxide of titanium, oxide of tantalum, oxide of chromium, or oxide of tungsten, but little or no silica. 5. Ores containing silica, lime, and another base, and which are fusible alone. Ores of the first class may be reduced without any flux ; but it is always better to employ one, as it greatly facilitates the formation of the button : borax may be used, or better, a fusi- ble earthy silicate, such as ordinary flint-glass. Ores of the second class require some base to serve as a flux, such as carbonate of soda, or a mixture of carbonate of lime and clay, or of carbonate of lime and dolomite. Ores of the third class are mixed with carbonate of lime in the proportion of from one-half to three-fourths of the weight of the foreign matter present in the ore. Ores of the fourth class require as a flux silica in the form of rounded quartz, and generally also some lime. The manganesian spathic ores which belong to this class, may be assayed with the addition of silica alone, but the magnesian spathic ores require lime. Ores of the fifth class require no flux. Method of conducting the assay. One hundred grains of the ore finely pulverized, and passed through a silk sieve, are well mixed with the flux, and the mixture introduced into the smooth concavity made in the centre of a crucible that has been lined with charcoal. The lining of the crucible is effected 318 QUANTITATIVE ANALYSIS. by partially filling it with coarsely powdered and slightly damped charcoal, which is then rammed into a solid form by the use of a light wooden pestle. The mingled ore and flux must be covered with charcoal. The crucible thus filled, is closed with an earthen lid luted on with fire clay, and is then set on its base in the air-furnace. The heat should be very slowly raised ; the damper remaining closed during the first half-hour. In this way the water of the damp charcoal exhales slowly, and the deoxidation of the ore is completed before the fusion begins ; if the heat were too high at first, the luting would probably split, and moreover the slag formed would dissolve some oxide of iron, which would of course be lost to the button, and thus give an erroneous result. After half an hour the damper is gradually opened, and the furnace being filled with fresh coke, the temperature is raised progressively to a white heat, at which pitch it must be maintained for a quarter of an hour : the damper is then closed, and the furnace allowed to cool. As soon as the temperature is suffi- ciently reduced, the crucible is removed, and opened over a sheet of brown paper ; the brasque is carefully removed, and the button of cast iron taken out and weighed. If the experi- ment has been entirely successful, the iron will be found at the bottom of the crucible in a small rounded button, and the slag will be entirely free from any adhering metallic globules, and will resemble in appearance green bottle-glass ; should, how- ever, the slag contain small metallic particles, the experiment is not necessarily a failure, as they may generally be recovered by washing, and the magnet : but if, on breaking the crucible, the reduced metal should be found in a partially melted state, and not collected into a distinct mas.s, it indicates either too low a temperature or an improper selection of fluxes, and the experiment must be repeated. The iron obtained is not chemi- cally pure ; it contains carbon, and, if the ore be manganiferous, manganese : the result is therefore somewhat too high, though indicating with sufficient exactness for all manufacturing pur- poses, the richness of the ore assayed. Humid assay of iron ores. Too great care cannot be bestowed on the sampling of ores intended for analysis. To expend so much time and labour on an isolated specimen (un- less for a special object) is worse than useless. The sample operated upon should be selected from a large heap which should be thoroughly gone over, and several pieces taken from IRON. 319 different parts ; these should be coarsely powdered and mixed, and about half a pound taken from the mass should be pre- served in a well-stoppered bottle for the analysis. (1.) Determination of water (hygroscopic and combined). About 50 grains of the ore are dried in the water-oven, Fig. 61 B, till no further loss of weight takes place ; the loss indicates hygroscopic water ; the residue is introduced into a tube of hard glass, to which is adapted a weighed tube con- taining chloride of calcium ; the powder is then gradually raised to a low red-heat, the combined water is thereby ex- pelled, and its amount determined by the increase in weight of the chloride of calcium tube. Some ores (the hydrated hsema' tites) contain as much as 12 per cent, of combined water. (2.) Sulphuric acid and sulphur. From 30 to 50 grains of the ore are digested with hydrochloric acid, filtered and washed. The filtrate, concentrated, if necessary, by evapora- tion, is precipitated by excess of chloride of barium. The insoluble residue on the filter is fused in a platinum or gold crucible, with nitre and carbonate of soda ; the fused mass is dissolved in hydrochloric acid, evaporated to dryness, redis- solved in dilute hydrochloric acid, filtered, and precipitated as before, by chloride of barium ; even 100 parts of the sulphate of baryta produced indicate 13'734 parts of sulphur, corre- sponding with 25-48 parts of pyrites. In the analysis of hema- tites it is necessary to bear in mind that perchloride of iron is partially reduced when boiled with finely divided iron pyrites and hydrochloric acid, sulphuric acid being formed. (Dick.) (3.) Phosphoric acid. From 50 to 75 grains of the ore are digested with hydrochloric acid and filtered. The clear solu- tion, which should not be too acid, is boiled with sulphite of ammonia, added gradually in small quantities, till it either becomes colourless or acquires a pale green colour, indicating that the peroxide of iron originally present has been reduced to protoxide. The solution is nearly neutralized with carbo- nate of ammonia, excess of acetate of ammonia added, and the liquid boiled ; strong solution of perchloride of iron is then added, drop by drop, until the precipitate which forms has a distinct red colour : this precipitate, which contains all the phosphoric acid originally present in the ore, is collected on a filter, washed, and redissolved in hydrochloric acid, tartaric acid added, and then ammonia. From this ammoniacal solu- tion the phosphoric acid is finally precipitated as ammonio- 320 QUANTITATIVE ANALYSIS. magnesian phosphate, by the addition of chloride of ammonium, sulphate of magnesia, and ammonia. The precipitate is allowed 24 hours to subside, it is then collected on a filter, and if it has a yellow colour, which is almost invariably the case, it is redissolved in hydrochloric acid, and more tartaric acid being added, it is again precipitated by ammonia : 100 parts of the ignited pyrophosphate of magnesia correspond to 63*8 parts of phosphoric acid. (4.) Determination of the remaining constituents. 25 or 30 grains of the finely-powdered ore are digested for about half an hour with strong hydrochloric acid, diluted with boiling dis- tilled water, and filtered. The residue on the filter being thoroughly washed, the solution is peroxidized, if necessary, by the addition of chlorate of potassa, nearly neutralized by ammonia, boiled with excess of acetate of ammonia, and rapidly filtered while hot ; the filtrate (which should be colourless) together with the washings, is received in a flask, ammonia is added, and then a few drops of bromine, and the flask closed with a cork. In a few minutes, if manganese be present, the liquid acquires a dark colour ; it is allowed to remain at rest for 24 hours, then warmed, and rapidly filtered and washed ; the brown substance on the filter is hydrated oxide of manga- nese : it loses its water by ignition, and then becomes Mn 3 O 4 , 100 parts of which correspond to 92 - 14 parts of protoxide. The liquid filtered from the manganese contains the lime and magnesia ; the former is precipitated by oxalate of ammonia, and the oxalate of lime formed converted by ignition into carbonate, in which state it is either weighed, having been previously evaporated with carbonate of ammonia, or it is con- verted into sulphate by the addition of a few drops of sulphuric acid, evaporation, and ignition. The lime being separated, the magnesia is thrown down as ammonio-magnesian phosphate by phosphate of soda and ammonia, and after standing for twenty- four hours it is collected on a filter, washed with cold ammonia- water, dried, ignited, and weighed ; 100 parts of carbonate of lime correspond to 56'0 of lime ; 100 parts of sulphate of lime to 40*1 of lime, and 100 parts of pyrophosphate of magnesia to 36*22 of magnesia. The red precipitate collected on the filter after the boiling with acetate of ammonia, consists of the basic acetate of iron, and perhaps of alumina, together with the phosphoric acid. It is dissolved in a small quantity of hydrochloric acid, and then IEON. 321 boiled in a silver or platinum basin with considerable excess of pure caustic potassa ; the alumina (with the phosphoric acid) is hereby dissolved, the insoluble portion is allowed to subside, and the clear liquid is then decanted, after which the residue is thrown on a filter and washed ; the nitrate and washings are supersaturated with hydrochloric acid, nearly neutralized with ammonia, and the alumina finally precipitated by carbonate of ammonia. From the weight of the ignited precipitate, the corresponding amount of phosphoric acid determined by a separate operation is to be deducted, the remainder is calculated as alumina. The residue left after digesting the ore with hydrochloric acid, consists principally of silica, but it may also contain alumina, peroxide of iron, lime, magnesia, and potassa. For practical purposes it is rarely necessary to submit it to minute examination ; should such be desired, it must be dried, ignited, and weighed, then fused in a platinum crucible with four times its weight of mixed alkaline carbonates, the fused mass dissolved in dilute hydrochloric acid, and evaporated to dryness, the residue moistened with strong hydrochloric acid, and after standing at rest for some hours, digested with hot water, filtered, and the silica on the filter ignited and weighed. The alumina, lime, oxide of iron, and magnesia in the filtrate are separated from each other according to the instructions given above ; the potassa is estimated by a distinct process. (5.) Carbonic acid. This acid, which constitutes a considerable part of the weight of that large and important class of ores, the clay ironstones, is estimated by noting the loss sustained after adding to a weighed portion of the ore sulphuric acid, and thus evolving the gas (see " Carbonic Acid") ; or more roughly, by the loss sustained in the entire analysis. Another method is to fuse 20 or 25 grains of the ore with 60 or 80 grains of dry borax, and noting the loss, which consists of water and car- bonic acid ; by deducting the water obtained in a previous ex- periment, the quantity of carbonic acid is obtained. This me- thod, however, can scarcely be recommended on account of the corrosion of the crucible, though the results are very accurate. (6.) Determination of the iron. This is performed on a se- parate portion of the ore, either by the volumetric method of Marguerite, or by that of Dr. Penny : both give very exact results. (a.) Marguerite's method. This is based on the reciprocal ac- tion of the salts of protoxide of iron and permanganate of po- 322 QUANTITATIVE ANALYSIS. tassa, whereby a quantity of the latter is decomposed exactly proportionate to the quantity of iron. The ore (about 10 or 15 grains) is dissolved in hydrochloric acid, and the metal brought to the minimum of oxidation by boiling the solution with sul- phite of soda (or better, by pure metallic zinc) ; the solution of permanganate of potassa is then cautiously added drop by drop, until the pink colour appears, and the number of divisions of the burette required for the purpose accurately noted. The solution should be considerably diluted, and there must be a sufficient quantity of free acid present to keep in solution the peroxide of iron formed, and also the oxide of manganese. The whole of the iron must be at the minimum of oxidation, and the excess of sulphurous acid must be completely expelled : if the latter precaution be neglected an erroneous result will be obtained, as the sulphurous acid will itself take oxygen from the permanganic acid, and thus react in the same manner as iron. To prepare the permanganate of potassa, 7 parts of chlorate of potassa, 10 parts of hydrate of potassa, and 8 parts of per- oxide of manganese are intimately mixed. The manganese must be in the finest possible powder, and the potassa having been dissolved in water, is mixed with the other substances, dried, and the whole heated to very dull redness for an hour. The fused mass is digested with water, so as to obtain as con- centrated a solution as possible, and dilute nitric acid added till the colour becomes violet ; it is afterwards filtered through asbestos. The solution must be defended from the contact of organic matter, and kept in a glass stoppered bottle. If the solution be evaporated, it yields beautiful red acicular crystals. It is better to employ the crystals in the preparation of the test-liquor, as the solution keeps much better when no manga- nate is present. To prepare the normal or test-liquor, a cer- tain quantity, say 5 grains of pianoforte wire, are dissolved in pure hydrochloric acid ; after the disengagement of hydrogen has ceased, and the solution is complete, the liquor is diluted, with about a pint of water, and accurately divided by measure- ment into two equal parts, the number of burette divisions of the solution of permanganate required to produce in each the pink colour is accurately noted, and this number is then em- ployed to reduce into weight the result of the analysis of an ore. A useful normal liquor is made by dissolving 100 grains of the crystals in 10,000 grains of water. IRON. 323 (b.) Penny's method. This is based on the reciprocal action of chromic acid and protoxide of iron, whereby a transference of oxygen takes place, the protoxide of iron becoming con- verted into peroxide, and the chromic acid into sesquioxide of chromium. The process is conducted as follows : A conve- nient quantity of the specimen is reduced to coarse powder, and one-half at least of this is still further pulverized until it is no longer gritty between the fingers. The test solution of bichromate of potassa is next prepared, 444 grains of this salt, in fine powder, are weighed out, and put into a burette graduated into 100 equal parts ; warm distilled water is after- wards poured in until the instrument is filled to 0. The palm of the hand is then securely placed on the top, and the con- tents agitated, by repeatedly inverting the instrument, until the salt is dissolved, and the solution rendered of uniform density throughout. Each division of the solution thus pre- pared contains 0*444 grain of bichromate, which Dr. Penny ascertained to correspond to half a grain of metallic iron. The bichromate must be pure, and should be thoroughly dried by being heated to incipient fusion. 100 grains of the pulverized ironstone are now introduced into a Florence flask with \\ oz. by measure of strong hydrochloric acid and \ oz. of water: heat is cautiously applied, and the mixture occasionally agi- tated, until all effervescence caused by the escape of carbonic acid ceases ; the heat is then increased, and the mixture made to boil, and kept at a moderate ebullition for ten minutes or a quarter of an hour ; about 6 oz. of water are next added, and mixed with the contents of the flask, and the whole filtered into an evaporating-basin. The flask is rinsed several times with water, to remove all adhering solution, and the residue on the filter is well washed. Several small portions of a weak solution of red prussiate of potassa (containing 1 part of salt to 40 of water) are now dropped upon a white porcelain slab, which is conveniently placed for testing the solution in a basin during the next operation. The prepared solution of bichro- mate of potassa in the burette is then added very cautiously to the solution of iron, which must be repeatedly stirred, and as soon as it assumes a dark greenish shade it should be occasionally tested with the red prussiate: this may easily be done by taking out a small quantity on the end of a glass rod and mixing it with a drop of the solution on the porcelain slab. When it is noticed that the last drop communicates a 324 QUANTITATIVE ANALYSIS. distinct Hue tinge, the operation is terminated ; the burette is allowed to drain for a few minutes, and the number of divisions of the test liquor consumed, read oft'. This number multiplied by 2 gives the amount of iron per cent. The necessary calculation for ascertaining the corresponding quan- tity of protoxide is obvious. If the specimen should contain iron in the form of peroxide, the hydrochloric solution is de- oxidized by sulphite of ammonia. The presence of peroxide of iron in an ore is easily detected by dissolving 30 or 40 grains in hydrochloric acid, diluting with water, and testing a portion of the solution with sulpha-cyanide of potassium. If a decided blood-red colour is produced, peroxide of iron is present. If it be desired to ascertain the relative proportions of peroxide and protoxide in an ore, two operations must be performed, one on a quantity of ore that has been dissolved in hydro- chloric acid in a stout closed bottle, or in a flask through which a current of carbonic acid gas is maintained, and another on a second quantity that has been dissolved as usual and then deoxidized by sulphite of ammonia or by pure metallic zinc. It is advisable to employ the solution of bichromate much weaker than that proposed by Penny, and to employ a burette graduated to cubic millimetres a good strength is 1 grain of metallic iron=10 cubic centimetres of bichromate. Metals precipitable by Sulphuretted Hydrogen from the Hydrochloric solution. A weighed portion of the ore, varying from 500 to 2000 grains, is digested for a considerable time in hydrochloric acid : the solution is filtered oif ; the iron in the filtrate reduced when necessary by sulphite of ammonia, and a current of sulphuretted hydrogen passed through it. The small quantity of sulphur (which is always suspended) is col- lected on a filter and thoroughly washed ; it is then incinerated at as low a temperature as possible. The residue (if any) is mixed with carbonate of soda and heated upon charcoal before the blowpipe : any globules of metal that may be obtained are dissolved and tested. Analysis of Pig Iron. The most important constituents to be determined are carbon (combined and uncombined), silicon, sulphur, phosphorus ; those of less consequence, or of more rare occurrence, are manganese, arsenic, copper, zinc, chromium, titanium, cobalt, nickel, tin, aluminum, calcium, magnesium, and the metals of the alkalies. (1.) Determination of the total amount of Carbon. About .100 IRON. 325 grains of the iron in small pieces are digested, at a moderate temperature, in 6-oz. measures of a solution formed by dissolv- ing 6 oz. of crystallized sulphate of copper, and 4 oz. of com- mon salt in 20 oz. of water and 2 oz. of concentrated hydro- chloric acid. The action is allowed to proceed until all, or nearly all the iron is dissolved. Carbon and copper are left insoluble ; these are collected on a filter, and washed first with dilute hy- drochloric acid (to prevent the precipitation of subchloride of copper) , then with water, then with dilute caustic potassa, and finally with boiling water. The mixed carbon and copper are dried on the filter, from which they are easily removed by a knife- blade, and are mixed with oxide of copper, and burnt in a com- bustion tube in the usual way, with a current of air or, still better, of oxygen. The carbonic acid is collected in Liebig's appara- tus, and from its weight the amount of carbon is calculated. (2.) Graphite, or IJncombmed Carbon. A weighed portion of the finely divided iron (filings or borings may be used) is di- gested with moderately strong hydrochloric acid ; the combined carbon is evolved in combination with hydrogen, while the gra- phite is left undissolved. It is collected on a filter, washed, and then boiled with a solution of caustic potassa, sp. gr. 1*27, in a silver dish ; the silica which existed in the iron in the form of silicon is hereby dissolved; the clear caustic solution is drawn off by a pipe" or siphon, and the black residue repeatedly washed ; it is dried at as high a temperature as it will bear, and weighed ; it is then heated to redness in a current of air, until the whole of the carbon is burnt off. A reddish residue gene- rally remains, which is weighed, and the weight deducted from that of original black residue, the difference gives the amount of graphite. This residue frequently contains titanic acid. (3.) Silicon. The amount of this element is determined by evaporating to dryness a hydrochloric solution of a weighed quantity of the metal : the dry residue is re-digested with hy- drochloric acid, diluted with water, boiled and filtered ; the in- soluble matter on the filter is washed, dried, and ignited, until the whole of the carbon is burnt off; it is then weighed, after which it is digested with solution of potassa, and the residue, if any, washed, , dried, ignited, and weighed : the difference be- tween the two weights gives the amount of silicic acid, 100 parts of which indicate 47 parts of silicon. (4.) Phosphorus. A weighed portion of the metal is digested in nitro-hydrochloric acid, evaporated to dryness, and the resi- PAKT u. z 326 QUANTITATIVE ANALYSIS. due re-digested with hydrochloric acid. The solution is treated precisely as recommended for the determination of phosphoric acid in ores ; every 100 parts of pyrophosphate of magnesia indicate 27'87 parts of phosphorus. (5.) Sulphur. In grey iron this element is very conveniently and accurately estimated by allowing the gas evolved by the ac- tion of hydrochloric acid on a weighed quantity (about 100 grains) of the metal, in filings or borings, to pass slowly through a solution of acetate of lead acidified by acetic acid : the sulphur, the whole of which takes the form of sulphuretted hydrogen, enters into combination with the lead, forming a black precipi- tate of sulphide of lead, which is collected, washed, and con- verted into sulphate of lead, by digesting it with nitric acid, evaporating to dryness, and gently igniting : 100 parts sulphate of lead=10'55 sulphur. The most minute quantity of sulphur in iron is detected by this process. If, however, crude white iron is under examination, this method does not give satisfac- tory results on account of the difficulty with which it is acted upon by hydrochloric acid ; it is better, therefore, to treat the metal with nitro-hydrochloric acid, evaporate to dryness, re- digest with hydrochloric acid, and then precipitate the filtered solution with excess of chloride of barium; or the finely divided metal may be fused in a gold crucible with an equal weight of pure nitrate of soda and twice its weight of pure alkaline carbonates ; the fused mass is extracted with water acidified with hydrochloric acid, and finally precipitated by chlo- ride of barium. M. Nickles (Journ. de Pharm. efc de Chimie) suggests the use of pure bromine mixed with distilled water as a solvent for cast iron or steel. (6.) Manganese. This metal is determined by the process de- scribed for its estimation in iron ores ; the iron must exist in the solution in the form of sesquioxide. (7.) Arsenic and Copper. The nitro-hydrochloric solution of the metal is evaporated to dryness, re-digested with hydrochloric acid, and filtered. The iron in the clear solution is reduced to protochloride by boiling with a sufficient quantity of sulphite of ammonia ; the solution is boiled till it has lost all smell of sul- phurous acid. It is then saturated with sulphuretted hydro- gen, and allowed to stand for twenty -four hours in a closed vessel, the excess of gas is boiled off, and the precipitate, if any, collected on a small filter and well washed ; it is digested with monosulphide of potassium, which dissolves the sulphide of IRON. 327 arsenic, leaving the sulphide of copper untouched ; the latter is decomposed by heating with nitric acid, and the presence of copper evinced by the addition of ammonia, which produces a fine blue colour ; the sulphide of arsenic is precipitated from its solution in sulphide of potassium by dilute sulphuric acid ; it may be redissolved in aqua-regia, and the nitric acid having been expelled by evaporation, the arsenic may be reduced in Marsh's apparatus. (See ARSENIC.) (8.) Nickel and Cobalt. These metals, if present, willbe found in the solution from which the copper and arsenic have been precipitated by sulphuretted hydrogen. The solution is per- oxidized, and the sesquioxide of iron precipitated by slight ex- cess of carbonate of baryta, after which the nickel and cobalt are precipitated by sulphide of ammonium. (9.) Chromium and Vanadium. These metals, which should be looked for in the carbonaceous residue obtained by dissolving a large quantity of the iron in dilute hydrochloric or sulphuric acid, are detected as follows (Wohler) : The ignited residue is intimately mixed with one-third of its weight of nitre, and exposed for an hour in a crucible to a gentle ignition. When cool, the mass is powdered and boiled with water. The filtered solution is gradually mixed and well stirred with nitric acid, taking care that it may still remain slightly alkaline, and that no nitrous acid is liberated which would reduce the vanadic and chromic acids. The solution is then mixed with an excess of solution of chloride of barium as long as any precipitate is pro- duced. The precipitate, which consists of vanadate and chro- mate of baryta, is decomposed with slight excess of dilute sul- phuric acid, and filtered. The filtrate is neutralized with am- monia, concentrated by evaporation, and a fragment of chloride of ammonium placed in it. In proportion as the solution be- comes saturated with chloride of ammonium, vanadate of am- monia is deposited as a white or yellow crystalline powder. To test for chromium only, the mass after fusion with nitre is ex- tracted with water, and then boiled with carbonate of ammonia ; the solution is neutralized with acetic acid, and then acetate of lead added : the production of a yellow precipitate indicates chromic acid. (10.) Aluminum. This metal is best separated from iron, by first reducing the latter to the state of protoxide by sulphite of ammonia, then neutralizing with caroonate of soda, and after- wards boiling with excess of caustic potassa, until the precipi- z 2 328 QUANTITATIVE ANALYSIS. tate is black and pulverulent. The solution is then filtered off, slightly acidulated with hydrochloric acid, and the alumina pre- cipitated by sulphide of ammonium. (11.) Calcium and Magnesium. These metals are found in the solution from which the iron and aluminum have been separated ; they both exist probably (together with the aluminum) in the cast-iron in the form of slag, and are best detected in the black residue which is left on dissolving the iron in dilute sulphuric or hydrochloric acid. After digesting this residue with caustic potassa, and burning away the graphite, a small quantity of a red powder is left, which is composed of silicic acid, oxide of iron, alumina, lime, magnesia, and perhaps titanic acid ; if 500 grains of cast-iron are operated upon, a sufficient quantity of insoluble residue will be obtained for a quantitative determi- nation of its constituents. Moderately strong acid should be used, and no heat applied ; it should be allowed to act on the iron for about a week, being repeatedly agitated. (12.) Determination of Carbon in Cast Iron and Steel. About 100 grains of the steel in small pieces are digested in 6-oz. mea- sures of a solution of chloride of copper made by mixing 6 oz. of crystallized sulphate of copper, and 4 oz. of chloride of so- dium with 20 oz. of water, and 2 oz. of concentrated hydrochlo- ric acid. The mixture is warmed only, not heated so as to evolve hydrogen. The action is allowed to proceed until all, or nearly all, of the steel is dissolved. The residue, consisting of carbon and metallic copper, is collected on a filter and washed with dilute hydrochloric acid to prevent the precipitation of subchloride of copper, then with water, then with dilute caustic potassa, to ensure the removal of all the acid, and finally with water, it is then dried. The carbon and copper are easily re- moved from the filter by the blade of a knife, they are mixed with oxide of copper, and burnt in a combustion furnace with a tmrrent of air purified from carbonic acid by passing through solution of caustic potassa ; the carbonic acid is collected as usual in Liebig's bulbs after drying by chloride of calcium. If graphite be present, it will be necessary to burn the carbon in a current of oxygen gas. 170. URANIUM. Oxide of Uranium is, according to Eose, completely precipi- tated from its acid solutions, having previously saturated them with ammonia, by the addition of sulphide of ammonium. No URANIUM. 329 inconvenience results from the solution containing any am- moniacal salts, except carbonate of ammonia and all alkaline carbonates. The precipitate is Hack or reddish-brown if the sulphide be greatly in excess. It is washed in water to which sulphide of ammonium is added. The precipitate is essentially protoxide of uranium, the composition of which is One equivalent of U . . . . 60 . . 88'23 One ditto of O . 8 11-77 One ditto of UO . . . 68 . . 100*00 After being dried, it is ignited, to expel any little sulphur which may be retained : then it is calcined in a current of hydrogen at a high temperature, and left to cool in the gas. The protoxide is thus obtained pure. Should the solution contain much salts of potassa or of other strong non-volatile bases, the precipi- tate will retain a small quantity of these bases. Separation of Oxide of Uranium from other metallic Oxides: (a.) From those which are completely precipitated from their solutions by Sulphide of Ammonium. Add to the solution excess of carbonate of ammonia mixed with sulphide of ammonium ; all the oxides which sulphide of ammonium transforms into sulphides are precipitated, whilst the protoxide of uranium is dissolved in the carbonate of am- monia. Leave the mixture to deposit in a closed vessel ; wash the precipitate by decantation, with water containing carbonate of ammonia and sulphide of ammonium, and then filter. Gently heat the filtered liquid to expel most of the carbonate ; de- compose the sulphide with hydrochloric acid, oxidize the pro- toxide of uranium by nitric acid, precipitate the oxide by am- monia, and, before weighing, calcine in a current of hydrogen. The tendency of the oxide of uranium to combine with bases, gives rise to its frequently carrying down along with it, in its precipitation by ammonia, bases which are not precipitated when alone, by that reagent, such as baryta and lime. If the liquid contain much potassa, a considerable quantity is often found in the precipitate, which is easily recognized by the orange-yellow colour perceptible at various parts of the heated precipitate. Sulphuretted hydrogen separates uranium from a great number of metals. The oxide may be separated very easily from the peroxide of iron by means of carbonate of am- monia ; but those metals whose oxides dissolve in part, or wholly, 330 QUANTITATIVE ANALYSIS. in carbonate of ammonia, such as manganese, zinc, cobalt, and nickel, have been regarded difficult of separation from uranium. The use of the carbonates of soda and of potassa affords a very exact and simple means of effecting their removal ; for uranate of potash is not soluble in the carbonate of potassa, but it dissolves completely, and in very little time, in a liquid saturated with a bicarbonated alkali. When a solution of a salt of uranium is precipitated with a slight excess of carbonate of potassa, and the liquid diluted with water, the whole of the uranium is dis- solved, and imparts to it a yellow colour. In both cases the double salt of potassa is formed. The carbonates of zinc, cobalt, and manganese are, on the contrary, insoluble in carbonate of potassa. To effect their separation, the solution may be thrown down, either by potassa, the precipitate being well washed and digested with bicarbonate of potassa, which will only dissolve the the oxide of uranium; or by slight excess of carbonate of po- tassa, the precipitate being collected on a filter, and washed as long as the liquor which passes through is coloured. On add- ing to a solution of the phosphate or arseniate of uranium, in an excess of carbonate of potassa, a known quantity of sesqui- oxide of iron dissolved in nitric acid, those two acids may be completely separated from the uranium, and their proportion determined by the increase in weight of the peroxide of iron. The oxide of uranium may also be separated by potassa, and the phosphoric and arsenic acids left in solution. To separate the oxide of uranium in solution, it may be sa- turated with hydrochloric acid, boiled to expel the carbonic acid, and the uranium precipitated by ammonia ; but as the liquor contains much potassa, the precipitate retains a certain quantity, and is not entirely converted into the state of green oxide, which makes it necessary to redissolve it, and precipi- tate again with ammonia. To extract uranium with great precision from the uranate of potassa dissolved in hydrochloric acid, the whole is evaporated to dryness in a platinum crucible, gradually heated, and a cur- rent of dry hydrogen conveyed into it, as long as the gas which is given off possesses an acid reaction. The double chloride of potassium and uranium is converted into uranium (so called) in the form of a black powder, which is separated by washing from the chloride of potassium. (b.) Separation from Nickel, Cobalt, and Zinc. When the oxide of uranium in its preparation from, pitchblende is so far URANIUM. 331 purified as to be dissolved in carbonate of ammonia, Wohler di- rects to mix sulphide of ammonium with the solution as long as a black precipitate falls ; in this way the nickel, cobalt, and zinc are entirely separated, no uranium being precipitated. According to Berthier, uranium is completely separated from iron, manganese, cobalt, nickel, and zinc, by boiling the solution after the addition of sulphite of ammonia. (c.) From Oxide of Iron (Pisani). As oxide of uranium, dis- solved in carbonate of ammonia, is not precipitated by ain- moniacal sulphide, we have only to add to the liquid separated by nitration with the oxide of iron, several drops of the latter reagent, to eliminate from it in the state of sulphide the small quantity of iron which has been dissolved ; after filtering we get a liquid containing all the uranium without any trace of iron. Analysis of Pitchblende. Of the various methods which have been employed by different chemists, the following, adopted by Arfwedson, seems on the whole the best. The mineral is reduced to fine powder, and digested with boiling aqua-regia ; the solution is evaporated on a water bath, to ex- pel the excess of acid, then diluted with water, and precipi- tated by sulphuretted hydrogen, sulphurous acid being pre- viously added to reduce the arsenic acid to arsenious acid: arsenic, copper, lead, and bismuth are in this manner precipi- tated. The solution is filtered, the excess of sulphuretted hy- drogen expelled by boiling, and ammonia added : the precipi- tate thereby formed is washed and dissolved, while yet moist, in carbonate of ammonia, peroxide of iron remains undissolved, and is separated by filtration. The solution is evaporated till the ammonia is volatilized, whereupon the peroxide of ura- nium is rendered insoluble ; it is washed, dried, and calcined in a platinum crucible, by which it becomes converted into urano-uranic oxide of a green colour. The uranates of lime, zinc, cobalt, or nickel, which may have been present in the so- lution, are not decomposed by the calcination ; they are readily dissolved, however, by dilute hydrochloric acid, which acid serves therefore to separate them from the urano-uranic oxide, which is thus obtained pure. It is washed first with dilute acid, and then with distilled water. From the metallic uranates in solution in hydrochloric acid, oxide of uranium may be obtained. For this purpose the solution is precipitated by ammonia, the dried precipitate is reduced by 332 QUANTITATIVE ANALYSIS. hydrogen gas, and treated immediately with hydrochloric acid, which dissolves all the foreign metals, leaving protoxide of ura- nium insoluble. Another method (Kessler). The mineral is dissolved in nitric acid, water added, and sulphuretted hydrogen passed into the solution at the temperature of about 86 F., to separate ar- senic, copper, and lead. The iron in the liquid is then again oxidized either by chlorine or by hot nitric acid, tartaric acid is added, the fluid is saturated with ammonia, when everything remains in solution. Bicarbonate of ammonia well saturated with carbonic acid is then added, the solution is next submitted again rapidly to the action of sulphuretted hydrogen as long as a precipitate is formed ; sulphides of zinc, iron, nickel, sometimes cobalt, are separated, whilst the oxide of uranium remains in solution. These solutions are washed with a dilute solution of bicarbonate of ammonia saturated with carbonic acid, and con- taining sulphuretted hydrogen. During the passage of the sulphuretted hydrogen through the tartaric acid fluid, care must betaken to maintain an excess of carbonic acid, which is liable to be displaced by sulphuretted hydrogen, and which prevents the oxide of uranium from be- coming converted into sulphide, and the other metallic sul- phides from forming green sulphur-salts, and passing through the filter. This is easily done by adding some fragments of marble to the sulphide of iron employed in furnishing the sul- phuretted hydrogen. To extract uranium from pitchblende, Liebig directs to heat the mineral to redness, then reduce it to an impalpable powder, and digest with nitric acid diluted with four parts of water, taking care to employ a larger quantity of mineral than the acid can dissolve. By this process, the protoxide of uranium is converted into sesquioxide, which unites to the nitric acid almost to the total exclusion of the iron. Sulphuretted hy- drogen is then transmitted through the solution, by which copper, lead, etc., are separated. The solution is boiled to expel any free acid, and after being concentrated by evaporation, is set aside to crystallize. Nitrate of uranium crystallizes in lemon-coloured "flattened prisms. 171. MANGANESE. This metal is quantitatively estimated as protoxide, as red oxide, or as prot&sulphate, and as sulphide. MANGANESE. 333 Quantitative estimation as Protoxide. Ebelmen prefers re- ducing the higher oxides of manganese to the state of protoxide for the purpose of weighing. To perform the experiment, he introduces the oxide into a small platinum crucible, the lid of which has a small platinum tube in the centre, through which a current of dry hydrogen gas is conveyed by means of a glass tube, of a diameter nearly equal to that in the crucible cover ; heat is applied by means of a spirit lamp ; the reduction is complete after a few minutes, and a rapid current of gas is passed into the crucible while cooling ; the oxide obtained is quite pure, it dissolves in hydrochloric acid without rendering it black, and without disengaging chlorine. Its composition is One equivalent of Mn . . . 27'5 . . 77-5 One ditto of O ... 8*0 .. 22 '5 One ditto of MnO . . 355 . . Quantitative estimation as Red Oxide. The solution is heated to boiling with excess of carbonate of soda. Should ammoniacal salts be present, the solution must be treated in the same man- ner as in the precipitation of carbonate of zinc, under similar circumstances ; the precipitate is thoroughly washed, dried, and ignited, by which it is decomposed into the red oxide (manganoso-manganic oxide) ; sometimes the solution is pre- cipitated with caustic potassa, and the hydrated protoxide thus formed is converted into red oxide by strong ignition ; it must be remembered, however, that hydrated protoxide of manga- nese is soluble in sal-ammoniac. It is sometimes convenient also to precipitate manganese as sulphide by adding coloured sulphide of ammonium to the solution to which sal-ammoniac and ammonia have been added ; the precipitate, having been washed with water containing sulphide of ammonium, is digested with hydrochloric acid, and the solution precipitated by carbo- nate of ammonia. The composition of red oxide of manganese is Three equivalents of Mn . 82'5 . . 72*05 Four ditto of O . . 32-0 . . 27*95 One ditto of MnO + Mn 2 3 1M'5 . . 10OOO Quantitative estimation as Sulphate. Oxide of manganese is converted into protosulphate in the same manner and with the same precautions as magnesia ; the salt must be ignited very 334 QUANTITATIVE ANALYSIS. feebly, or it will lose sulphuric acid. Its composition is One equivalent of MnO. . 35'5 . . 47-20 One ditto of S0 3 . . 4Q-Q . . J52^80 One ditto ofMnO,S0 3 ^5 . . 10CH)0 Quantitative estimation as Sulphide. In the absence of am- moniacal salts, manganese is completely precipitated from its solutions by sulphide of ammonium. The dried sulphide is introduced into a crucible mixed with a little sulphur. The crucible is provided with a perforated cover, through which a current of dry hydrogen gas is passed. When the crucible is full of hydrogen, heat is applied for some time, the crucible is then allowed to cool whilst the gas continues to pass, at the bottom of the crucible a mass of sulphide (MnS) is found, which, when not much heated, is green and black, like the native sulphide when red-hot. The sulphate, and all the oxides of manganese, can, according to Rose, be thus converted into a sulphide of invariable composition. Separation of Oxide of Manganese from the Oxides of the Fourth, Third and Second Groups. (a.) From Oxides of Cobalt and Nickel. Rose's original method, which, though complicated, gives very accurate results, is the following : The metals are first precipitated together as oxides by caustic potassa, they are then ignited, weighed, and converted into chlorides by introducing them into a bulbed tube, and transmitting over them a current of dry hydrochloric acid gas, a moderate heat being at the same time applied ; it requires a long time thus to convert the oxides completely into chlorides ; when the conversion is complete, dry hydrogen gas is passed through the apparatus, the bulb containing the chlorides being strongly heated, the hydrogen displaces the chlorine from the chlorides of cobalt and nickel, forming with it hydrochlo- ric acid gas ; and the operation must be continued as long as white clouds are formed, on holding a glass rod that has been dipped in ammonia at the end of the apparatus where the ex- cess of gas escapes. The hydrogen is allowed to pass through the tube till it is quite cold, the chloride of manganese, which has not been decomposed by the process, is dissolved out by water, the reduced cobalt and nickel are washed with very di- lute hydrochloric acid, and, finally, with water ; the two metals are then separated from each other by one of the methods above given, and the manganese is precipitated from the solution of MANGANESE. 335 the chloride, to which have been added the washings from the cobalt and nickel by carbonate of soda. Rose states that nickel may be conveniently separated from manganese in the same manner as cobalt, namely, by chlorine and carbonate of baryta. A method of separating oxide of cobalt from oxide of manganese was proposed by Barreswil ; it consists in adding carbonate of baryta to the solution, and then passing a current of sulphuretted hydrogen through it, by which, according to Barreswil, cobalt only is precipitated: this method has not, however, been found successful by other chemists. A plan for the separation of oxides of cobalt and nickel from oxide of manganese was proposed by "Wakenroder, and has obtained the approval of Rose ; it is based on the fact that, although nickel and cobalt are not precipitated from their acid solutions by sulphuretted hydrogen, their sulphides precipitated by sul- phide of ammonium are not dissolved by very dilute hydro- chloric acid. The acid solution of the three oxides is made am- moniacal, and the metals are precipitated by sulphide of am- monium, the solution is then rendered slightly acid by dilute hydrochloric acid ; the sulphide of manganese is dissolved with facility, small portions also of the sulphides of cobalt and nickel may also be dissolved, but they may be removed by pre- cipitating the solution with ammonia and sulphide of ammo- nium, and treating it anew with dilute hydrochloric acid. From Cobalt (Liebig's method). The metals are precipitated by cyanide of potassium, an excess of which redissolves the cyanide of cobalt, and a part also of the protocyanide of man- ganese, while another portion remains undissolved ; this is fil- tered off and washed ; the filtrate is heated to ebullition, a few drops of hydrochloric acid, insufficient, however, to acidify the solution, are added from time to time, and the manganese and cobalt are separated from one another in the same manner as nickel is separated from cobalt. Ebelmen has recently proposed to separate manganese from cobalt or nickel, by exposing the mixture of the oxides in a platinum or porcelain tray to a current of dry sulphuretted hydrogen, at a temperature a little below redness. The oxides are converted into sulphides, the sulphide of manganese is dis- solved by very weak hydrochloric acid, in the cold, and after filtration is precipitated at a boiling heat by potassa, and the sulphide of cobalt remaining on the filter is redissolved in nitric acid, and precipitated by caustic potassa. The smallest quantity 336 QUANTITATIVE ANALYSIS. of cobalt may, lie says, be detected in this manner. He applies the same process to the separation of arsenic from tin. From Iron and Nickel (Schiel). A stream of chlorine is passed through the solution to which acetate of soda has been added, peroxide of manganese is precipitated, the nickel and iron re- maining in solution. If cobalt be present, it is separated as oxide. (b.) From Oxide of Zinc. The oxides are dissolved in excess of acetic acid, and the zinc is precipitated by sulphuretted hy- drogen ; the solution must be acid, and contain no other acid but acetic. Another method (Flajolot). The excess of acid in the so- lution is neutralized by carbonate of soda ; an excess of cyanide of potassium is then added, and afterwards carbonate of soda. At the temperature of ebullition, carbonate of manganese is precipitated alone. (c.) From Oxide of Chromium. This is effected in the same manner as the separation of oxide of zinc from oxide of chro- mium. (d.) From Alumina (Rose) . Add chloride of ammonium, heat to ebullition ; pour in caustic ammonia and boil until the excess of the latter is disengaged ; the alumina will then be com- pletely precipitated and perfectly free from manganese, even when operating with free access of air. (e.) From Magnesia and Lime. Add a solution of acetate of soda, heat, and pass a current of chlorine through the liquid, when permanganate will be formed. Supersaturate with am- monia and boil, upon which all the manganese is reduced to sesquioxide and precipitated, the magnesia and lime remaining in solution. If the quantity of magnesia be large, a certain quantity of chloride of ammonium must first be added. Determination of the value of Commercial Binoxide of Man- ganese : (1.) By estimating the amount of chlorine evolved during the solution of the ore in hydrochloric acid ; the value of the oxide being in exact proportion to the quantity of chlorine produced. Three methods of estimating the chlorine have been employed. (a.) By noting the quantity of Protosulphate of Iron ivhich it peroxidizes. If the oxide of manganese be perfectly pure, 4*35 parts (one equivalent) will produce 35*5 parts (one equivalent) of chlorine, which will peroxidize 278 parts (two equivalents) of crystallized protosulphate of iron. MANGANESE. 337 Hence 25 grains of pure oxide of manganese yield chlorine sufficient to peroxidize 159 grains of protosulphate of iron ; 25 grains of the powdered specimen are weighed out, and a quantity not less than 359 grains of crystallized protosulphate of iron. The oxide of manganese is thrown into a flask con- taining about one ounce of strong hydrochloric acid slightly diluted, and a gentle heat applied. The protosulphate of iron is gradually added in small quantities to the acid, so as to ab- sorb the chlorine as it is evolved, and the addition of that salt continued till the liquor, after being heated, gives a Hue pre- cipitate with ferridcyanide of potassium, and has no smell of chlorine, which are indications that the protosulphate of iron is present in excess ; by weighing what remains of this salt, the quantity that has been added is ascertained, say m grains. If the whole of the specimen consisted of peroxide, it would require 159 grains of protosulphate of iron, and that quantity would therefore indicate 100 per cent, of peroxide ; but, if a por- tion of the manganese only is peroxide, a proportionally smaller quantity of the protosulphate will be consumed, and that quan- tity will give the real proportion of peroxide, by the proportion, as 159 : 25 : : m to the quantity required. (b.) The chlorine evolved is passed through water in which lime is diffused ; chloride of lime is formed; a certain quantity of protosulphate of iron is dissolved in water, and the solution of chloride of lime is added thereto, until the iron liquor ceases to strike a blue colour with a drop of solution of red prussiate of potassa : then, comparing the quantity of the solution of chloride of lime required, with the quantity that was produced, the quantity of chlorine generated, and hence the total quan- tity of available oxygen, is known. (c.) Baumann conveys the chlorine into a solution of nitrate of silver, and calculates the amount of real peroxide of man- ganese in the specimen from the quantity of chloride of silver formed. (2.) By converting Oxalic Acid into Carbonic Acid, ~by means of the second atom of Oxygen which the Peroxide of Manganese con- tains. Mn 2 + C 2 O 3 , producing Mn 0, and 2 C0 3 . 100 grains of the specimen are introduced into a weighed flask, and 150 grains of oxalic acid dissolved in 500 grains of water are poured upon it ; to this, 350 grains of oil of vitriol are to be added, and the orifice of the flask closed by a cork, through which passes a tube containing fragments of recently fused chloride of cal- 338 QUANTITATIVE ANALYSIS. cium ; the weight of this cork and tube are to be included in the tare of the flask. On the addition of the oil of vitriol, a brisk effervescence takes place, owing to the escape of carbonic acid, which passing over the fragments of chloride of calcium in the tube is dried, so that the gas alone passes off. "When the action slackens, a gentle heat may be applied, until all the oxide of manganese has dissolved ; the small quantity of a light brownish sediment, which generally falls, is easily distinguished from the particles of black oxide. As soon as the action is quite over, the flask is allowed to cool, and as it still contains a quantity of carbonic acid gas, this is removed by taking out the cork, and blowing air gently into the flask by a glass tube; the cork is then to be replaced, and the flask with its contents weighed ; the difference of weight represents the amount of carbonic acid evolved, one-fourth of the oxygen of which had been derived from the peroxide of manganese by its conversion into pro- toxide, which remains combined with sulphuric acid in the liquor, and the quantity of peroxide in the 100 grains of the one are thus directly found. Example. Suppose the flask and the materials together to weigh 1876 grains, and after the ac- tion has terminated 1816'5 grains : the loss, 59'5 grains, is car- bonic acid, consisting of 16'3 grains of carbon, and 43'2 grains of oxygen ; the oxygen derived from the mineral is therefore A O.O = 10 - 8, which represents 59 grains of pure peroxide of manganese in 100 grains of the substance experimented upon. (3.) Method of Fresenius and Will. This is founded on the same principle. A certain weighed quantity of the finely pow- dered manganese ore is projected into B (see Fig. 69,p. 268), and about 2i parts of neutral oxalate ofpotassa, prepared by saturat- ing the common binoxal ate with carbonate of potassa and crystal- lizing, or two parts of neutral oxalate of soda, and as much water added as will fill about one-third of the flask ; the apparatus is then prepared as directed in p. 269, B is then closed, the appara- tus weighed, and, by sucking the tube e, some sulphuric acid is made to pass from a into . The evolution of carbonic acid com- mences immediately, and in a very uniform manner : as soon as it stops, some more sulphuric acid is sucked over from a into b, and the operation thus continued till all the manganese is com- pletely decomposed: this operation will require from six to ten minutes: its completion is ascertained not only by no further evolutions of carbonic acid taking place upon a further influx MANGANESE. 339 of sulphuric acid into b, but also by no black powder remaining at the bottom of the flask. Some more sulphuric acid is then sucked over into b, in order to heat the fluid contained therein, so as completely to expel all the carbonic acid evolved during the course of the operation ; the wax stopper is then removed from a, and the tube d sucked until the air no longer tastes of carbonic acid ; the apparatus is then allowed to cool, and weighed. The entire examination may thus, according to the authors, be performed in a quarter of an hour ; from the loss of weight of the apparatus (the amount of carbonic acid expelled) the amount of available oxygen is found, or, what in fact is the same, the amount of peroxide contained in the manganese under examination, according to the following arrangement. Two equivalents of carbonic acid stand to one equivalent of peroxide of manganese in the same proportion as the amount of carbonic acid found stands to #, x being the quantity of real peroxide contained in the specimen. Let us suppose that our experiment had been made with four grammes (61'76 grs.) of manganese ore, and that we had obtained 3 '5 grammes (54-03 grs.) of carbonic acid. The arrangement would be 4 : 43-67 : : 3'5 : x x = 3-47. Thus, four grammes of the manganese ore containing 3 '47 grammes of peroxide, 100 parts of the same substance must contain 86*7 parts. To render, however, this calculation un- necessary, we need only ascertain what amount of manganese must be taken to make the number of centigrammes of carbonic acid obtained immediately indicate the percentage amount of peroxide. The calculation must therefore be 4 : 43-67 : : 100 : x x =. 0-993. Thus, if we take 0-993 grammes (15'33 grains) of the speci- men, the number of centigrammes of carbonic acid expelled will indicate the percentage amount of peroxide. But as the quantity of carbonic acid obtained would be too minute to ad- mit of a direct determination of its weight, it. is advisable to take a multiple of 0'993 grammes, and to divide the number of centigrammes of carbonic acid obtained by the same number, which has served as a multiplier. The multiple by 3, i.e. 2'9S grammes, is deemed by the authors the quantity "best adapted for the examination. Should the manganese contain carbonated alkaline earths, which is sometimes the case, it must undergo a 34-0 QUANTITATIVE ANALYSIS. preliminary process previous to the examination. To ascertain the presence or absence of carbonate of lime or baryta in the manganese under examination, it is sufficient to moisten a sam- ple powder with dilute nitric acid ; their presence is certain if any effervescence take place. The specimen, in that case, is treated as follows : 2 '98 grammes of the specimen are projected into b, covered with very dilute nitric acid, one part acid to twenty parts of water, and allowed to stand at rest for a few minutes ; the supernatant fluid is then poured upon a small paper filter, the manganese remaining in the flask is repeatedly washed with water, as well as the solid particles on the iilter, (the supernatant water being always poured on the filter), and the latter then thrown into 5, taking care not to lose a particle of manganese ; the further operation is conducted as usual. The neutral oxalate of potassa is preferable to free oxalic acid, or to binoxalate of potassa, since, when employing the former substance, the evolution of carbonic acid commences only upon the influx of sulphuric acid into b, whilst free oxalic acid, or binoxalate of potassa, begin to evolve carbonic acid immediately upon coming into contact with manganese and water ; and this would, of course, interfere with the correctness of the results, rendering it almost impossible to determine the exact weight of the apparatus before the commencement of the operation. Such is the method of examining ores of manganese proposed by the German chemists : the results are most accurate, and the mani- pulation is so simple, that it will no doubt entirely supersede all the other methods that have been described. It will be ob- served that, from the almost absolute identity of the equiva- lent number of carbonic acid multiplied by 2, 22 x 2 = 44, with that of peroxide of manganese, 43*67, the calculation of the quantity of real peroxide in a specimen of manganese be- comes a problem of the utmost simplicity ; and that, if we take 100 grains of the specimen, the loss of weight in grains will denote the percentage proportion of pure peroxide. One atom of peroxide of manganese, Mn0 2 =r43'67, contains one atom of oxygen, separable by sulphuric acid, and capable of converting one atom of oxalic acid into two atoms of carbonic acid. Thus : Oxalic Acid, C 2 3 + = 2 C0 2 Ditto . . 3(5 +8 =44 Dr. Ure recommends 250 grains of oxalate of potassa to 100 grains of the sample. MANGANESE. 341 Various other methods of ascertaining the value of commer- cial oxide of manganese have been proposed ; but as none of them exceed in accuracy or convenience the method of Will and Eresenius just described, the principles of the processes alone will be briefly described. (4.) Method founded on the reducing action of Protochloride of Tin (L. Muller, Ann. der Chem. und Pharm., Oct. 1851, and Chem. Gaz., vol. x. p. 75). Protochloride of tin and perchloride of iron react mutually on each other, thus SnCl + Fe 3 Cl 3 =SnCl 2 4-2FeCl. In order to derive advantage from this reaction, two courses might be followed: first, a solution of protochloride of iron might be converted into perchloride by the action of the chlorine developed from hydrochloric acid by the oxide of manganese to be valued, and then the quantity of perchloride of iron thus formed in this solution, determined by means of a solution containing a known proportion of protochloride of tin ; or secondly, the chlorine evolved may be passed through a solution of protochloride of tin, and the perchloride pro- duced estimated by means of a solution of perchloride of iron of known strength. The author finds the second method to be perfectly applicable in practice, as the chlorine is rapidly and completely absorbed by a solution of protochloride of tin. (5.) Method founded on the conversion of Arsenious Acid into Arsenic Acid by means of Chlorine and Permanganate of Potash (Astley Price, Chem. G-az., vol. ii. p. 416). Ten or any number of grains of the specimen under ex- amination are placed in a small flask, to which ten or more measures of a normal solution of arsenious acid in hydrochloric acid are added, and to the flask is adapted a Will's nitrogen apparatus, containing a solution of potassa, to retain any terchloride of arsenic which may be given off. The flask is then placed in a water-bath, or a gentle heat is applied until solution is effected. The contents of the flask, after being allowed to cool, are, together with the solution of potassa, transferred to a larger flask and diluted with water. The amount of arsenious acid remaining unchanged, is then deter- mined by the addition of a standard solution of permanganate of potassa, and the quantity thus indicated being deducted from the number of grains of arsenious acid employed in the first instance, will give the value of the specimen under analysis. PART IT. 2 A 342 QUANTITATIVE ANALYSIS. (6.) Method founded on the conversion of Copper into Bichlo- ride by the Chlorine evolved from Hydrochloric Acid l)y Oxide of Manganese (Nolte, Chem. Graz., vol. xvii. p. 296). Every atom of peroxide in the ore sets free one atom of chlorine from hydrochloric acid, which in its turn converts two atoms of copper into dichloride. Strong chemically pure hy- drochloric acid is poured over a known quantity of the native peroxide in a retort, and an excess of copper is added, the evo^ lution of free chlorine being avoided by keeping the retort as cool as possible until the ore is completely decomposed. The mixture is then heated to the boiling-point, so as to convert the protochloride at first formed into dichloride. The loss of copper in weight, gives the amount of peroxide in the ore, since every two atoms of copper dissolved answer to one of peroxide. GROUP V.-Section A. Lead, Thallium, Silver, Mercury, Bismuth, Cadmium, Copper, Palladium, Rhodium, Osmium, Ruthenium, Iridium. 172. LEAD. This metal is completely precipitated from its acid solutions by sulphuretted hydrogen ; it is also precipitated by sulphuric acid, by carbonate of ammonia, by oxalate of ammonia, by hydro- chloric acid, and by bichromate of potassa ; and it may therefore be weighed as sulphide, sulphate, chloride, chromate, and as oxide, the latter being produced by igniting with free access of air, the carbonate and oxalate. Lead is likewise estimated by volumetric processes. Precipitation as Sulphide. A stream of washed sulphuretted hydrogen gas is transmitted through the solution, rendered slightly acid with nitric acid until it is completely saturated, a gentle heat is then applied, and the precipitated sulphide is filtered as quickly, and with as little access of air, as possible, in order to avoid the decomposition of a portion of the sulphu- retted hydrogen, which would occasion a precipitation of sul- phur ; as it is difficult, even with great care, to prevent this, it is better to convert the washed sulphide of lead into sulphate, which is done by transferring it, filter and all, into a beaker, and pouring on it concentrated and fuming nitric acid, the LEAD. '343 sulphur becomes hereby oxidized into sulphuric acid ; a gentle heat is applied to assist the action ; the mixture is carefully transferred into a small Berlin crucible, and a few drops of sulphuric acid being added, it is evaporated to dryness and ignited. The composition of sulphate of lead is One equivalent of PbO . . 111-5 . . 73-60 One ditto of SO 3 ."' V . . 4Q-Q . . 26'40 One ditto of PbO, S0 3 . . 151-5 . . lOO'OO With respect to the precipitation of lead by sulphuretted hydrogen, the following observations have been made by Dr. Vogel. In precipitating lead with sulphuretted hydrogen it is usual to warm the liquid, to prevent any of the precipitate from passing through the filter, but this may give rise to con- siderable loss. If sulphuretted hydrogen be passed through a moderately strong solution of acetate of lead, until the filtered liquor contains no more metal, and is consequently not rendered turbid either by sulphuretted hydrogen or by sulphuric acid, and if the liquor be now warmed, a further precipitate is pro- duced by sulphuretted hydrogen after filtration ; the liberated acetic acid evidently decomposes the sulphide of lead. The same is the case with nitrate of lead, and remarkably so with the chloride. This is of considerable importance where lead has to be separated from chlorides, and where it cannot be pre- cipitated by sulphuric acid. "With mercury and bismuth no such decomposition of the sulphide occurs, and only in a slight degree with antimony. The above behaviour may even be em- ployed for the separation of some metals. If, for example, sulphuretted hydrogen be passed into a liquid containing nitrate of bismuth and nitrate of lead, until no further precipi- tate occurs, and the whole be then heated to boiling, the sul- phide of lead is entirely redissolved, while the sulphide of bismuth is not attacked. The lead may then be precipitated from the solution by sulphuric acid. According to E/ose, lead may be quantitatively estimated as sulphide by heating the washed precipitate by sulphuretted hydrogen in contact with sulphur in a crucible through which a current of sulphuretted hydrogen is made to pass. The heat should be a full red, and the sulphide (PbS) should be perfectly crystalline. Volumetric estimation of Lead by Sulphide of Sodium. A method of estimating lead by means of a normal solution of sulphide of sodium was proposed by Domonte, and is de- 2 A2 344* QUANTITATIVE ANALYSIS. scribed by him as simple, quick, and accurate, and well adapted for the examination of ivhite lead and acetate of lead, sub- stances which, as met with in commerce, are often greatly adulterated. To prepare the normal solution of sulphide of sodium, about 300 grains are dissolved in a pound of water, a certain known quantity of lead is dissolved in nitric acid, pre- cipitated, and redissolved by caustic potassa ; the solution of sulphide of sodium placed in a graduated burette, is added care- fully to the alkaline solution, which is maintained at a tempe- rature near to ebullition. Each addition of the sulphide pro- duces a black precipitate of sulphide of lead; from time to time the liquid is boiled, and the precise point at which a drop of the reagent produces no further precipitate is carefully ob- served. The number of divisions of the burette which have been required represents the value of the lead employed in the experiment : thus, suppose 10 grains of lead to have been used, and 30 burette divisions of the alkaline sulphide have been required, then, in all analyses of commercial products, for every 30 divisions of the burette that are required the ope- rator may infer that 10 grains of lead are present. The author does not find the success of the experiment to be interfered with by the presence of tin, antimony, arsenic, iron, cobalt, nickel, or zinc. The first three metals are not precipitated by an alkaline sulphide ; in the presence of excess of free alkali, zinc is precipitated after the lead, but the co- lour of the sulphide being white, renders its presence rather an advantage than otherwise ; copper complicates the process. It is necessary in the first place to determine this metal by one of the methods to be described hereafter. Domonte then makes a synthetic essay on a mixture formed of a weight of copper equal to that found by experiment, and of lead equal in weight to that employed in preparing the standard solu- tion. This assay shows by how many divisions the plumbi- metric liquor ought to be diminished, on examining the alloy. The number is, in fact, the difference between the results of the assay of pure lead, and those of the assay of the mixture of lead and copper. This being done, he examines the alloy in the ordinary way. Bismuth cannot be separated from lead in this way, but, commercially speaking, it is not likely that this metal will be found to occur as a contamination, its higher price being a guarantee. Volumetric estimation of Lead fy Bichromate of Potassa LEAD. 345 (Schwarz; Dingler's ' Poly tech nisch 'Journal'). A standard solution of bichromate is formed by dissolving 14*730 grammes of the pure salt in sufficient water to form one litre. One cubic centimetre of this solution precipitates O0207 gramme of lead. The lead to be analysed is dissolved in the minimum of nitric acid, the solution diluted with water, carefully neu- tralized with carbonate of soda or with ammonia, an excess of acetate of soda added, and the solution precipitated by the bichromate of potassa solution. When the precipitation ap- proaches its end, or when the precipitate commences readily to subside, some drops of a neutral solution of nitrate of silver are deposited on a porcelain plate, and the chromate solution only, added by two or three drops at a time, to the liquid under examination; after each addition, the whole is well stirred, allowed to subside, and a drop of the clear supernatant liquor added to one of the drops of the . silver solution. As soon as the bichromate is in excess, a red colour is produced, while the precipitated chromate of lead has no effect on the silver test, but simply floats on the top as a yellow precipi- tate. It is only now necessary to read off the number of divisions of the burette that have been required, to learn the amount of lead in the solution. The operator must be careful to add abundance of acetate of soda. The only foreign metal which, according to the author, interferes with the reaction, is bismuth, in the presence of which it is inapplicable. Quantitative estimation as Sulphate. When lead is to be weighed in the form of this salt, the solution is precipitated by dilute sulphuric acid, and then mixed with twice its volume of alcohol, sulphate of lead not being altogether insoluble in water ; it is collected on a filter, on which it is washed with spirits of wine. Its composition has been given above. Accord- ing to Vogel, sulphate of lead precipitated from solutions con- taining nitric acid, is never free from that acid, and when lead is thrown down by sulphuretted hydrogen from compounds con- taining zinc, the sulphide of lead always contains sulphide of zinc. Quantitative estimation as Oxide : (a.) Precipitation as Carbonate. Carbonate of ammonia, mixed with a little caustic ammonia, is the precipitant em- ployed, and the solution is heated ; it is washed on the filter with pure water, and afterwards ignited in a porcelain cruci- ble, by which it is converted into protoxide of lead. The pre- 346 QUANTITATIVE ANALYSIS. cipitate is removed as completely as possible from the filter and the latter is ignited alone, the residue being afterwards mixed with the ignited precipitate ; the object of this is to pre- vent the reduction of a portion of the protoxide of lead by the organic matter of the filter. The same observation applies to the sulphate. (b.) Precipitation as Oxalate. Hose prefers occalate of am- monia as the precipitant of oxide of lead from its solutions, which must be either neutral or weakly ammoniacal. The pre- cipitated oxalate of lead is converted into protoxide by ignition in an open porcelain crucible, having, as in the two former cases, been removed as much as possible from the filter, the latter being ignited alone. The composition of protoxide of lead is One equivalent of Pb . . . 103-50 . . 92'83 One ditto of ... 8-QO . . 7*17 One ditto of PbO . . 111-50 . . 100*00 Quantitative estimation as Chloride. The solution is preci- pitated by excess of hydrochloric acid, concentrated by eva- poration on the water-bath, and the residue washed with abso- lute alcohol, mixed with ether. As chloride of lead is vola- tilizable, it must not be ignited, but dried at a temperature below redness. Its composition is One equivalent of Pb . . . 103-50 . . 74*46 One ditto of Cl . . . 35-50 . . 25-54 One ditto ofPbCl . . 139-00 . . 100-00 Estimation of Lead in small quantities in the presence of other metals^ by precipitation as Chromate. There are fre- quent instances in which the chromate of potassa offers con- siderable advantages as a precipitant of lead, and the effi- cacy of this reagent is increased by the great insolubility of chromate of lead, particularly in acetic acid, it may therefore be employed to precipitate this metal in the presence of copper and zinc. The chromate of lead precipitated should be dis- solved in a little hot dilute hydrochloric acid ; a small crystal of tartaric acid added, and the solution, rendered alkaline by ammonia, should be treated with hydrosulphuric acid, or mixed with sulphide of ammonium. The sulphide of lead thus obtained is washed, and converted into sulphate by the usual method* SILVER. 347 Analysis of Galena. Treat the pulverized mineral with concentrated nitric acid ; evaporate to dryness with a small ex- cess of sulphuric acid ; treat the dry mass with water, which dissolves out all the sulphates but that of lead. Tin, anti- mony, quartz, and sulphate of baryta would, if present, be left in the insoluble portion. Collect and weigh the insoluble portion ; digest it repeatedly with acetate of ammonia (sp. gr. 1'065), wash, dry, and weigh. The difference indicates the amount of sulphide of lead. The lead may be precipitated from its solution in acetate of ammonia by sulphide of ammo- nium, and the sulphide of lead may be oxidized into sulphate by nitric and sulphuric acids. From all the metals that have hitherto been treated of, lead may be separated by sulphuretted hydrogen ; the solution should be acidified by nitric acid, and considerably diluted. Separation of Sulphate of Lead from Sulphate of Baryta. Hyposulphite of soda dissolves sulphate of lead. To effect the separation, a concentrated solution of the hyposulphite is added to a mixture of the two salts, and the whole gently warmed, taking care that the temperature does not exceed 68 ; at a higher temperature sulphite of lead is formed, which is insoluble in the hyposulphite. 173. PROTOXIDE or THALLIUM.* General Characters. This oxide may be prepared by allow- ing granulated thallium to oxidize in warm, moist air, and then boiling in distilled water. By repeating this operation two or three times, a saturated hot solution of the oxide is formed. Upon filtering, the small quantity of carbonate which may have formed, separates at first in white needles, whilst upon further cooling the oxide crystallizes out in yellow needles. Anhydrous protoxide of thallium is formed by exposing these yellow crystals in a vacuum over sulphuric acid ; it then forms a reddish-black mass, retaining the shape of the crystals. When heated to about the melting-point of the metal, it melts to a brown, limpid liquid, which at a high temperature evolves reddish-black vapours, partially oxidizing at the same time to the peroxide. Upon cooling, the browa liquid solidifies to an * For the information contained in this section we are entirely indebted to the able and exhaustive memoir on thallium, published in the ' Journal of the Chemical Society' (Series 2, vol. ii., April, 1864), by William Crookes, F.R.S., the discoverer of this singular element. 348 QUANTITATIVE ANALYSIS. almost black crystalline mass. The fused oxide attacks glass and porcelain, removing silica. When heated it always per- oxidizes slightly. Oxide of thallium is decomposed by hydro- gen at a red-heat, forming water and metallic thallium. The decomposition is however never perfect, owing to the oxide fusing and volatilizing. When fused with sulphur it forms sulphide of thallium ; and in aqueous solution with metallic zinc, the metal is precipitated, and oxide of zinc formed ; when an electric current is passed through a solution of the oxide, it is a^so reduced to the metallic state. The best method of preparing perfectly pure hydrated oxide of thallium, is to add water to the oily compound of oxide of thallium and alcohol. This at once separates the oxide in the form of a bright yellow crystalline mass, which may be separated from alcohol and water, by exposure to warm, dry air. It is a powerful base, dissolving readily in water, and forming a colourless, strongly alkaline solution. It has a slight odour, similar to that of potassa, dissolves the skin, and feels greasy. It has a strong action on the hair and nails, staining them a deep and very permanent brown colour. It blues red litmus-paper, browns turmeric-paper, has a metallic alkaline taste, and neutralizes acids completely. It eliminates ammonia from chloride of ammonium, and reacts with hydro- chloric acid, iodide of potassium, sulphide of ammonium, etc., in the characteristic manner of a thallium salt. An aqueous solution of protoxide of thallium has a greater similarity to potassa than to ammonia in its reactions with metallic salts. When added to solutions of magnesium, cerium, manganese, zinc, cadmium, lead, iron, cobalt, nickel, copper, mercury, silver, and peroxide of thallium, it precipitates the respective oxides, without redissolving them in excess. From salts of aluminum and chromium it precipitates the hydrated oxides, and easily re- dissolves them when in excess, forming with alumina a solution unaltered by boiling, but precipitated by a current of carbonic acid, and with chromium a green solution precipitated on boil- ing. The salts of thallium are highly poisonous, 3 or 4 grains of the sulphate being sufficient to kill a small animal; the symptoms are somewhat similar to those produced by lead. When ignited, salts of thallium generally fuse at temperatures below redness, and then volatilize without change. On charcoal before the blow-pipe, they volatilize, communicating an intense green colour to name. PROTOXIDE OF THALLIUM. 349 Comportment of Salts of Protoxide of Thallium with re- agents : Hydro sulphuric Acid produces no effect when added to a solution of sulphate or nitrate with excess of acid ; if the solution be neutral, a small portion of the metal is precipitated. In solutions of oxide of thallium, combined with a weak acid such as acetic, this reagent throws down the whole of the metal in the form of a deep brown sulphide. Sulphide of Ammonium precipitates thallium salts completely, the precipitated sulphide being insoluble in sulphide of ammo- nium, in caustic alkalies, their carbonates and cyanides, and only slightly soluble in acetic acid. Hydrochloric Acid and soluble Chlorides precipitate a diffi- cultly soluble white chloride. Hydrobromic Acid and soluble Bromides precipitate a white, nearly insoluble bromide. Hydriodic Acid and soluble Iodides precipitate an insoluble yellow iodide. Alkalies, Alkaline Carbonates or Eicarbonates produce no change in thallium protosalts. Phosphate of Soda gives a white precipitate, nearly insoluble in ammonia, easily soluble in acids. Chromate of Potassa gives a yellow precipitate, insoluble in cold nitric or sulphuric acid, but turned orange-red on boiling in the acid solution. Bichloride of Platinum precipitates a very pale yellow, in- soluble double salt. Phosphuretted Hydrogen precipitates a black phosphide. From aqueous solutions of salts of protoxide of thallium the metal is precipitated in metallic crystals by zinc, and slowly by iron. Peroxide of thallium (T10 3 ) is always formed when metallic thallium is heated, or even when a solution of the protoxide is evaporated in the air : when the metal is burnt in oxygen the product is chiefly peroxide. It is best prepared by adding potassa, ammonia, or protoxide of thallium to a solution of the peroxide, washing the precipitate formed, and drying at a temperature of 500 E. Anhydrous peroxide of thallium is a dark-brown powder, fusing with difficulty, and evolving oxygen at a red-heat, be- coming reduced to protoxide. It is neutral to test-paper, and insoluble in water. It dissolves readily in sulphuric, nitric, 350 QUANTITATIVE ANALYSIS. and hydrochloric acids, forming hygrometric and insoluble salts. The hydrated peroxide (T10 3 HO) is obtained by drying the precipitated peroxide at a temperature of 212 P. ; it forms a brown powder, a shade lighter than the anhydrous oxide. Properties of Thallium. The most characteristic property of thallium is the intense green colour which the metal, or any of its compounds, communicates to a colourless flame. When examined in the spectroscope this colour is seen to be absolutely monochromatic, appearing as one intensely brilliant and sharp green line. The delicacy of the spectral reaction of this metal is so great, that Mr. Crookes was able to produce it with the five-millionth of a grain of the sulphate. The specific gravity of thallium varies according to the treat- ment it has undergone. A lump, melted, and slowly cooled under cyanide of potassium, was found by Mr. Crookes to be 11*81 ; after being strongly compressed it became H'88. De la Rive states it to be 11'85 after being melted, and 11'86 after being drawn into wire. Thallium is the softest metal known, admitting of free ex- posure to the atmosphere. The finger-nail, or even a piece of lead, scratch it readily. It marks paper like plumbago, form- ing a streak, grey at first, then turning yellow, and in a day or two fading nearly out. Sulphide of ammonium or sulphu- retted hydrogen will at any time temporarily restore the dark streak. Thallium has less tenacity than lead, and does not become brittle at any temperature between F. and its melt- ing-point, which Mr. Crookes places at 561 F., and M. Lamy at 554 F. It is very malleable, and can be hammered into foil as thin as tissue-paper. It can be drawn into wire only with difficulty ; in this state it is almost devoid of elasticity, retaining any form into which it is bent, with scarcely a ten- dency to spring to its original position. Thallium is a very crystalline metal, and crackles almost as much as tin when bent. When several pounds of the metal are fused and allowed to cool slowly, and the interior liquid portion poured off from that which has solidified, well- defined crystals, in octahedrons and fern-like forms are pro- duced. When heated in the air, thallium begins to volatilize at a red-heat, evolving brown vapours of oxide ; it boils below a white-heat, and may be distilled in a current of hydrogen : when heated to redness, and plunged into oxygen, it burns brilliantly. SILVER. 351 The atomic weight of thallium has been found by Lamy to be 204. The results of five experiments gave Crookes 202'96, a number which he thinks may be somewhat too high. In electro-chemical position thallium is very near cadmium, being precipitated from the sulphate by zinc and iro*n, but not by cadmium, tin, or copper. The position of thallium amongst the elementary bodies has given rise to considerable discussion. On the Continent it is generally classed amongst alkali metals ; in England it is, on the other hand, generally regarded as belonging to the silver and lead group. Amongst the facts in favour of the relation- ship of thallium to the alkali metals, the following have been adduced : It forms a readily soluble, highly alkaline oxide, a soluble and alkaline carbonate, an insoluble platino-chloride, and with alumina a double sulphate, having the crystalline form of common alum, with a similar composition. As reasons, among others, for classing thallium with the heavy metals, have been urged the complete or nearly complete insolubility of its peroxide, sulphide, phosphide, iodide, bromide, chloride, chromate and phosphate ; its ready reduction from aqueous solutions of its salts by metallic zinc ; the highly poisonous character of its compounds ; the production of a brown inso- luble peroxide by electrolytic means ; its high atomic weight ; the complexity of its photographic spectrum, shown by Dr. "W. A. Miller to contrast strongly with the simplicity of those of the alkali metals ; its low conducting power for electricity, which is close to that of lead and tin, and much inferior to that of the alkali metals ; its specific heat, which coincides with that of lead ; its density and melting-point, very near those of lead, and finally its physical appearance and character, which approach so nearly to those, of lead, that few persons would notice at first sight any difference between the two metals. 174. SILVEE. This metal is weighed as chloride, as sulphide, as cyanide, or in its pure metallic state. It is likewise estimated by volu- metric processes. Quantitative estimation as Chloride. To the solution con- tained in a long-necked flask, and acidified with nitric acid, hydrochloric acid is added in excess ; the whole is then well agitated, and allowed to remain for several hours in a warm place ; the clear fluid is carefully separated by decantation 352 QUANTITATIVE ANALYSIS. from the precipitated chloride, which is then washed with water acidulated with hydrochloric acid. It is then trans- ferred to a weighed porcelain crucible, in which the washing is continued with distilled water till all traces of acid are re- moved. The'various washings are collected in a beaker, and, if the whole be not perfectly clear, it must be allowed to stand in a warm place for several hours, and the precipitate, if any should form, must be added to the contents of the porcelain crucible. The chloride of silver, having been perfectly washed, is heated to incipient fusion, and weighed ; it may afterwards be completely removed from the crucible, by reducing it by means of sulphuric acid and zinc. Should the quantity of chloride obtained in the experiment be small, it may be ad- visable to collect it on a filter, from which it should, after washing, be removed as completely as possible, and the filter, with the residue remaining on it, burnt, on the cover of the crucible, the ashes being mixed with the bulk of the chloride, which is then heated to incipient fusion in a counterpoised porcelain crucible as before. According to Rose, it is not admissible to precipitate silver by the chloride of ammonium, as this salt is capable of retaining traces of silver in solution. In cases where the presence of much chloride of ammonium is unavoidable, Gay-Lussac and Liebig recommend to evapo- rate the solution filtered from the chloride of silver nearly to dryness, and to treat the residue with nitric acid ; on exposing the whole to heat, the alkaline chloride is converted into ni- trate, while the small quantity of chloride of silver remains un- altered, and does not dissolve when the mixture is diluted. The composition of chloride of silver is One equivalent of Ag . . . IOS'0 . . 75'26 One ditto of 01 . . . 35 -5 . . 24'74 One ditto of AgCl . . 143-5 . . 100-00 Quantitative estimation as Sulphide. Washed sulphuretted hydrogen gas is transmitted through the solution as long as a precipitate continues to form ; the whole is then warmed, and allowed to settle ; it is finally collected on a weighed filter, washed as rapidly as possible out of contact of air, and dried at 212. Silver may likewise be precipitated as sulphide from neutral and alkaline solutions by sulphide of ammonium ; but, as in this case the precipitated sulphide is invariably accom- panied by sulphur, it is necessary to convert it for weighing SILVEE. 353 into chloride, by digesting it with nitric acid, and then preci- pitating it with hydrochloric acid. Even when precipitated from acid solutions by sulphuretted hydrogen, sulphide of silver very frequently contains sulphur, from which it may be freed by digesting it with a hot, moderately strong solution of sulphite of soda. The composition of sulphide of silver is One equivalent of Ag ... 108 . . 87'09 One ditto of S .... 16 . . 12-91 One ditto of AgS . . .124 . . 100 -00 Quantitative estimation as Cyanide. Cyanide of potassium is added in sufficient quantity to redissolve the precipitate which is at first formed, dilute nitric acid is then added, and a gentle heat applied, by which the whole of the cyanide of silver is precipitated. It is washed and dried at 212. The composition of cyanide of silver is One equivalent of Cy . . . . 26 . . 19 '4 One ditto of Ag ... 108 .. 8Q-6 One ditto of AgCy . . Tisi . . lOO'O Silver when combined with organic acids, is estimated in the metallic state, to which it is reduced by igniting the salt in a porcelain crucible. The heat should be gentle at first and the cover on, to prevent loss during the ignition. The lid is after- wards removed, and a strong heat applied for a considerable time, in order to effect complete combustion of the carbon of the organic acid. Silver is completely separated from all the metals of the first four groups by sulphuretted hydrogen. The solutions must in all cases be acid. Fom lead it may be sepa- rated by hydrochloric acid ; or by heating the solution contain- ing both metals with cyanide of potassium, which precipitates the lead in the state of carbonate, retaining the silver in solution as argento-cyanide of potassium. The silver is subsequently precipitated in the form of cyanide of silver by the addition of nitric acid. A method of determining the amount of silver in argenti- ferous galena has been founded by Mene on the solubility of oxide of silver, and the insolubility of oxide of lead in caustic ammonia in excess. About 300 grains of the specimen to be analysed are pulverized, and boiled with nitric acid diluted with three or four times its volume of water in a porcelain capsule. 354 QUANTITATIVE ANALYSIS. In a little while, all the sulphur is separated, and the lead dis- solved; the filtered liquid is precipitated by a great excess of ammonia, then filtered again rapidly, and the precipitate washed with ammoniacal water. By this reagent all the oxides are precipitated at first, and afterwards those which are soluble, are again dissolved. The liquid is treated with an excess of hydrochloric acid mixed with a few drops of nitric acid to facilitate the precipitation of the silver which is separated and weighed as chloride. This process is said to be applicable in every case of analysis, whatever may be the elements contained in the specimen under examination. Estimation of Silver in alloys (wet method). The analysis of alloys of silver by a solution of chloride of sodium may be effected in three different ways : 1. The silver may be pre- cipitated by an excess of the alkaline chloride, the weight of the chloride of silver produced indicating the amount of silver. 2. The quantity of chloride of sodium in a given weight of its aqueous solution being known, the amount of silver in the alloy may be ascertained by observing the weight of the alkaline chloride required to precipitate it completely. 3. The quantity of chloride of sodium in a given volume of its aqueous solution being known, the amount of silver in the alloy may be ascer- tained by observing the volume of the solution of alkaline chloride required to precipitate it. The last of these methods was proposed by Gay-Lussac in 1829, and as it is one of great simplicity and accuracy it is generally adopted in the British, Continental, and American mints for the assay of bars and coins of silver. The process is conducted in the following manner (Pelouze and Fremy) : . (a.) Preparation of pure Silver. A certain quantity of ordi- nary silver is dissolved in nitric acid ; should any residue re- main, it is separated by decantation, the solution is diluted with water and precipitated by slight excess of chloride of sodium ; the resulting chloride of silver having been well washed and dried, is reduced at a red-heat in a Hessian crucible with a mixture of chalk and charcoal, the proportions being 70'4 parts of chalk, and 4'2 parts of charcoal for every 100 parts of dry chloride of silver. The decomposition takes place in accordance with the following equation : Oxychloride of Calcium. The reduced silver forms a button at the bottom of the cruci- SILVER. 355 ble ; it is removed, well washed, re- dissolved in nitric acid, the solution again precipitated by common salt, and the chloride again reduced by chalk and charcoal ; the metal is now perfectly pure. It should be reduced to a granular state, or rolled into thin plates to facilitate its solution in nitric acid. (b.) Preparation of a standard or normal solution of Chloride of Sodium. A normal solution of chlori'de of sodium may be prepared by dissolving 5'414 grammes of the pure salt in 1 litre of distilled water at 15 C. ; a decilitre (100 cubic centi- metres) of this solution will precipitate exactly 1 gramme of pure silver. It is better, however, to prepare the normal liquor from ordinary culinary salt. For this purpose 200 or 300 grammes should be dissolved in about 2 litres of water ; a few grammes of the solution should be evaporated to ascertain the exact amount of salt it contains, and the solution should then be diluted with such a quantity of distilled water as would be re- quired, supposing the salt to be pure. 1 gramme 'of the pure silver is dissolved in pure nitric acid, and the solution precipi- tated with the proper precautions by 100 cubic centimetres of the saline solution, but as the salt is not pure, this quantity will not be found sufficient to effect the complete precipitation of the silver, and it will be necessary to add a certain number of cubic centimetres of a decinormal saline solution, the exact com- position of which is known. The volume of this solution which has been required to complete the precipitation of the silver is noted, and a calculation is made to ascertain how much of it must be added to the first solution to render it normal ; the experiment should be repeated two or three times to ensure accuracy. It is not, however, absolutely necessary that 100 cubic centi- metres of the saline solution should precipitate exactly a gramme of pure silver, it suffices if it approximates very closely, and that the exact quantity of silver precipitated be known. (c.) Preparation of the Decinormal Saline Solution. 100 cubic centimetres of the normal solution are poured into a flask of the capacity of 1 litre which is then filled up with dis- tilled water and well mixed ; it is evident that a litre, or 1000 cubic centimetres, of this solution will be required to precipi- tate 1 gramme of pure silver ; and that one-thousandth part, or 1 cubic centimetre, will precipitate one- thousandth part of a gramme or 1 milligramme of silver. 356 QUANTITATIVE ANALYSIS. (d.) Preparation of the Decinormal Solution of Silver. One gramme of the pure metal is dissolved in 5 or 6 grammes of pure nitric acid, and the solution diluted with distilled water to exactly the volume of 1 litre. The decinormal silver solution and the decinormal saline solu- tion are thus prepared of such strengths, that on mixing together equal volumes of each, neither nitrate of silver nor chlo- ride of sodium remain in the liquor, but nitrate of soda is in solution, and chloride of silver precipi- tated in accordance with the equation In laboratories where the wet assay of silver alloys is constantly go- ing on, large quantities of the normal saline so- lution are prepared at once, and the pipette is arranged as a fixture in the manner represented in Tig. 73, where o re- presents the reservoir containing the normal saline solution, and n a tube through which air is admitted when the apparatus is in use. The solution is deliver- ed from the reservoir through the bent tube I, furnished with a stop- cock. The pipette gf is placed in communication with the bent delivery-tube by the tube k in which a thermometer is SILVER, 357 inserted ; i is a stopcock which serves to establish a communi- cation between the tube I k and the pipette. The apparatus a b is employed to facilitate the exact emptying of the pipette, the tray c d slides backwards and forwards between two grooves, its motion in either direction being limited by the stops a and b. As seen in the figure, the assay-bottle c is immediately under the lower extremity of the pipette, from which the liquor can be delivered without wetting the neck or sides of the bottle ; d is a small vessel for receiving the excess of the saline solution, and e is a small sponge so adjusted that when the tray is slid along the groove till it is stopped by #, the sponge just touches the lower extremity of the pipette. In order to fill the pipette, its lower aperture is closed by the finger, and the stopcocks I and i being opened, the solution enters from the reservoir, the air escaping through an aperture in the plug of the lower part of the stopcock, which aperture can be closed at pleasure by the plug h. The pipette being filled, and the aperture closed, the finger is removed, and the sponge brought in contact with its point, the plug li is now gently relaxed, and the liquor allowed to flow from the pipette until the mark g is reached, the plug is then closed, and the last drop having been removed by the sponge, the burette remains filled with exactly 100 cubic centi- metres of the saline solution. Previous to applying this process to the analysis of silver alloys, it is necessary to know approximately the value of the assay ; that is, to operate upon a quantity containing nearly one gramme of silver ; a tentative experiment on a gramme of the alloy with the normal saline solution will give the requisite in- formation on this point. Let us suppose by way of illustration that a piece of French silver coin is to be examined, which, in order to be of the standard quality as prescribed bylaw, should contain ^VoV^ 8 * silver. The quantity that should be taken to represent 1 gramme of silver is formed by the proportion 827 : 1000 : : 1000 : X x = 1-115. T115 grm. of the alloy are therefore dissolved with the aid of a gentle heat in 5 or 6 cubic centimetres of pure nitric acid, the solution is transferred to the bottle c, Fig. 73, and the charge from the pipette delivered into it. The bottle is closed with its stopper, and briskly agitated for two or three minutes ; if only one or two assays are being made, this may be done with 2 B 358 QUANTITATIVE ANALYSIS. the hand, but when several samples are being operated upon, the ac/itator, Eig. 74, is employed. This consists of a frame d provided with a series of compartments e e e, etc., each of which will hold exactly one of the bottles, the frame is suspended from the steel spring a &, between two strong springs of vulcanized india-rubber. The stoppers of the bottles having been well secured, the apparatus is grasped by the handle c d, and briskly agitated for two or three minutes ; the chloride of silver is soon completely precipitated, leaving the supernatant fluid quite clear. One cubic centimetre of the decinormal saline liquor is now delivered into the bottle from a small pipette, this pipette is a small tube open at both ends, the lower aperture being con- siderably contracted ; it has two marks upon it corresponding to 1 and 2 cubic centimetres ; it passes through the cork, and nearly to the bottom of the bottle containing the deci- normal saline liquor, so that the requisite quantity can be easily withdrawn, by applying the finger to the upper end, then removing it from the bottle, and allowing the liquor to drop out from the lower aper- ture by relaxing the finger un- til the mark indicating 1 cubic centimetre has been exactly reached. Should the solution still contain silver, it is re- vealed by the formation of a white cloud ; hereupon the bottle is again agitated, and a second cubic centimetre of the decinormal saline solution in- troduced. Suppose that 3 cubic centimetres have in this way been added, and that the Fig. 74. fourth no longer produces a white cloud, thequestionarises, was the whole of the third cubic centimetre required or only part of it ? But as this question cannot be decided, one-half of SILVER. 359 the third cubic centimetre is assumed as the real quantity ; the error arising from this arbitrary measure cannot amount to more than half a millimetre, since 1 cubic centimetre of the decinormal saline liquor corresponds to 1 milligramme of silver. The result of the assay then is this : (1) 100 cubic centimetres of the normal saline liquor, equivalent to 1*000 gr. of silver ; (2) 2*5 cubic centimetres of the decinormal saline liquor, equi- valent to 2'5 milligrammes of silver: the quantity of alloy sub- mitted to analysis contains therefore 1*000 gr. + 0*0025 gr. = 1*0025 gr. of silver. The standard of the alloy is therefore found by the following proportion : 1*115 : 1000 : : 1000 : x 899 x 0*899, or the standard of the alloy is Suppose, however, that no white cloud has arisen after the introduction of the first cubic centimetre of the decinormal so- lution ; this shows that the whole of the silver has been preci- pitated by the 100 cubic centimetres of the normal saline liquor, and that quantity has probably been too much ; to ascertain this, the cubic centimetre of the decinormal saline solution which has just been added, must be neutralized by 1 cubic centimetre of the decinormal silver solution, and the liquor must be agitated till clear, the operator then adds successive cubic centimetres of the decinormal silver solution, until a white cloud is no longer produced. Let us suppose that a cloud ceases to make its ap- pearance on the addition of the fourth cubic centimetre, then, as before, one-half of the third cubic centimetre is taken as re- presenting the real quantity that has sufficed ; the 1*115 gramme of alloy contains therefore 1*0050 0025 = 0'9975 gramme of silver, and the standard is formed by the proportion 1-115 : 1000 : : 1000 : x #=0*8946, which is below the legal French standard, viz. 0'897. Dilatations and contractions are occasioned in the volume of the normal saline liquor by variations in temperature, and M. Gay-Lussac constructed a table of corrections to be applied. It is better, however, when a series of assays is about to be made, to make at the same time a standard assay with a gramme of pure silver; the exact value of the normal solution is thus ascertained, and all the assays made on the same day must be corrected in accordance with the divergence from the true value of the liquor indicated by the test experiment. If the operator is working with English weights and mea- 2B 2 360 QUANTITATIVE ANALYSIS. sures, the solution of common salt should be made of such a strength that 1000 grain measures of it precipitate exactly 10 grains of silver, and the -decimal solution of silver should be prepared by dissolving 10 grains of pure silver in nitric acid, and diluting it with distilled water till the solution occupies the bulk of 10,000 grain measures of water, each 10 grain mea- sures of this solution will therefore contain O'Ol grain of silver. The only metal the presence of which interferes with the ac- curacy of this process is mercury, a metal not likely to be met with in silver alloys ; if present, however, it would be precipi- tated by the chloride of sodium together with chloride of silver in the form of calomel. To avoid this inconvenience, Levol re- commends to supersaturate the nitric solution with caustic ammonia, then to add the test-liquor, and afterwards to super- saturate the excess of ammonia with acetic acid ; he states that by this modification he is able either with the presence or absence of copper to estimate accurately silver containing a tenth part its weight of mercury. Gay-Lussac simplifies this process by adding to the nitric solution of the alloy the am- monia and acetic acid at one and the same time, but in suffi- cient quantity to saturate the whole of the nitric acid, both that in combination with the silver, and that in the free state. He finds acetate of soda to answer quite as well as acetate of ammonia. Estimation of Silver in alloys (dry process). Cupellation : This method, which is the one usually adopted at Goldsmiths' Hall, and by refiners, is founded on the property possessed by silver of being unoxidizable, and nearly fixed, at a red- heat ; whilst copper in the presence of lead oxidizes, and forms with the oxide of lead a fusible glass, which passes into, and is absorbed by the cupel, the silver being retained, as it were on a filter, in the form of a bright globule or button. The cupel, which is best made of a mixture of finely levigated ashes of birch-wood and calcined bones, is thus prepared : the ash, slightly moistened, is laid in a brass mould somewhat deeper than that of the cupel intended to be made; in this is placed a curved and polished steel pestle, which is then struck smartly with a hammer. The operator must be careful to put as much ash into the mould as is required to make the cupel at once, it is otherwise apt to separate in layers when it comes to be heated. The little vessel thus made is dried with great care, and heated to redness before it is used. One SILVER. 361 part, by weight, of the cupel absorbs during the process of eu- pellation the oxide formed by two parts of lead. The assayist is thus furnished with a guide to the size of the cupel required for any particular experiment. The proportion of lead required varies with that of the copper in the alloys ; it is necessary therefore before proceeding with a definite experiment to determine approximatively, the standard or fineness of the silver. In the case of a piece of coin this is generally tolerably well known, but in all cases the fineness of the silver should be ascertained by a cupellation experiment with O'lOO gr. of the alloy and 1 gr. of lead, the weight of the button obtained furnishes a sufficiently close approximation, and from this preliminary trial, the quantity of lead which should be added is ascertained by reference to the following table con- structed by M. d' Arcet. The quantity of alloy usually employed is 1 gramme, the weight of the button in milligrames indicates the standard of the alloy: thus a button weighing Ogr. 900 milligrammes represents an alloy of Table showing the Quantity of Lead required for the Cupellation of various Alloys of Silver and Copper. Standard Amount of Quantity of Quantity of Lead of Copper Lead in relation to Silver. alloyed. necessary. that of Copper. 1000 . . T 3 o - . 950 . . . 50 . 3 . . eotoi 900 . . . 100 . 7 . . 70 to 1 800 . . . 200 . 10 . . 50 to 1 700 . . . 300 . 12 . . 40 to 1 600 . . . 400 . 14 . . 35 to 1 500 . . . 500 . . 16 to 17 . . 32 to 1 400 . . . 600 . . 16 to 17 . . 27 to 1 300 . . . 700 . . 16 to 17 . . 23 to 1 200 , . . 800 . . 16 to 17 . . 20 to 1 100 . . . 900 . . 16 to 17 . . 18 to 1 Pure co pper 1000 . . 16 to 17 , . 16 to 1 Example : Suppose a piece of silver of the approximative fineness of T 9 $nr * s * be analysed. The quantity of lead in- dicated in the table, viz. 7 grammes, is placed in the cupel, which is then introduced into the muffle, and heated to bright redness in the cupel furnace. When the lead is melted, and its .surface brilliant, 1 gramme of the alloy enveloped in a piece 362 QUANTITATIVE ANALYSIS. of pure sheet lead is introduced into the cupel ; it soon enters into fusion, and the surface of the liquid mass which is at first plane, becomes by degrees convex, in appearance it is now like a drop of oil. The fused oxide of lead is absorbed rapidly by the cupel, and a portion is volatilized and makes its escape from the muffle in the form of fumes ; when the volume of the alloy is reduced to about two-thirds, the cupel is advanced to the mouth of the muffle ; the surface of the button soon loses its brilliancy, and iridescent bands appear on its surface, which are occasioned by very thin layers of oxide of lead. The object of advancing the cupel to the mouth of the muffle, is to reduce the temperature of the button, which at this period becomes agi- tated with a rapid circular movement ; this suddenly ceases, arid the button having for a moment emitted a bright flash of light, technically called fulguration or coruscation, becomes white and fixed. The button should at the moment of brightening be covered with another cupel which has been kept hot for the pur- pose ; a portion of the metal may otherwise be lost by disper- sion, from sprouting or spitting. This phenomenon appears to arise from the sudden escape of the oxygen, which the silver had mechanically absorbed while in the melted state : if the but- ton has been cooled too rapidly, a crust is formed over the sur- face, while the interior portion remains fluid, and upon this soli- difying, the crust is ruptured by the sudden expansion of the metal, and little tubes or globules of the metal are expelled by the sudden escape of the gas. The cupel still covered, is now removed from the muffle, and when cold the globule is de- tached, brushed, and weighed. If the assay has been a good one, it adheres but loosely to the cupel, and its surface is clean, brilliant, and convex ; but if it has been too strongly heated, it is removed with difficulty, and its surface is dull and irregular. The adherence of the button to the cupel, and the flatness of its surface, may also arise from a deficiency of lead. In Fig. 75, A is the muffle, on each side of which is a slit o o ; the size of this muffle is in proportion to that of the furnace, one of a convenient size should be capable of containing six or eight cupels. The muffle when introduced into the furnace is so arranged that while its closed end is supported by a shelf of refractory brick #, its open end corresponds exactly with the aperture d of the furnace, to the sides of which it is luted by ^a little moistened fire-clay. The fuel (charcoal or coke) is sup- plied through the aperture/, which during the operation is SILVER. 363 closed by the door a. The interior of the muffle is in this way constantly traversed by a current of air, and the draught of the furnace is increased by the tall chimney H. The furnace having been lighted by introducing ignited charcoal by the opening, i, and all the apertures except that of the ash-pit having been closed, the muffle is allowed to get red-hot ; the cupels which have been previously drying round the ledge of the chimney are then placed by a pair of tongs on the floor of the muffle, on which a small quantity of pounded bone-ash has been strewed, the opening d is now closed ; the cupels are soon raised to the temperature of the muffle, and when this is the case the door is removed, and the assay dropped carefully into the cupels by a pair of light steel tongs. Fig. 75. From a series of experiments made first by Tillet and con- firmed and extended by D'Arcet, the following conclusions may be drawn : (1.) That by the process of cupellation the amount of silver is always slightly under-estimated, in consequence of a portion being volatilized and a portion being absorbed by the cupel. 364 QUANTITATIVE ANALYSIS. (2.) That the button is not pure silver, but contains both lead and copper ; a wet analysis shows that as an average, 1000 parts of the button contain four parts of foreign metals. (3.) The more copper the alloy contains, the larger the quan- tity of lead required ; this however is only applicable to alloys the standard of which is above 500. Copper containing a few thousandths of silver only does not require more lead than an equal weight of an alloy of 500. Bismuth acts in cupellation in the same manner as lead, and M. Claudet has drawn up a table showing the proportions that should be used, in different standards of alloys, but no applica- tion of these facts appears to have hitherto been made. To estimate the amount of loss during the operation of cupellation, recourse is sometimes had to a "table of errors," which however cannot be made exact, because such a table im- plies certain invariable conditions which in practice cannot always be complied with. It is better to use with each set of assays a check ov proof, that is, to make an assay with pure silver. The proof is placed in the muffle by the side of the assays ; e. g. suppose an assay has to be made of a piece of coin the approximative standard of which is yV^o '> 950 milligrammes of pure silver and 50 of copper are placed in a cupel and in- troduced into the muffle by the side of the alloy. Suppose that after cupellation, the proof has lost two milligrammes, this quan- tity must be deducted from the weight of the button furnished by the alloy. These proofs are especially useful when the alloy contains, gold, platinum, or palladium, as these metals tend to occasion an overweight in the button. Some experiments were made by Mr. Hambley (Chem. Gaz., vol. xv. p. 185), to ascertain whether the loss of silver by cnpellation is the same, when varying quantities are employed with a constant ratio of lead, and also to find the loss of silver when cupelled with a gradually increasing ratio of lead. The conclusions drawn from the experiments are 1. That accord- ing to the decrease in weight of the silver cupelled, so the loss of that metal very slightly increases, provided the ratio of lead employed be constant ; 2. That an increasing ratio of lead produces an increasing loss of silver. According to Field (Chem. News, vol. i. p. 277), the pre- sence of copper exercises a material influence upon the loss of silver in the process of cupellation, in the estimation of that metal; when it exists as a double sulphide with copper, the loss SILVER. 365 is very apparent, so much so indeed, that in the assay of these double sulphides, Field found it necessary to abandon the usual method and to adopt the following : The finely powdered mineral is digested in strong nitric acid until the sulphur is yellow ; the solution is somewhat largely diluted with water, and one or two drops of hydrochloric acid added. After heat- ing gently, and allowing the supernatant liquor to become clear, it is filtered. The residue on the filter, consists of earthy in- soluble matter and chloride of silver. It is dried, and the greater part mixed in a mortar with carbonate of soda, a small quantity of litharge, and a few grains of bitartrate of potassa. About half the mixture is put in a crucible, and the filter is then folded up and placed upon the powder, the remainder of which is now introduced, and the whole covered with a little fused borax. The crucible is now introduced into the furnace ; an argentiferous button of lead is obtained, which is cupelled in the usual manner. Among the various other methods of separating silver from copper, two may be specially noticed. The first consists in fusing the argentiferous copper with 2 parts of lead, and cooling the fused mass in thick round cakes. These cakes are then intro- duced into a furnace of peculiar construction, and the heat is raised sufficiently high to fuse the alloy of silver and lead, but not to fuse the copper. On cooling, the whole of the silver is found combined with the lead, from which it is separated by cupellation. The second method consists in dissolving the ar- gentiferous copper in sulphuric acid, in a platinum vessel, and precipitating the silver from the solution by means of plates of copper ; the precipitated copper, which is in the form of a grey metallic powder, is washed and fused with a mixture of nitre and borax ; it is thus purified from the copper which may have been precipitated with it. The method has two advan- tages ; first, the copper is recovered in a marketable form (that of blue vitriol) ; and second, the gold amounting to from T oW to YaVo^h P ar t, is saved, this metal remaining undissolved by the sulphuric acid. To estimate the amount of silver in a compound (commercial lead for example), Pisani avails himself of the fact that iodide of starch is decolorized when poured into solutions of some salts, whilst in others it retains its blue colour. The decolori- zation is produced by silver, mercury, the protosalts of tin, anti- mony, arsenic, iron, and manganese, and perchloride of gold* 366 QUANTITATIVE ANALYSIS. The salts of lead and copper have no action on it. In order, therefore, to estimate the amount of silver in a compound, a standard solution of iodide of starch is dropped into the solu- tion until the last drop is undecomposed, and causes a blue tint to remain in the liquid. Instead of forming a standard solu- tion of iodide of starch, Field dissolves iodine in iodide of po- tassium, and drops it into the solution of the silver salt to which a little starch-water has previously been added. The silver compound is dissolved in nitric acid, gently evaporated nearly to dryness, a slight excess of carbonate of soda introduced, and the carbonate of silver brought into solution by very weak acetic acid ; a few drops of clear starch-water are added, and the ioduretted iodide of potassium dropped in from a burette. "When the two liquids inert, a bright blue ring is formed, which immediately disappears on agitation, yellow iodide of silver being precipitated. When a permanent blue tinge is produced, no more silver exists in the solution. Three burettes are em- ployed ; in the first one, each division corresponds to -^ of a grain of silver, in the second to T i-Q, and in the third to T oVo- When it is evident that the greater part of the silver has been precipitated by the solution from the first burette, a small quantity from the second is introduced, and the last traces are thrown down by the third. No difficulty is experienced in compounds of silver and tin, but the presence of mercury en- tirely vitiates the results, as a salt of this metal decomposes the starch compound with great facility. 175. MERCURY. This metal is in general most conveniently weighed in the metallic state. It may also be estimated in the forms of sub- chloride and sulphide. Estimation as metallic Mercury : (1.) Reduction in the dry way. The solid mercurial com- pound is heated in a tube of hard glass, with an excess of soda lime, precisely in the same manner and with the same precautions as are observed in the analysis of ammoniacal salts (see page 273). The open end of the combustion tube is drawn out, and bent at a somewhat obtuse angle, in order that it may be in- serted into a flask containing water, into which the mercury is received ; when the analysis is over, the fluid metal is collected into one large globule, by agitating the flask, it is then decanted into a porcelain capsule, and the adhering water having been MERCURY. 367 removed by blotting-paper, it is dried in vacuo over sulphuric acid, without the application of heat. Rose recommends to introduce into the combustion tube, first a column of bicarbonate of soda, then one of quicklime, and then a well-blended mixture of the mercurial salt and quick- lime, and finally a column of quicklime ; the tube is to be heated first at the open end, finishing with the bicarbonate of soda. When the operation is ended, the bent end of the tube is cut, the mercury, collected in a flask, dried with paper and afterwards over sulphuric acid. Combinations containing iodide of mercury are not entirely decomposed when heated in this way, subiodide being condensed in the extremity of the tube simultaneously with the metallic mercury. To analyse these combinations recourse must be had to metallic copper, the ope- ration being similar to that with quicklime. In their experiments for determining the atomic weight of mercury, Erdmann and Marchand adopted the following some- what complicated method of reducing the oxide, their object being the attainment of the greatest accuracy (Journ. fur Praktische Chemie, xxxi. p. 385). A combustion tube about 3 feet long was drawn out in front to an open point, from 9 to 10 inches in length, and curved downwards. A loose stopper of copper shavings, which had been first oxidized by heating them while exposed to the air, and then reduced in a current of hydrogen, was introduced through the other end, and thrust forward near the point. This copper was followed by a stratum 5 to 6 inches in length, of small fragments of strongly ignited sugar-charcoal, from which every trace of dust had been carefully removed by sifting, and then the oxide (which had previously been strongly ignited in a current of air so as to remove every trace of mercurial vapour) was intro- duced. To advance to the front every trace of oxide which might have remained adherent to the hinder portions of the tube, it was finally rinsed with pulverulent copper : the tube, thus arranged, was treated precisely in the same manner as in organic analysis, and then placed in a long furnace. To the hinder extremities, a broad tube, filled with chloride of cal- cium, was fixed by means of a caoutchouc tube ; to this was applied a Liebig potassa apparatus, filled with sulphuric acid, and at last a large gasometer filled with carbonic acid. The point in front of the tube was connected by a caoutchouc tube with a weighed recipient, destined to receive the mer- 363 QUANTITATIVE ANALYSIS. cury, and in the arm proceeding from the last bulb of the latter, there was placed some gold leaf, to retain any trace of mercurial vapour which had not been condensed in the bulb. A gentle stream of the dry carbonic acid gas was first allowed to pass from the gasometer through the apparatus, the tube was then immediately surrounded with incandescent charcoal, proceeding from the front towards the hinder part, in the same manner as in an organic analysis. The carbon in the anterior portion of the tube is seen to burn at the expense of the libe- rated oxygen, and the mercury, which distils over in the cur- rent of carbonic acid, collects perfectly bright in the recipient. On the combustion of the charcoal, some water is formed, which passes over along with the mercury. This water is en- tirely removed, as well as the carbonic acid contained in the apparatus, by a current of atmospheric air at the close of the operation. By this mode of analysis these chemists obtained from four experiments the following numbers : 92*594, 92'596, 92-598, and 92'596, the mean of which, 92-596, gives as the equivalent number of mercury the figure 10O07. A modification of this process, consisting in the employ- ment of a current of hydrogen gas to assist in the reduction of the mercurial compound, was adopted by M. Millon (Ann. de Chim., 1846). Hydrogen gas, he observes, is more easily obtained in a regular current than most other gases, and it greatly facilitates the decomposition of all mercurial compounds : it favours the expulsion of the water which accompanies the reduction, and likewise the condensation of the mercury in the expansion of the tube in which it is to be collected and weighed. The pro- cess is conducted as follows : The dry gas is passed through a tube containing copper turnings heated to redness, a plan which was found the most efficacious for preserving the perfect metallic lustre of the mercury, while it also ensured the purifi- cation of the hydrogen. On leaving the tube with the copper turnings, the gas enters the tube containing the mercurial salt. This tube should be from 14 to 16 inches long, and of the diameter of an ordinary tube for organic analysis. At a small distance from its free extremity, it is contracted, and then again contracted and drawn out at the point, and curved upwards : it thus presents a space of from 3 to 4 inches be- tween the two contractions. A little asbestos is first placed in the tube next the first MERCURY. 369 contracted part ; then caustic lime in small fragments, to the extent of from 6 to 8 inches : the mercurial compound is next introduced, the quantity of which may vary from 15'5 to 62 grains, and then the tube is filled with caustic lime similar to the other. The tube is now placed in the furnace used 'for organic analysis : it receives the current of hydrogen at its wider and uncontracted extremity, and heat is applied in the usual manner. The ignited charcoal is gradually approached to the part of the tube containing the mercurial compound, and then a few pieces are placed behind it, to prevent the con- densation of the metal there. The water is first seen in the por- tion of the tube between the contractions : it is dissipated by gently heating it. This part of the tube is then allowed to cool, and the mercury now makes its appearance, condensing in its turn without any difficulty. At the end of the operation the part of the tube containing the mercury is separated by slightly moistening the heated tube ; the portion of the tube is weighed with the mercury it contains ; the metal is poured out, the tube is rinsed out by nitric acid ; it is then washed, dried, and weighed again. The difference between the two weights gives the weight of the mercury. The author states that by this mode of operating he has been able to drive 45 to 60 grains of mercury from one extremity of the tube to the other, condensing it between the contracted parts without the smallest loss. Two analyses of chloride of mercury yielded 73 '87 and 73'82 per cent, of mercury, figures which give as the atomic weight of mercury 100'07, the same as that ob- tained by Erdmann and Marchand. (2.) jReductionin the wet way. The reducing agents are pro- tochloride of tin and phosphorous acid. The process with the former is conducted as follows: The mercurial compound, if a solid, is digested with strong hydrochloric acid ; a concen- trated solution of protochloride of tin, which has been ren- dered perfectly clear by the addition of a few drops of hy- drochloric acid, is then added. The whole is boiled, but only for a few minutes, to avoid the risk of the volatilization of a portion of the mercury in company with the aqueous vapours. On cooling, the mercury is usually found deposited in the form of a black precipitate ; the supernatant fluid is removed by a siphon, and the precipitate is boiled with hydrochloric acid, on which it generally loses its pulverulent appearance, and becomes converted into running globules. It is washed, 370 QUANTITATIVE ANALYSIS. first with very dilute hydrochloric acid, and finally with dis- tilled water ; it is then received into a porcelain crucible, dried, first with bibulous paper, and lastly over sulphuric acid, in the manner already directed. When mercury has to be esti- mated in a liquid containing nitric acid, it is necessary to de- stroy this acid before a correct determination can be made : this is done by adding hydrochloric acid gradually to the so- lution, and concentrating by evaporation. The nitric acid is thus destroyed, free chlorine being at the same time disen- gaged, and the addition of the hydrochloric acid must be con- tinued as long as the odour of chlorine is perceptible ; pro- tochloride of tin is then added, and the remainder of the ope- ration is conducted in the manner already described. It is very difficult to obtain correct results when the solution con- tains much nitric acid. Whenever it is admissible, therefore, it is advisable to precipitate the mercury as sulphide, by sul- phuretted hydrogen, or colourless sulphide of ammonium, having previously nearly neutralized the solution with potassa, and mixed it with excess of cyanide of potassium. When phosphorous acid is employed, the mercury is precipi- tated as subchloride. When the liquid contains free hydro- chloric acid, the temperature may be raised to 140 without the subchloride being reduced to the metallic state. The precipi- tated subchloride is not immediately formed in very weak liquids ; it is necessary to leave the mixture undisturbed for twelve hours. The subchloride deposits itself readily, espe- cially when the liquid is sufficiently acid. The precipitate is collected on a filter, washed with hot water, and dried at 212. This process is, according to Eose, very applicable to cases where the liquid contains much nitric acid, only then the solution must be diluted with a sufficient quantity of water. Mercury can, according to the same authority, be easily se- parated by hydrochloric and phosphorous acids, from copper, cadmium, zinc, and even from bismuth, antimony, and arsenic. When the liquid contains bismuth, a large excess of hydro- chloric acid must be added, to prevent the precipitation of oxychloride of bismuth. The mercury being separated in the state of subchloride, the bismuth is precipitated in the filtered liquid as oxychloride, then reduced to the metallic state by fusion with cyanide of potassium. If the liquid contain anti- mony, this metal must be retained in solution by tartaric acid, which does not hinder the precipitation of the mercurial chloride. MERCUEY. 371 The composition of subchloride of mercury is Two equivalents of Hg . . . . 20OO . . 84-92 One ditto of Cl .... 35 "5 . . 15-Q8 One ditto of Hg 2 Cl . . . 235'5 . . 10OOO Estimation as Sulphide. A stream of washed sulphuretted hydrogen is transmitted through the acid solution of the salt ; in solutions of the suboxide, the precipitate formed is black at once; but, when a compound of oxide of mercury is under examination, in the beginning of the experiment white-coloured compounds of mercurial salts, with sulphide of mercury, are produced ; the addition of larger quantities of the gas causes the precipitate to assume various colours, but it ends in be- coming pure black. If the whole of the mercury exist in the original solution as oxide, it may be determined in the state of sulphide ; with which view, the precipitate, occasioned by sul- phuretted hydrogen, is received on a weighed filter, quickly washed with cold water, dried at 212, and weighed; but if the mercury, or any portion of it, existed as suboxide, it is inadmis- sible to estimate it as subsulphide, because it is liable to be partly decomposed, even by a gentle heat, into sulphide of mer- cury and metallic mercury ; and as the latter may be partly vo- latilized by a very gentle heat, an error of greater or less amount would be introduced (Hose). The sulphide containing a mini- mum of sulphur must therefore undergo further treatment as follows : It is collected on a filter, and transferred, filter and all, into a wide-mouthed flask, capable of being closed by a glass stopper. A small quantity of dilute hydrochloric acid is then poured into the flask, and a slow current of chlorine con- ducted into the solution ; the sulphide is hereby decomposed into chloride, sulphuric acid, and free sulphur ; as soon as the latter is observed to have a clear yellow colour, the stream of chlorine is stopped, and the flask is exposed to a gentle heat to expel the free chlorine ; it is then filtered off from the sulphur, and the mercury in the filtrate is estimated by pro- tochloride of tin. If the purity of the sulphide of mercury is questionable, Hose directs it to beredissolved and operated on as follows : To the washed precipitate add a weak solution of caustic po- tassa, and pass a current of chlorine through the liquid ; the sulphide is soon dissolved, especially if gently warmed ; under these conditions powdered cinnabar also dissolves. The mer- 872 QUANTITATIVE ANALYSIS. cury is entirely precipitated by sulphide of ammonium in a neutral or ammoniacal solution, without an excess of the reagent dissolving the precipitate which forms, but this is not the case when the liquid contains free potassa or soda, or the carbonates of these bases; here the alkali must be supersaturated with hydrochloric acid before precipitating the mercury. If the precipitated sulphide of mercury contain sulphur, it may be removed by treating it with a moderately concentrated hot solution of sulphite of soda, which dissolves the sulphur and be- comes converted into hyposulphite. The composition of sulphide of mercury is One equivalent of Hg . . . . 10OO . . 86'20 One ditto of S .... 16 . . 18-80 One ditto ofHgS . . . 116-0. Separation of Oxides of Mercury from Oxide of Lead : (1.) By Sulphuretted Hydrogen gas. The two metals are first precipitated together as sulphides from a diluted solution ; the mixed sulphides are dried, and introduced into a bulb blown in a tube of hard glass, the weight of which, before and after the introduction of the sulphides, is accurately taken. The tube is connected with an apparatus for generating chlorine, and a stream of the gas, dried by passing through a tube filled with chloride of calcium, is sent through the tube. When the whole apparatus is filled with chlorine, the bulb containing the sulphides is gently heated, upon which the chloride of mercury volatilizes, and is completely separated from the chloride of lead. The sublimed chloride of mercury is driven forward by the flame of a small spirit-lamp, and received in a vessel con- taining water. As soon as all appearance of sublimation ceases, the tube leading from the bulb to the receiver is cut off with a file, and any crystals which may have collected in it are washed into the flask. The chloride of mercury dissolved in the re- ceiver is precipitated by protochloride of tin, and the chloride of lead remaining in the bulb is determined by first weighing it together with the bulb, and then dissolving it out and weigh- ing the bulb alone. (2.) By Cyanide of Potassium. This method, which is far more simple and easy of execution than the one just described, is conducted as follows : The diluted solution of the two oxides is mixed with carbonate of soda, and then heated with excess of cyanide of potassium ; the whole of the lead is precipitated MERCURY. 373 in. the state of carbonate, while the mercury remains in solu- tion in the form of double cyanide of mercury and potassium. The solution is filtered off from the insoluble lead salt, and pre- cipitated by sulphuretted hydrogen. (3.) By Hydrochloric Acid. All the mercury present in the compound must be in the form of oxide. The dry mixture is treated with hydrochloric acid, and evaporated to dryness at a gentle heat. Alcohol, mixed with ether, is added to the residue, and the whole is digested : chloride of mercury alone dissolves, the insoluble chloride of lead is received on a filter, washed with alcohol, dried, and weighed. The alcoholic solution of chloride of mercury is evaporated to expel the alcohol and ether, and the mercury is then precipitated by protochloride of tin. (4.) By Sulphuric acid. Add sulphuric acid, and then al- cohol forming about one-sixth of the volume of the liquid ; if this does not contain sufficient hydrochloric acid, and if the proportion of sulphuric acid be insufficient, yellow subsulphnte of mercury may be precipitated, which is avoided by the addi- tion of sulphuric acid. The sulphate of lead requires washing with weak alcohol acidulated with sulphuric acid. Separation of Oxide of Mercury from Oxide of Silver : (1.) By Hydrochloric acid. Add a sufficient quantity of hydrochloric acid to the diluted solution ; after the deposition of chloride of silver, decant the supernatant liquid, then heat the chloride precipitate with a small quantity of nitric acid, add some water, and a few drops of hydrochloric acid, and then filter. Precipitate the mercury from the filtered liquid by phosphorous acid (Rose). (2.) By Cyanide of Potassium. The solution is nearly neu- tralized with potassa ; excess of cyanide of potassium is then added, by which the precipitate which first forms is entirely re- dissolved, and the solution contains the double cyanides of silver and potassium, and of mercury and potassium. On adding nitric acid, the cyanide of potassium is decomposed, the silver is precipitated in the form of cyanide of silver, while the cyanide of mercury remains in solution ; the former is separated by filtration, and from the filtrate, the mercury is precipitated by sulphuretted hydrogen. (3.) By Chlorine. This is effected precisely in the same manner and with the same apparatus as is employed in the separation of oxide of mercury from oxide of lead ; chloride of silver remains in the bulb, in which it is weighed, the bent tube, PART II. 2 C 374 QUANTITATIVE ANALYSIS. through which the volatilized chloride of mercury makes its escape into the receiver having been cut off. The fused chloride of silver must, after weighing, be removed from the bulb by zinc and sulphuric acid, in order that the weight of the latter may be taken. Separation of Suboxide of Mercury from Oxide of Mercury. The solution is diluted considerably with water, and the sub- oxide is precipitated as subchloride by hydrochloric acid : heat must be avoided. The subchloride, after standing for some time, is received on a weighed filter, and dried at a gentle heat. The oxide of mercury in the filtrate is precipitated by proto- chloride of tin. If the substance be a solid and insoluble, it is acted upon at a low temperature, with diluted nitric acid, until it is completely dissolved; hydrochloric acid is then added as before. Analysis of amalgams. If the metal or metals with which the mercury is combined are not volatile, or oxidizable by heat with access of air, the amount of mercury in the amalgam may be ascertained in the simplest manner by igniting it in a porcelain crucible. If, however, the metals are liable to alteration by exposure to heat in an open vessel, the ignition must be per- formed in a retort, the neck of which, after the volatilization of the mercury, must be closed up by the blowpipe, while the re- tort is still ignited. 176. BISMUTH. This metal is almost invariably weighed as oxide ; it is pre- cipitated from its solution in nitric acid by carbonate of am- monia ; the solution should be diluted, and it should be heated nearly to boiling for a few minutes before it is filtered, other- wise a portion of the oxide will be retained in solution by the precipitant. Neither carbonate of potassa nor carbonate of soda can be substituted for carbonate of ammonia; a portion of the former is carried down with the precipitate, and is not easily removed by subsequent washing, and the latter fails to effect a perfect precipitation. The carbonate of bismuth, hav- ing been washed, is separated from the filter, and ignited in a porcelain crucible, by which it loses carbonic acid, and becomes converted into protoxide of a yellow colour. The filter is burnt on the cover of the crucible, and its ashes added to the oxide. The composition of teroxide of bismuth is One equivalent of Bi . . . . 210 . . 8974 Three ditto of .... 24 .. 1O26 One ditto ofBi0 3 . . . 234 . . lOO'OO BISMUTH. 375 In the presence of sulphuric or hydrochloric acids, bismuth cannot be effectually precipitated by carbonate of ammonia, as in the former case a basic sulphate, and in the latter a basic chloride, is at the same time precipitated, and neither of these salts can be subsequently decomposed even by protracted di- gestion with excess of carbonate of ammonia. In such cases it is necessary, in the first place, to mix the solution of the bis- muth salt with acetic acid (water will not do, as it would cause the formation of an insoluble basic salt), and then to precipi- tate the bismuth in the form of sulphide by sulphuretted hy- drogen, or by sulphide of ammonium, having previously rendered the solution alkaline by the addition of caustic ammonia. The precipitated sulphide, having been washed, is decomposed by digesting it, filter and all, with nitric acid ; the solution is di- luted with weak acetic acid, filtered, the filter washed with the same diluted acid, and the filtrate finally precipitated by car- bonate of ammonia. Separation of Oxide of Bismuth from Oxides of Lead: (1.) By Sulphuric Acid. An excess of sulphuric acid is added to the solution containing the two oxides, and heat is applied till the sulphuric acid begins to volatilize : it is then quickly filtered, and the sulphate of lead is washed with water acidu- lated with sulphuric acid. The sulphate of bismuth in the filtrate is precipitated by carbonate of ammonia. The results are not very accurate, since sulphate of lead is not altogether insoluble even in dilute sulphuric acid. A better plan is the following: The solution of the two oxides in nitric acid is evaporated to dryness in a porcelain capsule on the water-bath : the dry residue is taken up by distilled water, and again eva- porated, when all free nitric acid is expelled, and the residue of bismuth is completely converted into a basic salt ; the capsule is allowed to cool with its contents, which are then treated with a cold aqueous solution of nitrate of ammonia of known strength. The solution is left for some time in contact with the solid residue, in order that the soluble nitrate of lead may be completely taken up by it. The solution of nitrate of lead is filtered oif, and the insoluble basic nitrate of bismuth which remains, is washed with the solution of nitrate of ammonia. The oxicle of lead may be precipitated as sulphate. The wash- ing must not be continued too long. (Lowe.) (2.) Ullyreris process. The two oxides are together precipi- tated with carbonate of ammonia, and redissolved in acetic 2 c 2 376 QUANTITATIVE ANALYSIS. acid. A strip of clean lead, of known weight, is then put into the solution, so that the whole of it is covered. The vessel is closed, and allowed to stand for some hours. The bismuth is separated in the metallic state, and that which remains on the strip of lead is washed off, and the strip dried and weighed. It is brought on to a filter, and washed with water that has been boiled and allowed to cool ; it is then dissolved in nitric acid, evapo- rated to dryness, ignited, and the oxide weighed. The solution of lead is precipitated with carbonate of ammonia, and the weight of the oxide determined from this, is to be deducted, the oxide corresponding to the loss which the strip of lead has suffered during the operation. (3.) By Heat. The process is the same as that employed for the separation of silver from mercury, and lead from mercury, namely, by converting the metals into chlorides by heating them in an atmosphere of chlorine, and expelling the volatile chlo- ride of bismuth by heat. Too high a temperature must be avoided, otherwise a portion of the chloride of lead may be also volatilized. The proportion of bismuth may be calculated from the loss of weight which the compound under examina- tion experiences. (4.) Liebig's method. To a cold solution of the nitrates of the two oxides carbonate of lime is added, which precipitates the bismuth, but not the lead. (5.) By Caustic Potassa. To the nitric solution caustic potassa is added; both oxides are precipitated; but on adding excess of the caustic alkali the oxide of lead is redissolved. Separation of Oxide of Bismuth from Oxide of Silver : (1.) By Hydrochloric Acid. Nitric acid is added to the so- lution, and then hydrochloric acid, which precipitates the silver as chloride. The bismuth in the filtrate is precipitated by sulphuretted hydrogen, the sulphide decomposed by nitric acid, and finally precipitated by carbonate of ammonia. (2.) By Cyanide of Potassium. This reagent is added to the solution of the two oxides, whereupon the bismuth is precipi- tated as carbonate ; the silver remains in solution as double cyanide of silver and potassium ; it is separated by filtration from the carbonate of bismuth. 177. CADMIUM. This metal may be quantitatively estimated as oxide or as sulphide. CADMIUM. 377 Estimation as Oxide. The solution under examination is precipitated by carbonate of potassa, and the resulting white carbonate of cadmium is decomposed by ignition into carbonic acid and water, which escape, and oxide of cadmium, which re- mains behind in the form of a brown powder. Ammoniacal salts interfere with the complete precipitation of oxide of cadmium by alkaline carbonates; carbonate of ammonia can- not, therefore, be employed as the precipitant. In precipi- tating oxide of cadmium by carbonate of potassa in the pre- sence of ammoniacal salts, the same precautions must be taken as in the precipitation of carbonate of zinc under similar cir- cumstances. The composition of oxide of cadmium is One equivalent of Cd ... 56 . . 87'5 One ditto of O .... 8 . . 12'5 One ditto of CdO . . . 64 . . lOO'O Precipitation as Sulphide. This may be effected either by sulphuretted hydrogen or by sulphide of ammonium ; in acid solutions the former reagent is employed. It should be largely diluted, and the gas allowed to pass slowly through the liquid for a long time ; the resulting sulphide is of a yellow or orange colour, according as the solution has been more or less di- luted. It is collected on a weighed filter, washed with dis- tilled water, and dried at 212. The composition of sulphide of cadmium is One equivalent of Cd . . . 56 . . 77'78 One ditto ofS . . . . 16 . . 22-22 One ditto of CdS . . . 72 . . 100-00 When sulphide of ammonium is employed as the precipitant, the sulphide of cadmium should be decomposed by digestion with hydrochloric acid, and the solution precipitated with carbonate of potassa. It has been observed by Dr. Keinsch (Jahrb. fiir Prakt. Phar., xiii. p. 72) that when a salt of cadmium is dis- solved in hydrochloric acid, and then treated with sulphuretted hydrogen, no precipitate takes place till after the gas has been passed through the solution for a long time, and that then the precipitate is a combination of sulphide of cadmium with chloride. On edulcorating with water, the chloride is dissolved, and the yellow sulphide is left on the filter. This observation shows the necessity of diluting the solution before passing the 378 QUANTITATIVE ANALYSIS. gas through it. He has observed, also, that in precipitating a hydrochloric solution containing copper, lead, bismuth, and cadmium by sulphuretted hydrogen, the three former metals are completely thrown down before the chloride of cadmium is at all decomposed ; the filtered liquor gives no traces of either of the three first metals, but on continuing to pass the gas slowly through it for a long time, the cadmium is precipitated of its characteristic yellow colour. Separation of Oxide of Cadmium from Oxides of Lead : (1.) By Cyanide of Potassium. The diluted solution of the oxides is first rendered slightly alkaline by carbonate of soda; cyanide of potassium is then added, and heat applied. The lead is precipitated as carbonate, the cadmium remains in so- lution as double cyanide of cadmium and potassium, and may be precipitated from the filtered . solution by sulphuretted hy- drogen. (2.) By Sulphuric Acid. The solution is concentrated, and excess of dilute sulphuric acid added ; sulphate of lead preci- pitates, which is received on a filter and washed, first with dilute sulphuric acid, and finally with alcohol ; the cadmium in the filtrate is precipitated by carbonate of potassa; the former method with cyanide of potassium gives the best results, in consequence of the partial solubility of sulphate of lead, even in dilute sulphuric acid. Separation of Oxide of Cadmium from Oxide of Silver : (1). By Hydrochloric Acid. To the diluted solution, nitric acid is added, and then slight excess of hydrochloric acid, the liquor filtered from the precipitated chloride of silver contains the cadmium, which is precipitated by carbonate of potassa. (2) By Cyanide of Potassium. The nitric solution of the two metals is rendered neutral by the addition of potassa ; cyanide of potassium is then added in such quantity that the precipi- tate, which is at first formed, is entirely redissolved. To the clear solution, excess of nitric acid is next added, whereupon the whole of the silver is precipitated as cyanide : the cad- mium remains in solution, and may either be precipitated by sulphide of ammonium, the liquid having been previously rendered slightly alkaline by the addition of potassa ; or the hy- drocyanic acid may be expelled by evaporating with sulphuric acid, and the cadmium precipitated as carbonate by carborite of potassa. Separation of Oxide of Cadmium from Oxide of Mercury. COPPEK. 379 This may be effected \)j formiate of soda, but the process is tedious, and requires considerable care. Hydrochloric acid is added to the solution, which is then nearly saturated with po- tassa ; excess of formiate of soda is then added, and the solution is set aside for some days at a temperature not above 170. The mercury precipitates in the form of calomel ; the solution filtered from the subchloride of mercury contains the cadmium, which is then precipitated by carbonate of potassa. Cyanide of potassium may also be employed to separate these two oxides in the following manner : The clear solution is ren- dered neutral by potassa, and excess of cyanide of potas- sium is then added. To the solution of the two cyanides very dilute nitric acid is added, and the whole is boiled ; the mer- cury salt is not decomposed: the cyanide of cadmium, on the other hand, is converted into nitrate, and may, consequently, be decomposed by carbonate of potassa : the filtrate, from the carbonate of cadmium, containing the oxide of mercury, may be precipitated by sulphuretted hydrogen. Separation of Oxide of Cadmium from Oxide of Bismuth. This likewise is effected by cyanide of potassium ; an excess of the cyanide is added, and heat applied, the whole of the bismuth is separated as carbonate, and the cadmium is precipitated either by sulphuretted hydrogen or by carbonate of potassn, the so- lution filtered from the carbonate of bismuth, having been pre- viously boiled with hydrochloric acid. Advantage may also be taken of the insolubility of chromate of bismuth, and the solu- bility of chromate of cadmium, to effect a separation of these two metals. Separation of Oxide of Cadmium from Oxide of Zinc (Aubel and Kamdohr). To the solution of the two oxides tartaric acid is added, and then solution of soda; it is then largely diluted and boiled for two hours, the oxide of cadmium is completely precipitated, the oxide of zinc remaining in solution, from which, after filtering off the oxide of cadmium, it is precipi- tated by sulphuretted hydrogen. The sulphide of zinc is re- dissolved in hydrochloric acid, the solution poured into a boil- ing solution of carbonate of soda, and the basic carbonate of zinc thus obtained, is converted into oxide by calcination. 178. COPPER. There are various methods of estimating this important metal. It may be weighed as oxide, as sulphide, as subiodide, 380 QUANTITATIVE ANALYSIS. and in the metallic state ; it may also be determined by various volumetric processes. (1.) Determination as Oxide. If the salt to be examined be. soluble in water, or in nitric acid, provided no organic substance be present, it is best precipitated J5y caustic potassa. The so- lution is considerably diluted, and raised to the boiling tempe- rature in a capacious porcelain basin, and the caustic alkali gradually added as long as a brownish-black precipitate is pro- duced ; the boiling is continued for a few minutes, and the pre- cipitated oxide is then received on a filter, and washed with boiling water. The greater part of it is then removed from the filter, and transferred to a platinum crucible, in which it is ignited. The filter, with its adhering oxide, is dried, burnt on the cover of the crucible, and the ashes added to the main bulk of the oxide. As oxide of copper absorbs moisture rapidly from the atmosphere, it must be weighed as soon as it is sufficiently cold to be placed in the scale of the balance, and it is advisable to allow the crucible to cool underneath a receiver by the side of a vessel containing concentrated sulphuric acid. If the so- lution be not dilute, the precipitation by caustic potassa is in- complete, as is rendered evident by the tiltrate becoming dis- coloured when mixed with sulphuretted-hydrogen water, the same is the case if any organic matter be present ; in either of these cases, the filtrate must be concentrated by evaporation, precipitated by sulphuretted hydrogen, and the precipitate treated as will.be presently described. The oxide on the filter must be well washed with boiling water, to remove all traces of the alkali, which is invariably carried down in company with the oxide, during its precipitation. The composition of oxide of copper is One equivalent of Cu . . . 31' 75 . . 79'87 One ditto of O ... 8 . . 20-13 One ditto of CuO . . . 39-75 . . 10OOO (2.) Precipitation and estimation as Sulphide. The solution is acidified with hydrochloric acid, diluted with water, and a stream of washed sulphuretted hydrogen gas passed through it till it is perfectly saturated. The precipitated sulphide is received on a filter, and washed as quickly as possible with water impreg- nated with sulphuretted hydrogen ; it is removed as completely as possible from the filter, which is then dried and ignited, and its ashes mixed with the bulk of the precipitate. The sul- COPPEK. 381 phide is decomposed by digestion with dilute nitro-hydrochloric acid, and, the sulphur being separated, the solution is evapo- rated with sulphuric acid, until the nitric acid is entirely ex- pelled ; a large quantity of water is then added, and the oxide is precipitated by caustic potassa at the boiling temperature. If the solution be neutral or alkaline, sulphide of ammonium may be employed as the precipitating agent. According to Rose, copper may be quantitatively estimated as sulphide, by calcining the precipitate by sulphuretted hydrogen in a current of hydrogen. The operation is performed in a cru- cible, through which a stream of dry hydrogen is made to pass. The heat must not be applied until the whole apparatus is full of hydrogen, and it is indispensable to continue the disengage- ment of gas until the crucible is quite cold. (3.) Estimation as Subiodide (Pisani). To the solution previ- ously freed from all metals whose iodides are insoluble, sul- phurous acid is added ; then, after having applied a gentle heat, iodide of potassium until the supernatant liquor has lost the colour due to the presence of copper, and the formation of a precipitate ceases. The subiodide of copper being very dense, settles readily, especially when heated, like chloride of silver. In this precipitation, the sulphurous acid must always be in a slight excess to avoid the formation of the brown compound. After having heated the liquor nearly to boiling, it is filtered on a counterpoised filter ; the precipitate is washed with hot water and dried ; it is heated with the filter to from 230 to 248 P., after which the subiodide is weighed, and from its weight the quantity of metal is calculated. The composition of subiodide of copper is Two equivalents of Cu . . . 63'5 . . 33 -33 One ditto of I ... 127'Q . . 66'67 One ditto of Cu 2 I . . 190-5 . . 100 00 (4.) Estimation as metallic Copper in ores. (F. Mohr, 'Re- pertoire de Chimie pure et appliquee,' and Chem. News, JSov. 8th, 1862.) (a.) Oxygenated ores, comprising Oxide, Siiboxide, 3falachite } and Phosphate of Copper. According to the richness of the ore to be assayed, from 75 to 150 grains of the mineral reduced to an impalpable powder, are heated with weak sulphuric and nitric acids. The mixture is boiled, evaporated to dryness, and then calcined until it ceases to disengage vapours. The copper be- 882 QUANTITATIVE ANALYSIS. comes transformed into sulphate, not decomposable by moderate heat, and very soluble in sulphuric acid. On the contrary, iron is converted into slightly soluble ferric subsulphate, and lead into insoluble sulphate. With tin and antimony, nitric acid forms oxides which resist the action of solvents. When the capsule is cool, distilled water is added and the solution boiled. It is then filtered, the nitrate contains all the copper, and small quantities of ferric sulphate. On treating this solution with zinc, the ferric salt is reduced, and the copper precipitated by boiling, in the metallic state. To make sure of the complete precipitation of the metal, it suffices to add a little solution of sulphuretted hydrogen to a drop of the supernatant liquid. To free the copper from excess of zinc, the powder is treated with hydrochloric acid, until it ceases to disengage bubbles of hydro- gen gas. The metal freed from zinc is washed with hot dis- tilled water ; the action of the air must be avoided when the solution is acid, because a portion of copper would dissolve ; but it is easy to avoid this inconvenience by effecting the wash- ing promptly. The reduced copper is dried in the water oven till it ceases to lose weight. The presence in the ore, of zinc, manganese, iron, cobalt, and nickel does not affect the precision of the analysis, neither of these metals being precipitated by zinc. (b.) Sulphuretted ores. The treatment by sulphuric and nitric acid must be several times repeated, drying and calcining the metal between each treatment with acid. This method, the author thinks, has the advantage of being applicable to all ores, whatever their composition, and of not necessitating a prelimi- nary qualitative analysis. In the second place, the copper is separated from metals (lead, tin, and anlimony), from which it is freed with difficulty by ordinary methods, whilst four metals (zinc, iron, cobalt, nickel) do not influence the exactness of the estimation, and may be disregarded. LoveVs method of determining Copper. This is a modifica- tion of the method proposed by Fuch for the quantitative esti- mation of iron (see page 313). The cuprous solution is intro- duced into a fl isk that can be accurately closed with a glass stopper, ammonia is added, till the liquid assumes a transparent blue colour, and the flask is then filled with water, from which all atmospheric air has been expelled by boiling; a clean and accurately weighed slip of copper is introduced into the bottle, which is immediately closed; when the liquor has become per- COPPER. 383 fectly colourless, the slip is removed, washed, dried and weighed, the diminution in weight which it has undergone, indicates the amount of copper originally present in the solution. The re- sult which, when properly conducted, is very accurate, depends on the abstraction of one equivalent of copper from the slip by every equivalent of oxide of copper in the solution to form an equivalent of suboxide of copper, which forms w r ith ammonia a colourless solution, thus: It requires a considerable time to complete the process, which is obviously altogether inapplicable in the presence of foreign metals, which are capable of being precipitated by copper. Cassasecd's method. This consists in dissolving the copper compound in an acid, adding an excess of ammonia to the so- lution, and comparing the tint furnished by this solution, with that which a known weight of pure copper yields likewise in the state of ammoniuret. (5.) Volumetric methods of estimating Copper : (a.) Pelouze's method (' Comptes Keiidus,' Feb. 12, 1846). This mode of analysis was suggested to the author, by the ac- curacy and rapidity with which alloys of silver are analysed by the process discovered by M. Gay-Lussac. He succeeded in ef- fecting his object in several different ways, all based principally on the phenomena of precipitation, and simultaneous decolora- tion. The following was the mode of proceeding which he finally adopted : a certain quantity of very pure copper is dissolved in nitric acid, the solution is diluted with water, and excess of am- monia added ; a deep blue solution is obtained. On the other hand, some sulphide of sodium is dissolved in water, and poured into a tube graduated and divided into tenths of a cubic centi- metre; the ammoniacal solution is heated to boiling, and the solu- tion of the sulphide gradually added. If we suppose that it re- quired 31 cubic centimetres to decolorize 1 gramme of copper, we have a standard solution of known strength. To apply this to the analysis of copper alloys, a certain known weight is dissolved in aqua-regia, the solution is supersaturated with ammonia, heated to boiling, and the standard solution of the sulphide added until it is decolorized, taking care to add from time to time, a little dilute ammonia to replace that which is evaporated. The decrease in the depths of the blue tint, points out that the end of the experiment is more or less near, and when it is re- quisite to add the last portions of the sulphide in drops. When 384 QUANTITATIVE ANALYSIS. the operation is supposed to be finished, the number of divisions employed for the decoloration is read off, and compared with the number required to decolorize an equal weight of pure copper. It must be remarked that the ammoniacal liquor from which the copper has been precipitated does not long remain colourless, but gradually becomes blue in consequence of the sulphide of copper becoming partially converted into sulphate by the absorption of oxygen. This mode of operating is not, according to the author, liable to an error amounting to more than T -oVo or T^oVoj though still greater accuracy is obtained by completing the decoloration of the blue liquid with a very weak solution of sulphide, precisely in the manner recommended by Gay-Lussac in his 'Analysis of Silver Alloys by Standard So- lutions of Common Salt.' Neither tin, zinc, cadmium, lead, an- timony, iron, arsenic, nor bismuth in anyway interfere with the success of this process, not being in the least affected by the sulphide of sodium while a trace of copper remains to be pre- cipitated ; indeed the author found that when the sulphides of zinc, cadmium, tin, lead, bismuth, and antimony are placed in contact with ammoniacal solution of sulphate of copper, they de- colorize it, some in the cold, others with the assistance of heat, which proves very evidently, that these sulphides cannot exist, except perhaps for an instant, in a solution of copper. Their formation subsequently to the decoloration, has no influence on the result of the analysis, as the termination of this is judged of by the decoloration of the liquid, without paying the least attention to the precipitates which subsequently form ; or, if any attention is paid, it is only with a view to obtain some knowledge of the nature of the metals which accompany the copper. Thus, if any alloy consists of copper, lead, tin, and zinc, the presence of zinc is readily detected by the white pre- cipitate which succeeds the black precipitate of sulphide of copper, the lead and tin being precipitated at the outset by ammonia. Cadmium, like zinc, is precipitated immmediately after the copper. The very moment the liquid is observed to be decolorized, a beautiful pure yellow precipitate of sulphide of cadmium is formed, if the addition of sulphide be continued. If the alloy contain silver, that metal is previously precipitated from the nitric solution by hydrochloric acid. In this method of estimating copper, an important property of ammonia, besides that of heightening the colour, is that it prevents the salts of copper being precipitated by sulphites and liyposul- COPPER. 385 pJiitcs, without which it would probably have been impossible to estimate the copper by means of solutions of the alkaline sulphides, since these salts almost always occur in the alkaline sulphide, and are moreover produced from them by contact with the air. A solution of sulphide of sodium becomes weaker by con- tact with the air, but the alteration is very slow, nor is it neces- sary to change the liquid as long as any remains in the flask in which a quantity has been prepared. The only precaution to be taken, and it is one which applies to all standard solutions, is to determine previously to each assay, the actual strength of the sulphide with a known weight of pure copper. Pelouze states in conclusion, that this method, applied to the analysis of copper ores, yields results of the greatest accuracy. (b.) C. Mohr's method (Liebig's 'Annalen,' xcii. p. 97). This is founded on the fact that salts of protoxide of copper are pre- cipitated by metallic iron, the latter being converted into pro- toxide. The quantity of the protosalt of iron is determined by permanganate of potash. The solution of the copper salt is put with a few drops of hydrochloric acid, and about one-fourth of pure chloride of sodium, into a stoppered bottle ; a quantity of soft iron wire is then introduced. , The reduction immediately commences, and should be assisted by a heat of from 89 to 100 F. All the copper is separated in a metallic form in an hour or two, when no trace of copper can be detected in the solution by sulphu- retted hydrogen. The following precautions must be observed ; the solution must not be too acid, as in that case an excess of iron is dissolved; and the heat applied must not be too strong, as this causes the separation of a basic protosalt of iron in the form of a flocculent precipitate, which has no action upon the permanganate of potassa ; when the reduction is completed, which may be known by the clearness of the fluid, a protosalt of iron has taken the place of the protosalt of copper, according to the formula CuO + S0 3 + Fe = FeO + S0 3 + Cu. The fluid with the separated pulverulent copper is then diluted, and a measured quantity drawn off by a pipette and treated with permanganate of potassa. (c.) Schwartz's method (Annal. der Chemie und Pharm. 84). This is based on the fact that suboxide of copper reduces iron from the state of sesquichloride to that of protochloride, thus : 386 QUANTITATIVE ANALYSIS. To a solution of the compound in nitric acid, or in water contained in a porcelain basin, is added a solution of neutral tartrate of potassa, and then potassa or soda in excess ; the blue fluid obtained, is warmed on the water-bath with an aqueous solution of grape or milk sugar ; the action is allowed to con- tinue until the blue fluid begins to turn brown, the precipitate, consisting of suboxide of copper, is allowed to subside, and then filtered and washed with hot water, till the washing water passes through colourless. The filter with its contents are then trans- ferred to the porcelain basin, and pure sesquichloride of iron added in slight excess, together with a little hydrochloric acid, gentle heat is applied, and when the subchloride of copper which is at first formed is redissolved, the solution is filtered into a capacious flask, and the remains on the filter well washed ; it is then allowed to cool to about 77, and the amount of pro- tochloride of iron determined by a standard solution of perman- ganate of potassa. Every 28 parts of iron found in the state of protochloride, indicate 81*75 parts of copper. (d.) Fleitmann 's method (Liebig's 'Annalen,' April, 1856, p. 141). When the solution of copper is free from nitric acid, or injurious metals, such as antimony and arsenic, the copper is precipitated with pure metallic zinc, the excess of which metal is got rid of by digestion with pure sulphuric acid ; the precipitate is washed, and dissolved in an acid solution of pure perchloride of iron. The solution of the copper takes place almost instan- taneously, and furnishes double its equivalent of protoxide of iron, which is determined by permanganate of potassa ; when nitric acid is present, ammonia is added in excess, and the pre- cipitation of the copper effected in the filtered solution, by pure zinc shavings. (e.) Terrell's method (' Comptes Eendus,' Feb. 1, 1858). The cuprous mineral is dissolved in nitric acid, which is completely driven off by concentrated sulphuric acid, the nitrates being converted into sulphates, ammonia is added in excess, and the liquid filtered ; it is then boiled with sulphite of ammonia until colourless, and the excess of sulphurous acid is driven oft' by boiling with hydrochloric acid, the solution is then diluted with water, and treated with a standard solution of permanganate of potassa. (f.) By Cyanide of Potassium (Parker and Mohr). The sub- stance is dissolved in an acid, and ammonia added in excess ; a normal solution of cyanide of potassium is then added from a COPPER. 387 burette until the blue colour disappears; two equivalents of cyanogen are necessary to decolorize one of copper in ammonia. In this method Field, who has subjected the various processes for determining copper to a critical examination (Chem. News, vol. i. pp. 24, 61, 73), remarks that great caution and consider- able practice are required to determine the necessary amount of cyanide, as towards the end of the process, the decolorization is not very distinct, the solution assuming a delicate violet tint which fades very gradually, leaving at length the liquid destitute of colour ; it is safer to leave a very slight tint, as it disappears after the lapse of twenty-four hours. If iron be present, it is better not to filter off the peroxide precipitated by ammonia, as it is exceedingly difficult to wash the copper out of it, and the estimation of the copper can be perfectly effected in the solution, in its presence. The solution of cyanide recom- mended by Field is 1300 grains of the salt dissolved in four pints of water, about 50 grains of the mineral being employed ; with alloys of tin and copper, antimony and copper, and ar- senic and copper, this method is available ; but it cannot be ap- plied to the analysis of alloys of copper and zinc, or of copper and silver; the presence of nickel and cobalt likewise interferes with its accuracy. Fleck (Polytech. Centralblatt, 1859, p. 1313) recommends to dissolve the copper compound in carbo- nate of ammonia instead of ammonia, and to add a drop of ferro- cyanide of potassium to the blue liquor. The moment the cupro- ammoniacal compound is destroyed, the liquid becomes red. With regard to the precipitation of copper by iron or zinc, Field remarks that though the estimation is never perfectly correct, traces both of zinc and iron being found with the pre- cipitated copper, however carefully the operation is performed, nevertheless in a commercial point of view it is not far from the truth ; but the precipitated copper should be washed with water at the temperature of 100 or 120, to which one or two per cent, of hydrochloric acid has been added ; if a stronger acid be used there is a danger of some of the finely precipitated copper being dissolved. (g.) Streng's method. The oxide of copper is reduced by grape sugar, a solution of starch and iodide of potassium added, afterwards a standard solution of bichromate of potassa ; the following reaction occurs : 3Cu 2 0+2Cr0 3 =6CuO-|-Cr 2 03. "When the whole of the subchloride of copper is converted 388 QUANTITATIVE ANALYSIS. into chloride, a permanent blue colour is produced ; the chromic acid then reacting on the iodide of potassium and expelling iodine. Field finds this method to give accurate results, pro- vided care be taken in the addition of the iodide of potassium. Diniodide of copper is very insoluble in hydrochloric acid, unless the latter be in excess, so that when a considerable quantity of alkaline iodide is employed, and little hydrochloric acid present, there is not much dichloride of copper in solution. MM. Plessy and Moreauhave founded a method of estima- ting copper on the following reaction : CuCl+Cu=Cu 2 Cl. The solution of protochloride is made as nearly neutral as possible by the addition of ammonia, and the liquor is then rendered green by the cautious addition of hydrochloric acid : a strip of copper is boiled in this solution until it becomes colourless ; the loss of weight indicates the amount of metal in the chloride. E. 0. Brown (Quart. Journ. Chem. Soc. vol. x. p. 65) dis- solves the copper compound in nitric acid, adds carbonate of soda, and afterwards acetic acid in excess. Iodide of potassium is then added, equal to at least six times the weight of the copper. A standard solution of hyposulphite of soda is poured into the flask from a burette, until the brown colour nearly dis- appears. Clear starch liquid is then introduced, and a further addition of hyposulphite, until the blue colour is destroyed. The presence of peroxide of iron is fatal to this method, not only on account of the deep colour of the peracetate of iron, but more particularly because the peroxide becomes partially deoxidized by the hyposulphite, and thus interferes with the reaction. In concluding his valuable review of the various methods of estimating copper, Field remarks that much must be left to the knowledge and experience of the operator. In a mixture of lead, arsenic, and copper, the cyanide of potassium process could be advantageously adopted, whilst the method by pre- cipitation would be worthless, as all the metals would be reduced. Manganese and zinc do not, on the other hand, affect the pre- cipitation of copper upon iron, but render the cyanide estima- tion valueless. Kunsel has recently proposed the following modification of Pelouze's volumetric method for the estimation of copper: Sulphide of zinc is employed for indicating the complete pre- COPPER. 389 cipitation of the copper, that substance being instantly decom- posed in a hot ammoniacal solution of copper. The solution of sulphide of sodium is of such a strength that one cubic centi- metre precipitates a centigramme of copper. The sulphide of zinc is prepared by dissolving the metal in hydrochloric acid, supersaturating with ammonia, and then boiling with a little sulphide of zinc to remove the lead, which is always present in commercial zinc. The ammoniacal solution of zinc is filtered and decomposed with sulphide of sodium, a small quantity of zinc being allowed to remain in solution. The moist sulphide of zinc, with excess of zinc solution, is then spread evenly upon filter-paper, several layers thick ; when the paper has ab- sorbed most of the solution, the moist white layer of sulphide of zinc is ready for use. According to 1'ield, the most expeditious method of converting protosalts of copper into subsalts by means of alkaline sulphites, is to mix about equal quantities of sulphite and carbonate of soda, and to make of these salts a strong cold solution. This liquid poured into a protosalt of copper, instantly decomposes it. After a few minutes' boiling, the whole is converted into suboxide, which dissolves in hydrochloric acid or ammonia with- out a shade of colour. Separation of Oxide of Copper from Oxide of Lead : (1.) By Cyanide of Potassium. To the diluted solution of the two oxides, carbonate of soda is added in slight excess, and then cyanide of potassium, and heat applied. The lead precipitates as carbonate ; the copper remains in solution in the form of double cyanide of copper and potassium : it is evaporated with sulphuric acid till all the hydrocyanic acid is expelled, then largely diluted with water, and precipitated at the boiling temperature by caustic potassa. (2.) By Sulphuric Acid. The solution is concentrated, and the oxide of lead precipitated by dilute sulphuric acid, the sul- phate of lead is collected on a filter, and washed first with di- lute sulphuric acid, and then with spirits of wine. The filtrate is boiled for some time, to drive off the alcohol, and finally pre- cipitated by caustic potassa. Separation of Oxide of Copper from Oxide of Silver : (1.) By Hydrochloric Acid. Nitric acid is added to the solu- tion, the silver is then precipitated as chloride, by hydrochloric acid, and the oxide of copper is thrown down from the filtrate bv caustic potassa. J 390 QUANTITATIVE ANALYSIS. (2.) By Cyanide of Potassium. The solution is neutralized with potassa, cyanide of potassium is then added till the pre- cipitate which is first formed is entirely redissolved. To the clear solution nitric acid is added, which precipitates the silver completely as cyanide. The filtrate is evaporated with sul- phuric acid till all the "hydrocyanic acid is expelled, and the copper is finally precipitated by caustic potassa. Fresenius gives another method of treating the solution of the oxides of these two metals in cyanide of potassium. Sulphuretted hy- drogen is passed into the solution, which precipitates only the silver, provided sufficient cyanide of potassium he present. The solution filtered from the sulphide of silver is heated to expel the excess of sulphuretted hydrogen ; cyanide of potas- sium is again added, and then, having completely decomposed this salt by evaporating with a mixture of sulphuric and nitric acids, the copper is precipitated by caustic potassa. (3.) Berlandt (Archiv fur Pharm., Band civ. 279) dissolves the mixed metals in nitric acid, and evaporates the solution to dry- ness, to get rid of excess of acid. He then dissolves one ounce of the salts in five ounces of water, filters the solution, adds fourteen ounces of a solution of protosulphate of iron (five and a half parts sulphate to eight and a half parts water) mixes, and stirs well. The greyish-white deposit, washed with very dilute sulphuric acid, and afterwards with water, is found to be pure silver. For the various methods of assaying alloys of silver and copper, see " SILVER." Separation of Oxide of Copper from Oxide of Mercury : (1.) By Cyanide of Potassium. This is effected in the same manner as the separation of oxide of silver from oxide of copper; the solution of the two oxides in cyanide of potassium is treated with sulphuretted hydrogen, by which mercury alone is preci- pitated. (2.) By Formiate of Soda. Hydrochloric acid is added to the solution, which is then nearly neutralized with potassa, and the mercury precipitated as subchloride by formiate of soda, in the manner directed in treating of the separation of oxide of mercury from oxide of cadmium. Separation of Oxide of Copper from Oxide of Bismuth : f 1.) By Carbonate of Ammonia. On adding this reagent to tne solution of the two oxides, in considerable excess, oxide of bismuth alone is precipitated ; it is allowed to remain at rest COPPER. 391 for some time in a warm place, and then filtered, the oxide of bismuth being washed on the filter with carbonate of ammonia. The filtrate is gently evaporated to expel the excess of carbonate of ammonia, caustic ammonia is then added, and the oxide of copper is finally precipitated by caustic potassa. This method is not to be recommended, it being difficult to remove all traces of oxide of copper from the precipitated oxide of bismuth, even by protracted washing with carbonate of ammonia. (2.) By Cyanide of Potassium. Carbonate of soda is first added in slight excess to the solution, the oxide of bismuth is then precipitated in the form of carbonate, by heating with cyanide of potassium ; the copper in the filtrate is determined by precipitating it by caustic potassa, having previously expelled the hydrocyanic acid by evaporating with sulphuric acid. (3.) By Chlorine. The solution of the two metallic oxides is precipitated by sulphuretted hydrogen, and the mixed sul- phides, having been washed and weighed, are introduced into a bulb, blown in a tube of hard glass, connected with an appara- tus for generating chlorine. A stream of this gas is trans- mitted through the tube, which at the same time is heated by a spirit lamp, first gently, and then to redness ; the sulphides are thus converted into chlorides, and the chloride of bismuth, being volatile, is gradually expelled ; it may be received into a vessel containing water impregnated with hydrochloric acid, or the amount of oxide may be estimated by the loss of weight sustained by the analysed substance. The apparatus employed may be the same as that used for separating mercury from lead, mercury from silver, or bismuth from lead. The chloride of copper remaining in the bulb is washed out with a little dilute nitric acid, evaporated with sulphuric acid, and finally precipi- tated by caustic potassa. Separation of Oxide of Copper from Oxide of Cadmium : (1.) By Carbonate of Ammonia. This reagent is added in ex- cess to the solution of the two oxides ; carbonate of cadmium is precipitated, while oxide of copper remains in solution. (2.) By Cyanide of Potassium. This reagent is added until a clear solution is obtained, sulphuretted hydrogen is then passed through the mixture, by which the cadmio-cyanide is completely decomposed, sulphide of cadmium being precipi- tated, while the whole of the sulphide of copper remains in solu- tion. The excess of sulphuretted hydrogen is expelled by heat, and a little more alkaline cyanide added. The copper may then 392 QUANTITATIVE ANALYSIS. be precipitated as sulphide, by the addition of an acid ; or the double cyanide may be decomposed by evaporating with sul- phuric acid, and the metal precipitated as oxide by caustic potassa. (3.) By Iodide of Potassium (Pisani). The copper is pre- cipitated as iodide, by the addition, first of sulphurous acid, a gentle heat being applied, and then of iodide of potassium until the supernatant fluid has lost the colour due to the pre- sence of copper, and the formation of a precipitate ceases. From the filtered liquor the cadmium is precipitated by sul- phuretted hydrogen. (4.) The metals are precipitated as sulphides, the sulphides are boiled with dilute sulphuric acid (one part concentrated acid, and five water) and rapidly filtered. The filtrate con- tains the whole of the cadmium, which may then be precipitated by sulphuretted hydrogen. Separation of Oxide of Copper from Oxide of Zinc : Analysis of Brass. Although oxide of zinc is, when alone, completely soluble in caustic potassa, this reagent cannot safely be em- ployed to separate zinc from copper, since the oxide of copper invariably carries down with it a greater or lesser quantity of oxide of zinc. Sulphuretted hydrogen is, however, an effectual reagent for this purpose. (1.) The alloy, being dissolved in nitric acid, is treated with a current of the gas ; sulphide of copper alone precipitates, and from the filtrate the oxide of zinc may be completely thrown down, by boiling with carbonate of potassa or by evaporating it to dry ness, and igniting in a platinum crucible. (2.) The copper is precipitated by sulphurous acid and iodide of potassium, according to Pisani's method, and the zinc in the filtrate, by carbonate of soda or potassa. (3) MM. Eivot and Bouquet's method. The alloy is dissolved in nitric acid, diluted with water, and saturated with ammonia. Into the ammoniacal solution is now put a slight excess of pure potassa in fragments, and the whole is gently heated on a sand-bath until it is completely decolorized, and until the liquor no longer smells of ammonia. The oxide of copper is now thrown on a filter, and washed with boiling water. Into the alkaline liquor hydrochloric acid is poured, until the whole dis- plays an acid reaction, when the zinc is precipitated by car- bonate of soda. Before filtration, the solution is heate'd du- riug seven or eight hours on a sand-bath, for the purpose of COPPER. 393 eliminating free carbonic acid. The precipitate is DOW filtered, washed with boiling water, separated from the filter, and cal- cined. The two metals are thus estimated in the state of oxide. Separation of Copper from Nickel. (Dewilde, 'Bulletin de la Societe Chimique de Paris.') Dissolve about 30 grains of the alloy in hydrochloric acid with the addition of nitric acid ; eva- porate the excess of acid, and dissolve the chlorides iti about two pints of water. To the solution add pure cream of tartar, double the weight of the alloy. Heat slightly to favour the solution, and add, little by little, a solution of caustic potassa in alcohol. The metals are precipitated as hydrates, which are redissolved by an excess of the alkali. The blue liquid is boiled for a few minutes with a solution of grape sugar ; the copper is hereby precipitated as suboxide, which is washed, dried, and calcined, and then converted into nitrate, and the copper esti- mated by one of the volumetric methods. The filtered solution containing the nickel is evaporated to dryness, the residue in- cinerated, then washed to remove the carbonate of potassa, and a second time incinerated and washed, it is then redissolved in aqua-regia from which the hydrated oxide is precipitated by potassa. As it is very difficult thoroughly to wash this volu- minous hydrate, Dewilde prefers, where great accuracy is re- quired, to dry, caleine, and then after another washing with hot water, to reduce it to metallic nickel in a platinum crucible in an atmosphere of hydrogen gas. Detection and estimation of small quantities of Antimony, Arsenic, Bismuth, and Lead in metallic Copper : Analysis of Commercial Copper. (Abel and Field, Quart. Journ. Chem. Soc., January, 1862.) (a.) Determination of Antimony and Arsenic. Two hundred grains of the metal are dissolved in nitric acid, a small quantity of solution of nitrate of lead is added, equal to about ten grains of the salt, and subsequently an excess of ammonia arid carbon- ate of ammonia. A precipitate is formed which may consist of oxide and carbonate of lead, arsenal e and antimoniate of lead, and oxide of bismuth, the whole of the copper remaining in so- lution. The precipitate is separated by filtration, thoroughly washed, and digested in a strong solution of oxalic acid, whereby the antimony and arsenic are dissolved. To the filtered liquid sulphide of ammonium is added, or what is preferable, it is ren- dered alkaline by ammonia, and hydrosulphuric acid passed through to saturation. Traces of sulphide of copper generally 394 QUANTITATIVE ANALYSIS. impart a greenish tinge to the liquid, and are deposited after some time, as it is nearly impossible to wash away the last traces of that metal, from the nitrate of lead precipitate. This sulphide is filtered off, and a slight excess of hydrochloric acid is added to the filtrate which is diluted to about 8 ounces. If any large amount of either antimony or arsenic be present, there will be an immediate precipitate. If smaller quantities exist (one or two-hundredths of a grain), the flask should be placed on the sand bath for a few hours and the temperature maintained at from 140 to 200 F., when the metals, if present, will make their appearance as sulphides. If the precipitate be orange or orange-red, the presence of antimony is certain; but if a pure canary-yellow, its absence may be presumed. The precipitated sulphides are oxidized by means of concentrated nitro-hydro- chloric acid, the clear solution is mixed with chloride of ammo- nium and excess of ammonia, and the arsenic is separated as ammonio-magnesian arsenate (2MgO,jSTH 4 O, AsO 5 ,HO). The filtrate from this, is slightly acidified with dilute hydrochloric acid, and the antimony is precipitated by hydrochloric acid, and ultimately determined as antimoniate of teroxide of antimony (Sb0 4 ). Should the presence of antimony and arsenic have been previously ascertained, it is of course unnecessary in the quantitative process, to precipitate them both as sulphides from their solution in oxalic acid, as the arsenic may be at once pre- cipitated by the addition of sulphate of magnesia and excess of ammonia, and the antimony determined in the filtrate. (b.) Determination of Lead and Bismuth. The nitric acid solution of about 200 grains of the copper is mixed with a small quantity of solution of phosphate of soda ; ammonia in excess is then added, and the resulting precipitate is collected and purified from copper, by washing with ammoniacal water. The precipitate is afterwards dissolved in hydrochloric acid, its so- lution is rendered alkaline with ammonia, and submitted to a current of sulphuretted hydrogen. The precipitated sulphides of lead and bismuth are thoroughly washed and dissolved in dilute nitric acid. The solution is nearly neutralized with ammonia, and then digested with a little hydrated oxide, or basic nitrate of copper, which precipitates the oxide of bismuth, while the lead remains in solution. The precipitate is tho- roughly washed, dissolved in dilute nitric acid, and the bismuth separated from the copper by the addition of excess of ammonia. The oxide of bismuth thus obtained is purified by washing, and COPPER. 395 its weight determined in the usual manner. The solution con- taining the nitrates of lead and copper is mixed with solution of carbonate of soda ; excess of acetic acid and a small quantity of bichromate of potassa are added, and the chromate of lead is collected. Should the copper contain iron, that metal would be obtained as oxide together with the bismuth. When this is the case, which may readily be known by the brownish tinge imparted to the oxide of bismuth, they must be separated. An extremely delicate method of testing copper for bismuth is the following, founded upon a curious reaction exhibited by iodide of potassium in the joint presence of lead and bismuth, first noticed by Field. About 100 grains of the copper to be examined are dissolved in nitric acid, a solution of nitrate of lead equal to about 5 grains of the salt is added, and subse- quently ammonia and carbonate of ammonia. The precipitate is washed with ammoniacal water to free it from copper, and dissolved in warm acetic acid. Considerable excess of iodide of potassium is introduced, and the liquid is warmed until the precipitate disappears. On cooling, crystalline scales make their appearance, which by their colour indicate the presence or absence of bismuth. If that metal be absent, the scales are brilliant gold-colour, but if it be present, or even the slightest trace, they assume a dark orange or crimson tint, varying in intensity of colour according to the amount of bismuth present. If it be desired to test the copper for arsenic, the nitrate of lead precipitate is digested with acetic acid. The oxalates of lead and bismuth are insoluble in acetic acid, but it dissolves the arsenic. Analysis of a mixture of Oxides of Lead, Bismuth, Silver, Copper, Mercury, and Cadmium. To a diluted solution, car- bonate of potassa is first added, and then excess of cyanide of potassium ; the oxides of lead and bismuth are precipitated in the form of carbonates ; they are received on a filter, washed, dissolved in nitric acid, and the oxide of lead separated from the oxide of bismuth by dilute sulphuric acid, as directed (p. 371). The washing from the precipitated carbonates being mixed with the filtrate, excess of diluted nitric acid is added, the silver is precipitated as cyanide, in the form of which salt it is estimated. The filtrate from the cyanide of silver, toge- ther with the washings, are again neutralized by carbonate of potassa, a fresh quantity of cyanide of potassium is added, and a stream of sulphuretted hydrogen gas is passed through the 396 QUANTITATIVE ANALYSIS. solution, the mercury and the cadmium are precipitated as sul- phides : the whole of the sulphide of copper is retained in so- lution by the cyanide of potassium, provided a sufficient quan- tity of that reagent has been added ; to ensure this, it is advi- sable to add a fresh portion after the action of the sulphuretted hydrogen. The precipitated sulphides of mercury and cadmium, after being well washed on the filter, are decomposed by diges- tion with aqua-regia, the solution is filtered off from the sul- phur, nearly neutralized with potassa, and the mercury esti- mated as subchloride. The cadmium in the solution, filtered from the subchloride of mercury, is precipitated by carbonate of soda. The sulphide of copper, which remains dissolved in the cyanide of potassium, is mixed with nitric acid and evaporated with the addition of sulphuric acid, until the whole of the hydrocyanic acid is expelled ; the sulphate of copper retained in solution is finally precipitated at a boiling temperature by caustic potassa. 179. PALLADIUM. This metal is remarkable for its great affinity for cyanogen, on which property is founded a method of separating it from its solutions; cyanide of mercury is added, the solution having pre- viously been neutralized with soda ; a bright yellow precipitate is produced, which, by drying, becomes yellowish-grey, and by ignition is decomposed, metallic palladium of a blue colour re- maining in the crucible. Prom all metals which are not preci- pitated by sulphuretted hydrogen, palladium may be separated by that reagent ; the resulting sulphide is converted by heat into basic sulphate, which, being dissolved in hydrochloric acid, and neutralized with soda, is precipitated by cyanide of mer- cury. Separation of Palladium from Copper. In crude platinum ores palladium occurs in combination with copper. Berzelius gives the following process for separating these two metals. (Poggendorfs ' Annalen,' Bandxiii. p. 561.) Both metals are precipitated from an acid solution by sulphuretted hydrogen. The precipitated sulphides are exposed to heat, while still moist, and roasted as long as they give off sulphurous acid ; they are thereby converted into basic sulphates of oxides. These salts are dissolved in hydrochloric acid, the solution is mixed with chloride of potassium and nitric acid, and then evaporated to .drymess. The dark saline mass thus produced contains chloride PALLADIUM. 397 of potassium, chloride of copper and potassium, and chloride of palladium and potassium. The first two of these salts are to be extracted by alcohol, specific gravity O833 ; the palla- dium salt, being insoluble therein, remains behind. It is brought on a weighed filter, and washed with alcohol ; it is then dried and weighed; it contains 28'84 percent, of palladium. The saline mass may also be dissolved in water, and precipitated with cyanide of mercury. The alcoholic solution of the copper salt is evaporated to expel the spirit, the saline mass is re- dissolved in water, and the copper precipitated by caustic po- tassa. Palladium has been imported into this country from Brazil, alloyed with gold, some specimens containing 5 or 6 per cent, of palladium. The operation of refining is thus described by Mr. Cock (Proc. Chem. Soc., vol. i. p. 162). The gold dust is fused, in charges of about 7 Ibs. troy, with its own weight of silver, and a certain quantity of nitre ; the effect of this fusion is to remove all earthy matter, and the greater part of the base metals contained in the gold dust, and in the silver melted w r ith it. The fused mixture is cast into ingot moulds, and when cooled, the flux or scoria3 is detached. Two of the bars thus ob- tained are then remelted in a plumbago crucible, with such an addition of silver as will afford an alloy, containing one-fourth its weight of pure gold, and which, first being well stirred to ensure a complete mixture, is poured through a perforated iron ladle into cold water, and thus very finely granulated. It is then ready for the process of parting. For this purpose about 25 Ibs. of the granulated alloy are placed in a porcelain jar upon a heated sand-bath, and subjected to the action of about 25 Ibs. of pure nitric acid, diluted with its own bulk of water ; after the action of this quantity of acid, the parting of the gold is very nearly effected ; but, to remove the least portions of silver, etc., about 9 or 10 Ibs. of strong nitric acid are boiled upon the gold for two hours. It is then completely refined, and, after being washed with hot water, is dried and melted into bars containing 15 Ibs. each. The nitrous gas, and the vapour of nitric acid arising during the above process, are conducted by glass pipes (connected with the covers of the jars) into a long stoneware pipe, one end of which slopes downwards into a receiver for the condensed acid, the other end being inserted into the flue for the purpose of carrying off the uncondensed gas. 398 QUANTITATIVE ANALYSIS. The nitrate of silver and palladium, obtained as above, is carefully decanted into large pans, containing a sufficient quan- tity of solution of common salt to effect the precipitation as chloride, of the whole of the silver, the palladium and copper remaining in solution in the mother-liquor, which is drawn off, and, when clear, run off, together with the subsequent wash- ings from the chloride of silver, into wooden vessels, and the metallic contents are then separated in the form of a black powder, by precipitation with sheet zinc, assisted by sulphuric acid. The chloride of silver, when washed clean, is reduced by the addition of granulated zinc and dilute sulphuric acid, washed on the filter with boiling water, dried, and melted in plumbago crucibles, without the addition of any flux. From the black powder obtained as above, the palladium is extracted byre-solution in nitric acid and supersaturation with ammonia, by which the oxides of palladium and copper are first precipi- tated and then redissolved, while those of iron, lead, etc., re- main insoluble. To the clear ammoniacal solution, hydro- chloric acid in excess is then added, which occasions a copious precipitation of the yellow ammonio-chloride of palladium, from which, after sufficiently washing it with cold water and ignition, pure metallic palladium is obtained. The mother- liquor and washings contain all the copper and some palladium, which are recovered by precipitation with iron. 180. BHODIUM. This metal is. according to Berzelius, (PoggendorfFs Annal., Band xiii. p. 254) best estimated by the following process: The solution is mixed with excess of carbonate of soda, and eva- porated to dryness ; the dry residue is then ignited in a plati- num crucible. Upon dissolving the mass in water, peroxide of rhodium remains behind, which is brought upon a filter, and washed, first with hydrochloric acid, and finally with water. It is then ignited with the filter, and subsequently reduced by hydrogen gas. The reduction is so easily effected, that it is scarcely necessary to assist the action of the gas by the appli- cation of heat. Separation of Rhodium from Copper. Berzelius directs to pour the solution into a flask which is furnished with a glass stopper, and to saturate it with sulphuretted hydrogen gas. The flask is then closed and allowed to remain for 'twelve hours in a warm situation. The sulphide of copper is, in that time RHODIUM. 399 fully, and the sulphide of rhodium, for the most part, precipi- tated. The solution is filtered, and, being heated and eva- porated, yields a fresh portion of sulphide of rhodium, which is added to the other sulphides. These are placed, while still moist, in a platinum crucible, and roasted as long as sulphurous acid exhales. When the roasting is finished, the mass is sub- jected to the action of concentrated hydrochloric acid, peroxide of rhodium remains undissolved, and is reduced by hydrogen gas ; the copper in the solution is precipitated by caustic po- tassa. Rhodium exists together with iridium, palladium, osmium, etc., in platinum ores. It is extracted from the liquor from which the palladium has been precipitated by cyanide of mer- cury, by the following process. Hydrochloric acid is added, and the solution is evaporated to dryness, the excess of cyanide of mercury is decomposed, and transformed into chloride. The dry saline mass is reduced to a very fine powder and washed with alcohol, sp. gr. O837. The double chlorides of sodium and platinum, sodium and iridium, sodium and copper, and so- dium and mercury, are dissolved, but the double chloride of sodium and rhodium remains behind in the form of a fine red powder. It is washed with alcohol and decomposed by gently heating in a current of hydrogen gas, by which the chloride of rhodium is reduced, and 'the metal is subsequently separated from the chloride of sodium by water. Another method of treating the saline mass is to mix it with about twice its weight of carbonate of potassa, and to calcine the mixture. The residue is treated with water, and the copper dissolved out by hydrochloric acid. The residual mass is next mixed carefully with five times its weight of anhydrous bisulphate of potassa and heated to redness in a well-covered platinum crucible; the heat is continued till the mixture is about to solidify. The oxide of rhodium dissolves in the bisulphate of potassa. The saline mass is extracted with boiling water, and the process is repeated with fresh bisulphate as long as the salt continues to receive colour ; excess of carbonate of soda is poured into the aqueous extract, the whole is evaporated to dryness, and the residue is calcined ; the residue is again treated with boiling water ; oxide of rhodium remains undissolved, and is then in a state to be reduced by hydrogen gas. 400 QUANTITATIVE ANALYSIS. 181. OSMIUM AND RUTHENIUM. The first of these metals, which is found in combination with iridium in platinum ores, is extracted by the following process (Berzelius's ' Traite de Chimie') : The grains of os- mium-iridium, which are exceedingly hard, are reduced to a very fine powder, which is then digested with hydrochloric acid, to remove the iron which has become rubbed from the mortar during the operation of trituration ; it is then dried, mixed with nitre, and heated gradually to redness in a porce- lain retort connected with a receiver containing caustic am- monia. During this operation both metals become oxidized, and a portion of the oxide of osmium, or osrnic acid, which is volatile, is carried forward with the nitric oxide gas, and condensed in the ammonia, to which it communicates a yellow colour. Another portion is condensed on the sides of the receiver in the form of a crystalline mass. All disengage- ment of gus having ceased, the apparatus is allowed to cool, the receiver containing the ammonia is then removed, and the mass which remains in the retort is dissolved in water. The solution is of a deep-brows colour, and contains a combination of the two oxides with potassa. It must not be filtered, since the organic matter of the paper decomposes it. It is there- fore at once distilled, at a gentle heat, in a retort connected with a receiver, and the greater part of the liquor drawn over. This liquor, which is colourless, and possessed of a strong and disagreeable odour, contains the osmium in the form of osmic acid. The following method of extracting iri- dium and osmium from the pulverulent residue, after digest- ing platinum ore with aqua-regia, was proposed by Wohler. The residue is mixed with an equal weight of dry and finely pulverized common salt, heated to redness in a long glass tube, and a stream of chlorine sent through, as long as it con- tinues to be absorbed. In this operation, the titanate of iron is not attacked, but there are formed compounds of iridium and osmium with chloride of sodium ; much osmic acid is also disengaged, being expelled with the aqueous vapour intro- duced with the chlorine gas ; this is condensed in a receiver containing ammonia, which is adapted to the tube. When no more chlorine is absorbed, the contents of the tube are treated with water, which dissolves the double salts of iridium and osmium. The solution has a deep red-brown colour. The OSMIUM AND RUTHENIUM, 401 titauate of iron and other matters remain undissolved. The solution is then distilled, by which, much osmic acid, arising from the decomposition of the chloride of osmium, is removed, and is condensed in a receiver containing ammonia ; when no more osmic acid passes over, the distillation is stopped, and the residual liquor, after being filtered, is mixed with excess of carbonate of soda, evaporated to dryness, and the residue feebly calcined. It consists of a mixture of suboxide of iridium with chloride of sodium ; the latter is dissolved out by water. The suboxide of iridium is not, however, pure, it still contains a notable quantity of iron and of osmium ; the former is removed by reducing the oxide to the metallic state in a current of hydrogen eas, and then acting on the residue with concen- trated hydrochloric acid. Under this. treatment the platinum residue loses in general from 25 to 30 per cent, of its weight ; it is not, however, exhausted ; and, by a second treatment with chloride of sodium, a further amount of 5 or 6 per cent, of a mixture of osmium, iridium and iron may be extracted. From the residue a small quantity of platinum may generally be dis- solved out by aqua-regia ; it not unfrequently also contains chloride of silver, which may be separated by ammonia. The following modification of this process has been recom- mended by Fritzsche (Journ. fur Prakt. Chem., xxxvii. p. 483) : Equal portions of caustic potassa and chlorate of po- tassa are melted together in a very spacious porcelain cruci- ble over a spirit-lamp, and into the fused mass is conveyed about three times its weight of osmium-iridium, without first reducing it to powder. As soon as, on further heating, the chlorate of potassa liberates oxygen, the fused mass begins to act on the osmium-iridium. The mass froths violently, so that the heat must be moderated when it becomes more tena- cious ; the action at last proceeds without any further heating, the mass becomes, nearly black, and the operation is discon- tinued as soon as the frothing has ceased. During the whole operation not a trace of osmium vapours is perceptible, but a slight evolution commences on the solidification of the mass, M r hich is increased by further heating ; this, however, is unne- cessary, so that there is scarcely any trouble with the vapours of osmic acid. Six hundred grammes of osmium-iridium may be fused with ease in a porcelain crucible capable of hold- ing 2 Ibs., with 100 grammes of caustic potassa and chlorate of potassa, over a spirit-lamp. The operation scarcely lasts an 402 QUANTITATIVE ANALYSTS. hour, and at least 50 grammes are decomposed. On treating the fused mass with water, an orange-coloured solution, con- taining osmium and. ruthenium, is obtained, and a blackish-blue precipitate, which may very easily be separated by suspension from the undecomposed iridium-osmium. By the above process of Berzelius, iridium cannot be com- pletely separated from osmium, for, on heating the former while exposed to the air, vapours of osmic acid are always disengaged. In order to complete the separation of these two metals, the following method is employed by Fremy ('Comptes E-endus,' Jan. 22, 1844) : 100 grammes of the residue of the platinum workings are mixed with 300 grammes of nitre ; the mixture is introduced into a large crucible and kept for an hour at a red-heat in a- wind furnace. After this calcination the mass is poured on a metallic plate, which operation should be performed in the open air, and it is even indispensable to cover the face, for without this precaution the vapours of osmic acid would act violently on the skin. During the cal- cination with nitre, a certain quantity of osmic acid is lost, but it was found that the proportion of this acid which might be condensed would never repay the inconveniences of calcining in a porcelain crucible. The decanted mass, which contains the osmiate and iridiate of potassa, is treated in a retort with nitric acid, which liberates the osmic acid, which is condensed in a concentrated solution of potassa. The residue is treated with water, which removes the nitre, and it is then acted upon with hydrochloric acid, which dissolves the oxide of iridium. In this manner the osmium is obtained in the state of osmiate of potassa, and the iridium in that of soluble chloride. Into the solution of osmiate of potassa Fremy pours a small quan- tity of alcohol ; the liquid becomes heated, acquires a red tint, and deposits a crystalline powder of osmite of potassa ; in this case the osmium is frequently precipitated entirely from solu- tion. The salt may be washed with alcohol, which does not dissolve it, and it then may be preserved without alteration. All the compounds of osmium may be prepared from it. On treating it with a cold solution of sal-ammoniac, it dissolves at first, and is then decomposed, giving rise to a new yellow salt, scarcely soluble in cold water. This salt, which is so easily prepared, affords, on calcination in a current of hydrogen, perfectly pure osmium. To extract the iridium, Fremy treats the chloride obtained OSMIUM AND RUTHENIUM. 403 as above with sal-ammoniac, a reddish-brown precipitate is formed, consisting of a combination of the bichlorides of os- mium and iridium with sal-ammoniac. To separate these two double salts a stream of sulphurous acid is passed into the water in which they are suspended, the double salt of iridium is decomposed, chloride of iridium is formed, which is very soluble in water; the double salt of osmium undergoes no change, it remains in the state of a red salt. The soluble salt of iridium crystallizes in large brown prisms out of solutions in sal-ammoniac ; it is therefore easy to purify it when calcined in a current of hydrogen ; it affords metallic iridium in a state of purity. Besides iridium and osmium, another metal has been found in the platinum residues by Professor Clauss (Bullet, de la Classe Physico-Math. de 1'Acad. de St. Petersbourg, t. iii. p. 311). It is called by its discoverer ruthenium. He pre- pared it in the following manner : The residue, which had been once fused with nitre and extracted with water and acids, was mixed with an equal quantity of nitre, and kept at a white- heat in a Hessian crucible for two hours. The mass was taken out while still red-hot, with an iron spatula, and, after cool- ing, was reduced to a coarse powder, which was extracted with distilled water, leaving it to stand with it till it became clear ; the perfectly clear liquid, which was of a dark yellow colour, was then decanted. It could not be filtered, since it was de- composed by the action of the organic matter of the paper. It contained rutheniate, chr ornate, and silicate of potassa, not a trace of either rhodium or iridium, and only a very minute trace of osmiate of potassa. Nitric acid was cautiously added to this solution, until the alkaline reaction of the liquid had disappeared : oxide of ruthenium and potassa, and some sili- cic acid, were hereby precipitated in the form of a velvet black powder, while chromate of potassa remained undissolved. After edulcoration, the oxide of ruthenium and potassa was dissolved in hydrochloric acid, and the solution was evapo- rated till the silica separated as a gelatinous mass. It was then diluted with water and filtered. It could not be evapo- rated to dryness for the more complete separation of the silica, because the chloride of ruthenium was thereby decom- posed into an insoluble protochloride. The filtered solution, which is of a very beautiful orange-yellow colour, was eva- porated down to a very small volume, and mixed with a concen- 404 QUANTITATIVE ANALYSIS. trated solution of chloride of potassium, when the salt KOI + Ru C1 4 separated in reddish-brown crystals. A further quan- tity was obtained by evaporating the liquid decanted from the crystals. The salt was further purified by recrystallization. Clauss has hitherto only been able to obtain the metal as a blackish-grey powder, which is considerably lighter than iri- dium : the aqueous solution of the chloride is precipitated in the form of a black oxide by ammonia, by which it is distin- guished from all the other platinum metals, none of which are precipitated by ammonia at the ordinary temperature. If a slip of zinc be inserted in the solution of the orange-coloured chlo- ride acidified with hydrochloric acid, a black metallic powder is after a time deposited, the liquid acquires a dark indigo- blue colour, but subsequently, after the whole of the metal is deposited, becomes colourless. GROUP V. Section B. Antimony, Arsenic, Tin, Platinum. Iridium, Gold, Se- lenium, Tellurium, Tungsten, Molybdenum, Tita- nium. 182. ANTIMONY. This metal is quantitatively estimated as sulphide, as anti- inonious acid, and as pure metal. Precipitation as Sulphide. The solution is diluted with water, and sufficient tartaric acid added to redissolve any basic salt, which the addition of water may have precipitated. A stream of washed sulphuretted hydrogen gas is then conducted through the solution until it smells strongly of it ; it is allowed to remain at rest for some time in a moderately warm situation, till the excess of sulphuretted hydrogen has been expelled. If the solution under examination contain antimony in the state of oxide only, the precipitated sulphide may be then collected on a weighed filter, washed thoroughly with distilled water, dried at 212, and weighed. Its composition is One equivalent of Sb . . 120 . . 71'43 Three ditto of S ... 48 . . 28-57 One ditto of SbS 3 . 168 . . lOO'OO ANTIMONY. 405 But if any of the higher oxides of antimony are also present, which will always be the case when the compound under ex- amination is dissolved in aqua-regia, correct results cannot be obtained by simply weighing the precipitated sulphide, its composition being variable. In this case the sulphide must be decomposed by nitric acid, and the amount of sulphur present determined by converting it into sulphuric acid in the following manner : The sulphide is collected on a filter, dried at 212, and its weight, filter included, accurately determined ; a por- tion is then removed for analysis, the weight of this portion is noted ; it is digested in a flask with fuming nitric acid, care- fully added at first, the action generally being very energetic ; hydrochloric acid is then added, and heat applied and continued for some time ; if a clear solution be obtained, it is diluted with water, tartaric acid being added to redissolve any basic salt which may be thereby precipitated. The sulphuric acid formed by the oxidation of the sulphur is precipitated in the form of sulphate of baryta by chloride of barium : it is received on a filter, well washed with boiling distilled water, ignited, and weighed : from the weight obtained, the amount of sulphur is calculated, which, being deducted from the weight of the sul- phide of antimony operated on, gives the quantity of metal ; should it happen that a portion of the sulphur escaped oxida- tion by the nitric acid, the solution will not be clear ; in this case, the solution, having been diluted with water with the requisite addition of tartaric acid, is passed through a weighed filter, and the weight of the separated sulphur, after being well washed and dried on the filter at 212, is added to that calculated from the sulphate of baryta. It may be observed, that it is easy to see whether sulphide of antimony precipi- tated from the original solution by sulphuretted hydrogen is the sulphide corresponding to Sb0 3 : a small weighed portion of it is boiled in a test tube with hydrochloric acid, if a clear solution be obtained, the sulphide is SbS 3 , and the remainder may at once be weighed and calculated, but if it do not com- pletely dissolve, it contains a mixture of one or more of the higher sulphides, and the residue must be treated as above directed. Two other methods have been proposed for treating the mixed sulphides : one is to decompose them at a gentle heat in a current of dry hydrogen gas, by which operation the sulphur is removed partly as sulphur, and partly as sulphuretted hydrogen, metallic antimony being left behind ; the other is to II. 2 E 406 QUANTITATIVE ANALYSIS. heat the mixture in a small retort out of contact of air, by which all the higher sulphides are converted into the lowest ; neither of these methods, however, gives such correct results as the first. When the reduction process is adopted, it is im- possible to regulate the heat so as to prevent the volatilization of a small portion of the antimony, and, in the process of heating the sulphides out of contact of air, a portion of SbS 3 is volatilized, and another portion is reduced to oxide, which is sublimed together with the sulphur. Estimation as Antimonious Acid. The compound is evapo- rated with nitric acid, and the residue ignited as long as it con- tinues to lose weight : its composition is : One equivalent of Sb . . . 120 . . 78 '94 Tour ditto of O 32 21-06 One ditto of Sb0 4 . . 152- . . 100'GO Dilute solutions of antimony in which the amount of metal is to be determined, must not be concentrated by evaporation when they contain hydrochloric acid, as is almost always the case, because terchloride of antimony escapes with the vapour of acid. The volatilization of the metal cannot be prevented by the addition of sulphuric acid, but it may be, pretty nearly, by nitric acid. The estimation of sulphur in precipitated sul- phide of antimony may be made by treating a weighed quantity of it with hydrochloric acid; terchloride of antimony (SbCl 3 ), corresponding in composition to teroxide of antimony (Sb0 3 ), is formed, and the equivalent quantity of sulphur escapes in the form of sulphuretted hydrogen. The remainder of the sulphur separates in the solid form, and after boiling with hydrochloric acid it may be collected on a filter, washed first with water, containing some hydrochloric and tartaric acids, and lastly with pure water ; its weight indicates the amount of the sulphide of antimony (Rose). Assay of Antimony Ores (Levol, Ann. de Chim. et de Phys., 3 ser. xlvi. p. 472). 100 parts of the ore (the sulphide) are mixed with 200 of ferrocyanide of potassium (anhydrous), covered with 50 parts of cyanide of potassium, and heated to a dull red-heat in an iron crucible ; a regulus of metallic anti- mony is obtained, which, according to the author, represents nearly exactly the quantity contained in the ore. Ores of lead may be assayed in a similar manner. ARSENIC. 407 183. AESENIC. This metal may be quantitatively estimated as arsenic acid by combining it with oxide of lead, and as ammonio-ar senate of magnesia ; it may also be conveniently weighed as sulphide. Quantitative estimation by Protoxide of Lead. To ensure the metal being in the state of arsenic acid, the compound under examination (which must contain no other acid) is di- gested with aqua-regia, and carefully evaporated to dryness ; the residue, after being somewhat more strongly heated in a platinum crucible, is dissolved in water, and a known weight of recently ignited and pure protoxide of lead added ; the mix- ture is carefully evaporated to dryness, and gently ignited : the amount of arsenic acid present, is learnt from the increase in weight, and from this the quantity of arsenic is calculated. Estimation as Sulphide. For this purpose the arsenic must be in the form of arsenious acid, to ensure which, the solution of the compound should be mixed with a concentrated aqueous solution of sulphurous acid, and gradually heated to gentle ebullition in a flask ; the heat must be maintained till the whole of the excess of sulphurous acid is expelled ; hydrochloric acid is now T added, and a stream of washed sulphuretted hydrogen is conducted through the liquid, till it smells strongly of it. It must not be immediately filtered, because the sulphuretted hydrogen liquor may hold in solution a portion of sulphide of arsenic. It is set aside in a warm place, until the excess of sulphuretted hydrogen gas is expelled. Fresenius recommends that this be done by transmitting through the solution a stream of washed carbonic acid gas. The liquid being freed from sul- phuretted hydrogen, is passed through a weighed filter, washed with hot distilled water, dried at 212, and weighed. Its com- position is One equivalent of As . . 75 .. 60*97 Three ditto of S . . . _48 . . 39-Q3 One ditto of AsS 8 . . 123 . . lOO'OO Should the process of treating the solution with sulphurous acid not have been adopted, and if there is reason for supposing that a portion of the arsenic may have been in a higher state of oxidation than that of arsenious acid, and if, moreover, there should be present certain other substances capable of decom- posing sulphuretted hydrogen, such as chromic acid, peroxide 2E 2 408 QUANTITATIVE ANALYSIS. of iron, etc., then the sulphide of arsenic cannot be weighed for the purpose of estimating the metal, until after the excess of sulphur has been removed : this is done precisely in the same manner as has been directed in the case of sulphide of antimony with excess of sulphur, viz. by converting the free sulphur into sulphuric acid, and determining its amount in the form of sulphate of baryta. A solution of pentasulphide of arsenic (AsS 5 ) in sulphide of ammonium yields instantly a precipitate of ammonio-arsenate of magnesia with a solution of magnesia. According to Lessen (Ann. der Chem. und Ph.), sulphides of tin and antimony are not precipitated under the same conditions. Estimation as Ar senate of Magnesia and Ammonia. For this purpose the arsenic must be in the form of arsenic acid, to which, if in a lower state of oxidation, it must be brought by warming the solution with hydrochloric acid, and then adding chlorate of potassa in small quantities at a time, till the fluid smells strongly of chlorous acid ; it is then allowed to cool, ammonia in excess added, and then a mixture of chloride of ammonium and sulphate of magnesia ; the solution is allowed twelve hours thoroughly to precipitate, and is then filtered through a weighed filter, the precipitate washed witk ammoniacal water, dried at 212 F., and weighed. Its compo- sition is Two equivalents of MgO ..... 40'32 . . 21' 18 One ditto of NH 4 O ..... 26-00 . . 13-66 One ditto of AsO 5 ..... 115-00 , . 60-43 One ditto of HO ...... 9'00 . . 4'73 One ditto of (2MgO,NH 4 O,As0 5 ,HO) 190'32 100-00 As the mixture of chloride of ammonium, sulphate of magnesia, and ammonia, is of constant use in the laboratory for the deter- mination of phosphoric acid, it should be kept ready prepared by dissolving 1 part of crystallized sulphate of magnesia and 1 part of chloride of ammonium in 8 parts of water and 4 parts of solution of ammonia, allowing the fluid to stand at rest fop several days, and then filtering (Fresenius). According to Kose (PoggendorfF's ' Annalen der Physik und Chemie,' vol. cxvi. p. 453) the arsenates of iron, manganese, zinc, lead, and copper may be successfully analysed by calcina- tion with sulphur in a current of hydrogen, the arsenic volati- lizing as sulphide, the base remaining under the form of fixed ARSENIC. 409 sulphide. In many instances arsenic acid may be expelled from arsenates by calcination with chloride of ammonium ; the alka- line arsenates are transformed into chlorides by a single calci- nation, the earthy arsenates offer more resistance, and in the case of arsenate of magnesia the arsenic acid is never com- pletely expelled. The analysis of native arsenate of iron may be perfectly effected by sulphuretted hydrogen: a weighed quantity of the ore is heated to faint redness in a current of the gas ; both metals are converted into sulphides ; the sulphide of arsenic is completely volatilized, and the sulphide of iron, which does not contain a trace of arsenic, is dissolved in hydro- chloric acid, peroxidized by chlorate of potassa, and precipitated by ammonia. Separation of Arsenious from Arsenic Acid (Fresenius). ' The solution containing the two acids is divided into two equal portions ; in one, the arsenic acid is reduced to arsenious acid by sulphurous acid, and the whole then precipitated by sulphu- retted hydrogen : the precipitated sulphide of arsenic is washed, dried at 212, and weighed. The arsenious acid in the other half of the solution is converted into arsenic acid by chlorine, for which purpose it is mixed with hydrochloric acid, and solu- tion of indigo added till the fluid acquires a blue colour. Chloride of lime, containing a known amount of chlorine, is then added from a weighed solution, until the blue colour dis- appears : from the quantity consumed, the amount of chlorine used is calculated. This operation depends on the circum- stance, that every equivalent of arsenious acid requires two equivalents of chlorine to convert it, in the presence of water, into arsenic acid : thus By calculating the resulting quantity of arsenious acid upon sulphide of arsenic, and subtracting the weight of this from the total weight obtained from the first portion of the solution, the amount of arsenic acid originally contained in the latter is found. A simpler method is to precipitate the arsenic acid as arsenate of magnesia and ammonia, and to determine the amount of arsenious acid in the filtrate by sulphuretted hy- drogen. Separation of Arsenic from Antimony. When the compound of the two metals is in the reguline state, the arsenic may be expelled by heating out of contact of air; when, however, other metals are also present, this process cannot be adopted, since 410 QUANTITATIVE ANALYSIS. most other metals retain a part of the arsenic when at a red- heat. In this case several methods have been proposed : (1.) The substance is dissolved in aqua-regia, tartaric acid is added, the solution is dissolved, and the antimony and arsenic are together precipitated by sulphuretted hydrogen : the sul- phides are intimately stirred together, collected on a filter, dried at a very gentle heat, and weighed. From a weighed portion, the amount of sulphur is determined by dissolving it in aqua-regia, and having added tartaric acid and diluted with water, precipitating the sulphuric acid formed by chloride of barium. Another weighed portion of the mixed sulphides is heated in a tube, through which a stream of hydrogen gas is passing; the excess of sulphur and the sulphide of arsenic are expelled, leaving the sulphide of antimony, which may then be weighed. This method, when carefully performed, is capable, according to Rose, of affording a result only about a half per cent, from truth ; the proportion of arsenic present is calcu- lated from the loss, the united weight of both metals having by the first experiment been determined in the state of sulphide. (2.) Behren's method. The arsenic and antimony are con- verted into sulphides, and to the mixture is added, while still moist, an equal volume of neutral nitrate of lead, and about ns much water. The mass is boiled in a porcelain dish, being stirred without interruption, and, the water which evaporates being renewed until the whole has acquired a dark-brown colour, it is then filtered. The residue contains the entire amount of antimony, and a portion of the arsenic. The solution contains the arsenic in the state of arsenious acid, nitric acid, and oxide of lead ; it is heated with carbonate of ammonia, as long as a precipitate is formed. To the liquid filtered from the carbonate of lead, some hydrochloric acid is added, until it has an acid reaction ; sulphuretted hydrogen gas is then trans- mitted through the solution, and the sulphide of arsenic thus obtained has no trace of antimony. To separate the arsenic contained in the state of sulpharsenide of lead, in the mass which remained on the first nitration, M. Behren digests it at a gentle heat with caustic ammonia, which converts the sulph- arsenide of lead into sulphide of lead, and sulphide of arsenic, which latter dissolves in ammonia. To the filtered solution, hydrochloric acid is added, and the sulphide of arsenic precipi- tated, is added to that first obtained. Presenius objects to the precipitation of the lead in the so- ARSENIC. 411 lution containing the arsenic by carbonate of ammonia ; he pre- fers separating the two metals by sulphide of ammonium, or by evaporating the solution to dryness, and fusing the residue with carbonate of soda and nitre. (3.) Antimony and arsenic in alloys may, according to Fre- senius, be separated, though not in a very accurate manner, by heating the mixture covered with common salt and carbonate of soda, in a glass tube, through which a current of dry carbonic acid gas is passing ; the tube is heated gently at first, but the heat is gradually increased to the highest degree of intensity: by this means the arsenic is driven off, while the volatilization of the antimony is prevented by the common salt and carbonate of soda. (4.) Hofmann's method (Quart. Journ. Chem. Soc.). This is based upon the dissimilar deportment of arseniuretted and antimoniuretted hydrogen with nitrate of silver, the former yielding arsenious acid which passes into solution the latter giving rise to the formation of antimoniuret of silver, which is insoluble in water This process presents no difficulty as far as the arsenic is con- cerned, which may be recognized in solution by ammonia if there be an excess of silver, or by sulphuretted hydrogen if the silver has been entirely precipitated. It is far less easy to find, according to this process, minute quantities of antimony in the presence of large quantities of arsenic, the silver compound of antimony being mixed with a bulky precipitate of metallic silver. By treating this precipitate with hydrochloric acid, there dis- solves, together with the antimony, a small quantity of chloride of silver, which is sufficient to darken the precipitate produced in the solution by sulphuretted hydrogen to such a degree as altogether to mask the presence of antimony. This inconveni- ence may easily be obviated by boiling the mixture of silver and antimoniuret of silver, after the arsenious acid has been care- fully washed out with boiling water, with tartaric acid, which dissolves the antimony alone. The solution thus obtained yields at once with sulphuretted hydrogen the characteristic orange- yellow precipitate of sulphide of antimony. In evolving the hy- drogen compounds of arsenic and antimony, care must be taken to add as little nitric acid as possible to the hydrochloric acid used in dissolving the sulphides of the metals, since the presence 412 QUANTITATIVE ANALYSIS. even of moderate quantities of this acid, greatly interferes with the free disengagement of the gases. If there be tin with the arsenic and antimony, the metal will be deposited upon the plates of zinc used in evolving the hydrogen, from which it may be mechanically detached, dissolved in hydrochloric acid, and tested by the usual process. (5.) Meyer's method (Liebig's 'Annalen,' May, 1848). This is founded on the insolubility of antimoniate of soda. The mixture of the two metals is deflagrated with three times its weight of nitrate and carbonate of soda, the fused mass is washed with cold water, which dissolves out the arsenic acid, which, after reduction by sulphurous acid, may be precipi- tated by sulphuretted hydrogen, and determined in the usual manner. 184. TIN. This metal is weighed as peroxide : it has also been attempted to estimate it by a normal solution of iodine. Quantitative estimation as Peroxide. The compound under examination is evaporated nearly to dryness, at a boiling heat, with strong nitric acid, to ensure the whole of the metal being in the state of peroxide ; if any hydrochloric acid be present, it must be completely decomposed and expelled. The peroxide of tin, which by the action of the nitric acid has become con- verted into the insoluble modification, is filtered, washed,ignited, and weighed. Its composition is One equivalent of Sn . . . . 59 . . 78 '66 Two ditto of O . . . . 16 . . 21'34 One ditto of Sn0 2 . . . 75 . . lOO'OO Tin may also be precipitated from its acid solution, whether in the state of a protosalt, or of a persalt, by sulphuretted hy- drogen ; in the former case the resulting sulphide is brown, in the latter case it is yellow ; the gas must be passed through the solution till it smells strongly of it, and a gentle heat must afterwards be applied, by which a small quantity of sulphide, which is dissolved in the saturated liquor, becomes again depo- sited ; the sulphide of tin must not be weighed as such, but converted into peroxide, which may be done by heating it, first gently and then intensely, in a porcelain crucible, till all the sulphur is removed in the form of sulphurous acid, the ex- pulsion of which is facilitated by adding a small fragment of TIN. 413 carbonate of ammonia. Rose recommends to convert the sul- phide of tin into peroxide by nitric acid, observing that, when the quantity of sulphide is at all considerable, its conversion into peroxide takes place by heat alone with extreme slowness. Volumetric estimation : (1.) Oaultier de Claiibry's method (' Comptes Rendus,' July 13, 1846). As a normal solution, he employs 1 gramme of iodine dissolved in a decilitre of alcohol, sp. gr. 0-932, and the solution of tin is prepared with 1 gramme of this metal dis- solved in hydrochloric acid, and diluted with water freed from air, so as to form a litre. By means of a graduated pipette, a demi- decilitre of the tin solution is measured off, and the bu- rette, divided into tenths of a cubic centimetre, is filled with the normal solution ; the latter is poured into the first, until it is no longer decolorized ; half a decilitre of the tin solution, containing 5 decigrammes of tin, decolorizes 100, or 10 cubic centimetres of the normal solution. If the tin ore under ex- amination is soluble in hydrochloric acid, the operation is perfectly simple ; if it does not dissolve in it, it is acted 011 with aqua-regia, containing much hydrochloric acid ; and, when the tin is become converted into perchloride, an excess of hydrochloric acid is added, and it is then boiled with some iron, by which it becomes reduced to protochloride ; the pro- cess is now conducted in the same manner as in the prece- ding case. If it happen to be an alloy containing only twenty per cent, of lead, hydrochloric acid will dissolve it ; beyond that, however, only imperfectly ; but, as aqua-regia scarcely acts on the compounds of lead, the alloy must be dissolved in nitric acid, evaporated to expel the excess of acid, and then treated with hydrochloric acid and iron. Stannic acid, espe- cially when it has not been dissolved, is readily converted into protochloride, in the presence of an excess of hydrochloric acid and protochloride of iron, so that the assay is brought to the same state as when the product could be acted upon imme- diately with hydrochloric acid. When the compound to be analysed contains arsenic, antimony, bismuth, copper, or lead, the iron precipitates it, and again reduces the assay to the state of a tin solution. To precipitate the whole of the copper, and not to leave any of the protochloride of that metal in solution, a considerable excess of hydrochloric acid must be employed, and the boiling with iron continued for some length of time. The analysis of a salt of tin can be made with the same ease, and 414 QUANTITATIVE ANALYSIS. if a mixture of a per- and proto-salt is examined, or any of the corresponding haloid compounds, the relative proportions can be determined by analysing the substance itself, and then making a second analysis of the product boiled with hydro- chloric acid and iron. Zinc and iron do not interfere with the analysis by iodine, while the protosalts of iron and the corre- sponding haloid compounds decolorize the sulphate of indigo, which M. Pelouze had attempted to employ for the estimation of tin, and renders this process impracticable. Iodine may, according to the author, be used to determine the quantity of tin in a solution containing various metals; but if there be present an arsenite, sulphite or hyposulphite, phos- phite or hypophosphite, the normal solution will be decolorized as with protochloride of tin. It will be requisite, therefore, to oxidize these salts by nitric acid or by chlorine, and to reduce the tin to protochloride by means of iron. (2.) M. Stromeyer's method. He introduces the solution of protochloride of tin into an excess of sesquichloride of iron. The salt of iron becomes reduced to a minimum according to the following equation : 811 Cl + Fe 2 C1 3 = Sn C1 8 -f 2 Fe Cl. It is then estimated by a standard solution of permanganate of potassa, as if it were a salt of protoxide of iron. The results are accurate, but the method is only applicable in the absence of copper and iron, as these two metals decompose per- manganate of potassa as well as of tin. Separation of Protoxide of Tin from Peroxide of Tin. Eose's method is the following : The solution is divided into two por- tions ; in one, the amount of the metal is determined by preci- pitating it by sulphuretted hydrogen, and then converting it into peroxide in the manner described above ; the other por- tion is dropped into solution of chloride of mercury, and the subchloride of mercury precipitated, is filtered, washed, dried at a gentle heat, and weighed. From this weight, it is easy to calculate how much protoxide or protochloride of tin was con- tained in the solution, since the quantity of chlorine contained in the subchloride of mercury is the same as that contained in the protochloride of tin, by which it w r as precipitated ; or, on the other hand, the quantity of chlorine in the subchloride of mercury is equivalent to the quantity of oxygen in the pro- toxide of tin, which acted as a precipitant. Analysis of Alloys of Tin and Copper. The process of M, TIN. 415 Cottereau (' Coraptes Kendus,' June 29, 1846) is founded on the principle that copper is precipitated from its solutions by zinc before tin. The alloy is reduced to a fine powder, a certain quantity weighed off, and digested with boiling hydrochloric acid. Into the resulting solution of the protochlorides of cop- per and tin, a plate of zinc is introduced. By a previous assay of the copper contained in the alloy by the cuprometric process of M. Pelouze (p. 379), the quantity of copper is determined, and consequently an equivalent amount of zinc may be added to the solution ; or the plate of zinc may be immediately intro- duced into the solution, and left there until a bright iron blade does not acquire a red tint on being immersed in the liquor. The strip of zinc is then removed, and the precipitate collected on afilter. Whichever plan be adopted, the filtered liquor is acted on just as if it were pure protochloride of tin. The proto- chloride of zinc formed does not in the least interfere with the reaction. Separation of Tin from Antimony. This is attended with difiiculty. The following methods have been proposed : (1.) JSerthier's method. The two metals being dissolved in concentrated hydrochloric acid, tartaric acid is added to the solution, which is then diluted with water, sulphite of ammonia added, and the whole boiled. The tin is precipitated while the antimony remains dissolved. (2.) Level's process (Ann. de Chim. et Phys., Jan. 1845). The alloy being reduced to a thin plate, is heated with hydro- chloric acid; after some minutes' boiling, a saturated aqueous solution of chlorate of potassa is added in small quantities at a time, until the alloy entirely disappears. The two metals are then thrown down together by means of a bar of distilled zinc ; the precipitate is removed from the zinc with the greatest care, a quantity of concentrated hydrochloric acid, about equal to what was employed at first, is added, and the whole is boiled so as to redissolve the tin without previously removing the chlo- ride of zinc; when the action has terminated, that is, when nothing remains but the antimony, which is always the case after an hour's boiling, this metal then forms a very fine blackish powder : it is collected on a weighed filter, and the tin may now be immediately obtained from the liquors by sulphuretted hydrogen. According to L. Eisner (Journ. fiir Prakt. Chem., vol. xxxv. p. 313), the above process has no claim to accuracy. 416 QUANTITATIVE ANALYSIS. He found, ou repeating the experiment with a mixture of 7i grains of tin and as much antimony, exactly according to the directions given by Levol, that the acid liquid filtered from the black powder, on being treated with sulphuretted hydrogen, yielded a precipitatewhich consisted evidently of two differently- coloured layers ; the inferior layer was clearly the orange -red precipitate of sulphide of antimony, above which was the cho- colate-brown protosulphide of tin. It is certain, therefore, that some antimony had dissolved with the tin. Eisner convinced himself that antimony is partly dissolved on being boiled with hydrochloric acid, and that consequently no quantitative me- thod of separating it from tin can be founded on its insolubi- lity in that acid. (3.) Ease's method (Chem. Graz., vol. v. p. 313). Strong nitric acid is cautiously poured upon the metals, and when the vio- lent oxidation has ceased, the whole is evaporated at a gentle heat, and the dry powder of the oxides fused in a silver cruci- ble over an argand lamp, with an excess of hydrate of soda. The fused mass is softened with a large quantity of water, gently warmed, and the antimoniate of soda allowed to subside. When perfectly cold, the clear solution is passed through a filter : if this is done while it is still warm, the solution will contain some antimoniate of soda. The insoluble salt is again treated once or twice with water, allowed to settle and cool, and the liquid, when perfectly clear, passed through the filter. "When the whole of the stannate of soda has been dissolved in this manner, the liquid which has been warmed with the antimoniate of soda remains opalescent ; it must not be poured on the filter, as it would pass through turbid. A small quantity of a dilute solution of carbonate of soda may be added to it, which ren- ders it clear ; but the edulcoration must not be continued for any length of time, as otherwise some antimoniate would be dissolved. The moist antimoniate of soda is now treated in a beaker with a mixture of hydrochloric and tartaric acids, in which it readily dissolves ; and the filter, upon which mere traces of the salt should have collected, is washed with the same mixture. The antimony is then precipitated from the solution by sulphuretted hydrogen, and the amount of antimony estimated from the quantity of sulphide obtained. Rose reduces the sulphide of antimony by hydrogen in a porcelain crucible, through the lid of which a thin porcelain tube passes. A gentle heat is care- TIN. 417 fully applied, until the crucible no longer decreases in weight. After the reduction, the inner side of the lid is coated with metallic antimony, which, however, in no way interferes with the accuracy of the experiment. The solution of stannate of soda is acidified with hydro- chloric acid. It is not necessary to add so much acid that the whole of the eliminated oxide of tin is again redissolved. It is merely necessary that the solution should strongly redden blue litmus-paper. Upon this, sulphuretted hydrogen is passed into it. The sulphide of tin is converted by roasting into oxide. When it has been dried, it frequently decrepitates, by which, if care be not taken, a very considerable loss may be occasioned. It is, on this account, preferable to place it with the filter, while still moist, in a porcelain crucible, and to heat it for a long time very gently, and with access of air, in order to expel the sulphur at the lowest possible temperature. If a strong heat be given at the commencement, white fumes of oxide escape, especially when the air has free access. The higher sulphide of tin has the property of subliming somewhat, at certain temperatures ; the vapours are oxidized by contact with the air, and form oxide of tin. This is also the cause of a white sediment of oxide of tin being formed upon the charcoal, when sulphide of tin is heated on charcoal before the blow- pipe. A strong heat should not be applied until there is no longer any perceptible odour of sulphurous acid. After being strongly ignited, a piece of carbonate of ammonia is placed in the crucible, and, after its volatilization, a strong heat applied, with access of air, in order to expel the whole of the sulphuric acid formed ; a small decrease of weight will be perceived. The oxide of tin thus obtained never appears perfectly white ; and the sulphide of tin, precipitated by sulphuretted hydro- gen, does not possess a purely yellow colour, which Kose attri- butes to melting the metallic oxides with hydrate of soda in the silver crucible, traces of oxide of silver being removed by the alkaline solution. The results do not attain the highest degree of accuracy. Stannate of soda containing an excess of hydrate of soda does not become turbid by boiling, as stated by Fremy, and the solution may even be evaporated until crystals separate, which, on the addition of water, entirely dis- solve, yet the antimoniate of soda contains a small quantity of oxide of tin ; consequently the sulphide precipitated by sul- phuretted hydrogen from the acid solution of the antimoniate 418 QUANTITATIVE ANALYSIS. of soda contains a small amount of sulphide of tin, which does not part with the whole of its sulphur at the temperature at which the sulphide of antimony is reduced by hydrogen. For this reason a somewhat smaller amount of tin, a larger quantity of antimony, and a slight excess, is found in the analysis. (4.) Tookey's method (Quart. Journ. Chem Soc., vol. xv. p. 464). The author finds the process of reduction by metallic iron to answer perfectly. The hydrochloric solution of the two metals is digested at a gentle heat, with pure thin sheet iron, until the whole of the iron is dissolved ; a considerable quan^ tity of cold water is then added, the antimony is collected on a weighed filter, washed and dried ; it should then be purified from traces of iron by digesting with cold dilute hydrochloric acid, dried in the water-oven, and weighed in the form of a grey powder. The tin is precipitated from the filtrate by sul- phuretted hydrogen, the protosulphide of tin dried, and con- verted into binoxide by careful ignition. Separation of Tin, Antimony, and Lead (Tookey). When the ordinary method of separation was used, viz. oxidation by nitric acid, evaporation at a steam temperature, and extrac- tion with water, the author found that it was impossible tho- roughly to remove the lead; varying portions in different trials being retained in an insoluble state by the oxides of tin and antimony, probably in the form of antimoniate of lead. He therefore availed himself of the volatility of the chlorides of tin and antimony, and the fixity of chloride of lead to effect a separation ; for this purpose about 10 grains of the alloy are introduced into a bulb expanded in the middle of a piece of hard combustion tube, having ingress and egress tubes bent at angles of 45 degrees, but in opposite directions ; one point- ing upwards, for introducing the alloy, and the other down- wards, for collecting the volatile products in water : sufficient nitric acid having been poured on the alloy, the funnel tube is closed with a cork, and the oxidation effected at a moderate temperature. When it is converted into a perfectly white mass, the excess of acid is expelled by causing a gentle current of air to traverse the apparatus, a moderate heat being at the same time applied. A current of dry hydrochloric acid gas generated from fused chloride of sodium and sulphuric acid, is then transmitted slowly through the tube, the downward egress-end being immersed in a small quantity of water. When the mass becomes liquid from the absorption of gas, a gentle TIN. 419 heat is applied, the chlorides of tin and antimony distil over, and are condensed in the water, chloride of lead remaining. When the distillation is nearly finished, a greater heat is ap- plied to expel the last traces of bichloride of tin, and lastly the chloride of lead, now perfectly free from tin and antimony, is heated to fusion, it is then removed from the tube and con- verted into sulphate, in which state it is weighed; the water containing the mixed chlorides of tin and antimony, after the addition of a little more hydrochloric acid, is digested with pure metallic iron until the whole of the iron is dissolved ; cold water is then added : the precipitated antimony is then col- lected on a weighed filter, washed with cold water, and dried at 212. The filtrate from the antimony is saturated with sulphuretted hydrogen; sulphide of tin is precipitated, which is converted into binoxide by ignition. Separation of Tin and Arsenic; Analysis of Ar senates of Tin. The mixed oxides, in a finely pulverized state, are placed in a small porcelain tray, and heated in a combustion-tube in a current of dry sulphuretted hydrogen ; both oxides are re- duced to sulphides, the sulphide of arsenic volatilizes com- pletely, leaving sulphide of tin, which is first roasted, and then strongly calcined in a platinum crucible, it is hereby con- verted into stannic acid, in which state it is weighed. Separation of Antimony, Tin, and Arsenic (PoggendorfTs ' Annalen,' Ixxvii. p. 110). The mixture of stannate and aiiti- moniate of soda, obtained by acting on the alloy with nitric acid, and subsequently fusing with hydrate of soda, is first washed thoroughly out of the crucible with water, and the in- soluble antimoniate of soda allowed to separate. The solution which contains all the stannate of soda is now mixed with so much alcohol (specific gravity, O83), that its volume relatively to that of the water is as 1 to 3. In this spirituous solution the stannate of soda is completely soluble, while the last por- tions of antimoniate of soda are completely separated. The clear liquid having been filtered off, the antimoniate of soda is washed with stronger alcohol till some of the filtered liquid, rendered acid with a little dilute sulphuric acid, gives no longer a precipitate with sulphuretted hydrogen. The solu- tion is heated to expel the alcohol, diluted with water, super- saturated with sulphuric acid, precipitated with sulphuretted hydrogen, and the sulphide of tin converted into oxide. The antimoniate of soda is dissolved in a mixture of hydrochloric 420 QUANTITATIVE ANALYSIS. and tartaric acids, from which solution the antimony is precipi- tated by sulphuretted hydrogen. "When arsenic is present in the alloy it is found in the solution containing the staunate of soda ; this solution is supersaturated at once with hydro- chloric acid, without previously driving off the alcohol, and sulphuretted hydrogen passed through it without previously separating the precipitated arsenate of tin. The whole is set aside till it scarcely smells of sulphuretted hydrogen, and the precipitate is collected on a weighed filter. The filtered liquid is heated for some time, by which any portion of sulphuretted hydrogen still present, and the greater portion of the alcohol, are expelled. It is then mixed with a solution of sulphurous acid, and again treated with sulphuretted hydrogen, when generally a further small quantity of sulphide of arsenic is precipitated. This treatment with sulphurous acid is not, however, necessary, if sulphuretted hydrogen is again passed immediately through the liquid separated from the sulphides. The small amount of sulphide of arsenic which is precipitated in this case is always free from every trace of sulphide of tin, and is kept separate, and only added to the sulphide of arsenic obtained in the separation of the arsenic and tin. These two metals contained in the precipitate by sulphuretted hydrogen are separated by being heated in a current of sulphuretted hydrogen, when the sulphide of arsenic is volatilized. The re- sidual sulphide of tin is converted into oxide, and the sublimed sulphide of arsenic into arsenic acid. This method proved on examination to be absolutely accurate. 185. PLATINUM. In whatever form this metal may be precipitated, it is in- variably weighed in the reguline state. It may be precipitated from its solution as ammonia-chloride of platinum, as potassio- chloride of platinum, and as sulphide of platinum. (1.) Precipitation as Ammonio-chloride of Platinum. The acid solution is concentrated ; it is then mixed with a concen- trated solution of chloride of ammonium, and a sufficient quan- tity of spirits of wine added to effect the complete precipita- tion of the double salt : the precipitate is allowed to subside perfectly ; it is then collected on a filter, and washed with spirits of wine till the fluid passes through quite colourless ; it is dried, and ignited, by which it is completely decomposed, metallic platinum, in the form of a grey spongy powder, remain- PLATINUM. 421 ing in the crucible. The ignition requires to be performed with great care : the dried double salt is transferred, filter and all, into a weighed platinum crucible, and is heated gently, the cover of the crucible being laid loosely on, as long as fumes of sal-ammoniac are seen to escape ; the cover is then removed, and a stronger heat applied till the organic matter of the filter is consumed ; the crucible is finally exposed to an intense heat. The reason why a gentle heat must be applied at first, is because a portion of the un decomposed double salt might otherwise be carried off with the fumes of the sal-ammoniac. (2.) Precipitation as Potassio-cliloride of Platinum. The so- lution is mixed with chloride of potassium in excess, and alcohol added ; the precipitated double salt is collected on a filter, dried at 212, and weighed. It would not be safe, however, to esti- mate the platinum from the weight of the double salt, and, in- deed, after ignition, it is not completely decomposed into me- tallic platinum and chloride of potassium ; where accuracy is required, therefore, a known portion of the double salt dried at 212 must be introduced into a weighed bulbed tube, and heated to redness, while a stream of dry hydrogen gas passes through it. It thus becomes completely reduced, hydrochloric gas being evolved ; when this has ceased, which is known by white fumes ceasing to be formed on bringing a glass rod or a feather dipped in ammonia-water near the end of the tube, the apparatus is allowed to cool, the chloride of potassium is then dissolved out with water, and the reduced platinum well washed. It is then again heated to low redness in a stream of hydrogen gas, and weighed; the weight obtained calculated upon the whole precipitate on the filter, gives the result. (3.) Precipitation as Sulphide. This is sometimes practised to separate platinum from such metals as are not precipitated from their acid solutions by sulphuretted hydrogen, as well as from others whose sulphides are insoluble in alkaline sulphides ; in the former case, the gas is passed into the acid solution. No precipitation frequently takes place till heat is applied, when the solution turns brown, and sulphide of platinum is precipitated ; it is not weighed as such, but after being washed, is ignited, by which it becomes reduced, and is then estimated in its metallic state : in the latter case, the solution is ren- dered neither neutral or alkaline, and sulphide of ammonium added in excess, in which the sulphide of platinum dissolves. Separation of Platinum from Antimony, Arsenic , and Tin. PAET II. 2 F 422 QUANTITATIVE ANALYSIS. This is readily effected by taking advantage of the volatility of the chlorides of the three last metals; all four are first preci- pitated together by sulphuretted hydrogen ; the mixed sul- phides are introduced into a bulbed tube, and heated in a stream of chlorine gas ; platinum alone remains. Analysis of Platinum ores (MM. Deville and H. Debray, 'Annales de Chimie et de Physique'). The following sub- stances are contained in the ores of this valuable metal : 1, Sand; 2, Osmide of Iridium ; 3, Platinum, Iridium, Rho- dium and Palladium f probably an alloy) ; 4, Copper and Iron ; 5, Gold, and a little Silver. (1.) Determination of the Sand. This is effected by heating a known weight (about 2 grammes) of the ore with five or six times its weight of pure granulated silver in a crucible, the sides of which have been previously glazed by melting borax in it ; over the metals, 10 grammes of fused borax, and one or two pieces of wood-charcoal are placed, the heat is raised a little above the melting-point of silver, at which it is kept for some time ; the vitreous matters are dissolved by the borax, and all the metals are contained in the button of silver which is found at the bottom of the crucible when cold ; it is heated to faint redness and weighed, the weight subtracted from the sum of the weights of the ore and silver employed, gives the amount of sand. (2.) Determination of Osmium and Iridium. About 2 grammes of the ore are repeatedly treated with aqua-regia until the whole of the platinum is dissolved, which is known by the solution being no longer coloured. The insoluble residue is osmide of iridium in the form of spangles, together with the sand ; it is washed by decantatiori, dried and weighed ; by subtracting the weight of the sand obtained in the former operation, we get the weight of the osmide of iridium. (3.) Determination of the Platinum and Iridium. The so- lution obtained in the last operation is evaporated to dryness on the water-bath, redissolved in a little water, alcohol added, and then crystals of sal-ammoniac in great excess ; after stand- ing for twenty-four hours the orange-yellow or reddish-brown precipitate is thrown on filter, and washed with alcohol, it consists of the ammonio-chlorides of platinum and iridium, but a portion still remains in the solution; the double chlorides of the two metals, after being dried, are heated to low redness in a platinum crucible, and the filter is burnt at the lowest PLATINUM. 423 possible temperature ; to favour the reduction of the iriclium, and the escape of the last traces of osmium, a piece of paper saturated with turpentine is introduced into the crucible. The heat is now raised gradually to whiteness, and the reduction of the metal is finished in a current of hydrogen. The filtrate from the double chlorides of platinum and iridium is concen- trated until the greater part of the sal-ammoniac crystallizes out ; it is then allowed to cool, upon which a deep violet-co- loured precipitate is formed ; this is ammonio-chloride of iri- dium, mixed with a little of the corresponding platinum salt; it is collected on a filter, and washed first with a solution of sal-ammoniac, and then with alcohol, after which it is dried and ignited in a current of hydrogen. The results of the two reduc- tions are weighed and digested at about 120 F., with repeated portions of aqua-regia diluted with five or six times its bulk of water, till the liquid is no longer coloured : all the platinum is dissolved, leaving as a residue iridium in a state of purity. (4.) Determination of the Palladium, Iron, and Copper. These metals are contained in the liquor from which the plati- num and iridium have been separated. The alcohol and chloride of ammonium are evaporated off, the removal of the latter being facilitated by the addition of nitric acid, which transforms it into nitrogen and hydrochloric acid ; the evaporation is con- tinued almost to dryness; the residue is transferred to a weighed porcelain crucible, and when dry, is moistened with sulphide of ammonium, and loosely covered with two or three grammes of pure sulphur. The crucible is covered, placed in a larger clay crucible, and surrounded with pieces of wood charcoal, a cover is then put upon the clay crucible, and it is brought slowly up to a bright-red heat, the fire being lighted at the top of the furnace in order to avoid the projection of any matter from the crucible, which might happen if it were too quickly heated. When cold, the inner crucible contains palla- dium in the metallic state, with the sulphides of iron and copper, and also the gold and the rliodium. The mixture is digested for a long time at 150 with concentrated nitric acid, which dis- solves the palladium, iron, and copper, converting them into ni- trates. These are dissolved out by water, and the residue well washed ; the solution and washings are evaporated to dryness and then calcined at a strong red-heat ; the palladium is reduced, and the iron and copper are converted into oxides, which are easily separated from the palladium by strong hydrochloric acid, 2 F 2 424 QUANTITATIVE ANALYSIS. aiid the latter, now in a state of purity, is strongly ignited and weighed. The chlorides of copper and iron are treated in the usual way for the determination of each metal. (5.) Determination of the Gold. This metal is contained in the residue (4.) insoluble in nitric acid : it is weighed and then treated with dilute aqua-regia, which dissolves the gold, and sometimes, but rarely, a trace of platinum, the presence of which is detected by chloride of ammonium ; the difference be- tween the weight of the crucible before, and after, the treatment by aqua-regia gives the weight of the gold. (0.) The residue left in the crucible is rhodium, which is reduced in a current of hydrogen and then weighed. The authors caution experimenters to beware of the ill effects of osmium, which particularly attacks the eyes ; and also to avoid breathing the vapour from the aqua-regia. 186. GOLD. This metal, like the preceding, is always weighed in the me- tallic state, in which it is precipitated from its solutions by protosulphate of iron, or by oxalic acid. A method of estimating the metal indirectly by means of standard solutions has also been published by O. Henry. Gold may also be precipitated from its acid solution, by sulphuretted hydrogen. The sulphide is readily reduced by heat alone. Precipitation ly Protosulphate of Iron. If the solution con- tain nitric acid, it must be evaporated nearly to dryness, with the addition of successive portions of hydrochloric acid, the residue is dissolved in water, acidulated with hydrochloric acid, and mixed with an excess of a clear solution of protosulphate of iron. It is set aside for some time in a warm place, till the reduced gold has completely subsided in the form of a fine brown powder, which is then filtered, gently ignited, and weighed. The object in expelling the nitric acid, is to prevent the formation of aqua-regia, which might redissolve a portion of the precipitated gold. Solution of protonitrate of mercury may be employed in the place of protosulphate of iron ; the latter, however, is preferable as the reducing agent. (2.) Reduction by Oxalic Acid. In certain cases it is not convenient to introduce another metal into the solution from which gold is to be precipitated ; oxalic acid is then employed as the reducing agent. The solution as before, is freed from nitric acid, and oxalate of ammonia added; if the solution GOLD. 425 does not already contain free hydrochloric acid, a quantity is next added, and the mixture set aside in a warm place ; the whole of the gold is deposited in the form of yellow scales, which are collected on a filter, washed, gently ignited, and weighed. In cases where gold is the only fixed substance pre- sent, its quantity may be estimated by simply evaporating the solution to dryness and igniting the residue, and, when in com- bination with other substances which are incapable of being precipitated from acid solutions by sulphuretted hydrogen, gold may be separated by that reagent, and the resulting sulphide decomposed by ignition in a platinum crucible. The rationale of the reduction of terchloride of gold by pro- tosulphate of iron is this : The rationale of the reduction by oxalic acid is as follows : AuCl 3 + 3(HO,C 2 O 3 ) = Au + 3HCl + 6C0 2 . Estimation of Gold l>ij standard solutions (Henry, Journ. de Pharm., Jan. 1847). This process, which was devised by the author in consequence of the difficulty which he experi- enced in appreciating very minute quantities of gold, either bv weighing, or by cupellation, in certain investigations in which he was engaged relative to the new processes of gilding and silvering of Elkington and Euolz, is founded on the prin- ciple, that in a mixture of terchloride of gold, a basic salt, and copper, a quantity of the latter metal, equivalent to that of the gold, separated either in powder, or upon the object to be gilt, is dissolved. "When the amount of gold upon a gilt object, or in a bath which has been, or is to be employed in gilding has to be ascertained, the following plan may be adopted. The objects, weighed with care, are digested with hot pure nitric acid ; as soon as the copper forming the basis is dissolved, the solution is diluted with distilled water, and the gold is soon seen to settle at the bottom of the vessel in small brilliant scales. These are collected, and, after washing, dis- solved in aqua-regia ; the solution is evaporated with great pre- caution nearly to dryness, so as to obtain a ruby-red product, soluble in water ; this is terchloride of gold with a little acid. This product is now dissolved in distilled water, and mixed with five or six times its weight of pure bicarbonate of potassa or soda, dissolved in distilled water ; the mixture is heated, conveyed into a ground stoppered flask, and a somewhat large amount of finely divided copper, which has been previously 426 . QUANTITATIVE ANALYSIS. heated in a current of hydrogen, added to it ; the mixture is now and then shaken, and after about an hour the liquid assayed. A very minute quantity of the liquid is poured upon a watch glass and treated with protosulphate of iron ; if the liquid does not yield a black or grey precipitate, it is a sign that it contains no gold in solution ; should the contrary occur, more copper must be added, and the liquid again agitated. AVhen the whole of the gold has been precipitated upon the copper, the liquid is carefully saturated with pure sulphuric acid so as to be slightly acid. By this means all the copper precipi- tated in the state of carbonate is dissolved, without the gold or metallic copper being at all acted upon. It is filtered, and a so- lution of pure ferrocyanide of potassium of known strength care- fully added by means of a graduated burette, until a precipitate ceases to be formed ; the number of divisions of the instrument employed to precipitate the copper is noted, and in this manner the quantity of the metal dissolved in the liquid is ascertained. When it is a solution which is to be, or has been, used for gilding, the author advises to precipitate the diluted acid solu- tion by a current of sulphuretted hydrogen, to collect the pre- cipitate and strongly calcine it after washing. The sulphide of gold being reduced to the metallic state, the calcined residue is redissolved in nitric acid, and the gold which has remained unattacked, dissolved in aqua-regia, and treated as above de- scribed. The use of the ferrocyanide of potassium, to determine the amount of copper which represents the gold in a compound, is founded on the fact, that this reagent is still very sensitive when sulphide of sodium has no longer any perceptible action. The conditions requisite for the success of the operation are : 1st, to take care that the copper employed is perfectly free from oxide; 2nd, to be certain that, after contact with the copper, no gold remains in solution ; 3rd, to saturate the mix- ture exactly with pure sulphuric acid after the reaction ; 4th, to mix as quickly as possible, at a gentle heat, the copper and the bicarbonate with the solution of the terchloride of gold ; 5th, to add the test liquid, which has been made shortly be- fore use, with precaution, and only by drops, when but a slight chestnut or dark-red precipitate is produced. M. Henry quotes a number of experiments, the results of which prove the exactness of the process, as well as the ease and quickness with which it may be performed. GOLD. 427 Separation of Gold from Platinum. The alloy is boiled in aqua-regia, and the gold reduced by oxalic acid ; from the filtered solution the platinum may be thrown down in the metallic state by formic acid ; or the platinum may first be precipitated from the hydrochloric solution by chloride of potassium, and the gold in the filtrate may then be reduced by protosulphate of iron. Separation of Gold from Antimony, Tin, and Arsenic. This may be effected in the same manner as the separation of pla- tinum from these metals, viz. by heating the mixed sulphides in a stream of chlorine gas. The chlorides of tin, antimony, and arsenic distil over, while the gold remains behind. Eisner employs metallic zinc as the precipitant for gold, which he finds to remove the metal from a solution of its chloride more effectually than protosulphate of iron. For the quanti- tative determination of gold, however, in a mixture of platinum, gold, and tin, he prefers protosulphate of iron, on account of the greater simplicity of the operation. He first precipitates the platinum by a concentrated solution of chloride of ammo- nium, and afterwards the gold from the filtrate by a recently- prepared solution of protosulphate of iron. He found that when an excess of this reagent was employed, the filtrate from the precipitate still gave a beautiful dark-red colouring with protoehloride of tin, though no precipitate was observed. Analysis of alloys of Gold. Several methods are employed. In the first place, an approximation to the relative proportions of the constituents is obtained by the touchstone and the assay- needle ; the former is a black and polished basalt ; black flint and pottery will serve the same purpose. The assay-needles are small fillets of gold, alloyed with different and known quan- tities of silver or copper: the sets may consist of pure gold; pure gold 23 J carats, with half a carat of silver ;* 23 carats of gold, with one carat of silver ; 22^ carats of gold, with 1^- carat of silver, and so on till the silver amounts to four carats, after which the additions may proceed by whole carats. Other needles may be made in the same manner, with copper instead * In estimating or expressing the fineness of gold, the whole mass spoken of is supposed to weigh twenty-four carats of twelve grains each, either real, or merely proportional, like the assayers' weights ; and the pure gold is called fine. Thus, if gold is said to be twenty-three carats fine, it is to be understood that, in a mass weighing twenty-four carats, the quantity of pure gold amounts to twenty-three carats. 428 QUANTITATIVE ANALYSIS. of silver, and other sets may have the addition, either of equal parts silver and copper, or of such proportions as the occasions of business may require. When a specimen of gold is about to be examined, it is rubbed on the touchstone, and the colour which it leaves is compared with that communicated to the stone by the assay-needles taken successively ; that which leaves a mark most nearly resembling that produced by the specimen, is in composition most nearly allied to it, and, as the compo- sition of the needle is known, so the operator is enabled to judge of the quantity of silver necessary to be added for the quartation proof. The alloy is next cut into small thin plates, and fused in the cupel (see SILYER), with 3^ times as much pure silver as it contains gold, and with three or four times its weight of lead. After the operation, the gold and the silver remain, the oxide of copper being absorbed with the oxide of lead by the cupel. This process is termed quartation, because the gold forms one-fourth part of the cupelled alloy, a propor- tion which admits of the complete subsequent extraction of the silver by the action of nitric acid. The alloy of gold and silver is next reduced to thin plates, weighed, and gently heated with nitric acid, diluted, and quite free from nitrous and hydrochloric acids ; the silver dissolves, the gold remaining untouched. The acid being saturated, a stronger acid is added, and the solution is gradually brought to boil, by which the perfect separation of the silver is accomplished, the gold retaining still the form of the original plate. It is therefore easily weighed; previous to which, however, it must be washed with distilled water, as long as any traces of silver appear, by the test of common salt, and then heated to redness. The loss which the mass experiences by the process of cupellation, indicates the quantity of copper contained in the alloy, and the proportion of silver is afterwards found by the action of the nitric acid. It must particularly be borne in mind, that if the nitric acid be not free from nitrous acid, and especially from hydrochloric acid, a portion of the gold, sufficient to affect seriously the result of the assay, may be dissolved. Mr. Makin has drawn attention to the loss which occurs in parting operations, and refining on a large scale, from the solu- tion of gold in nitric acid, even when quite free from hydro- chloric acid, iii consequence of the formation of nitrous acid. When the silver is present in large quantity, the solvent ac- tion appears to be restrained by electrical action ; but as the GOLD. 429 silver is removed, the solution of the gold goes on more rapidly. The cause of the evolution of nitrous acid is evident as long as there is any silver present, and it often results from the use of charcoal, to prevent u bumping." When charcoal is thoroughly carbonized, it does not materially affect the acid, but if it con- tain woody matter, nitrous acid is sure to be set free. Purification of Gold by Cementation. The alloy is reduced to a thin plate, and surrounded in a crucible with a pulverulent mixture of four parts of brick-dust, one of calcined vitriol, and one of common salt. It is then exposed for 16 or 18 hours to a strong red-heat. The vapours of hydrochloric and sulphuric acids which are formed, attack the metals mixed with the gold, and the mass is prevented from fusing by the brick-dust. If the first cementation has not been found sufficient to purify the gold, the operation is repeated, but, in the place of common salt, nitre is used. The same method is employed to refine the surface of gold articles after they are polished. The cementa- tion here produces the same effect as tartar and salt, in which silver goods are boiled to give them a white colour. Purification ~by Fusion with Sulphide of Antimony. Some borax is fused in' a crucible, so that the walls become lined with the vitrified flux ; a mixture of two parts of sulphide of anti- mony and one of the gold to be assayed is then introduced. The sulphur combines with the foreign metals, and the anti- mony unites with the gold ; the alloy being removed, the scoriae are a second time fused, with the addition of two parts more of sulphide of antimony, and, when the whole of the gold is ex- tracted, the various alloys are mixed together and heated in an open vessel, with two parts of sulphur. The sulphide of antimony volatilizes, leaving the gold ; to increase the action, a current of air from a pair of bellows may be directed on the surface of the melted mass, or, what is perhaps still better, the mixture of the two metals may be fused in a large crucible, with three times its weight of nitre, by which the antimony be- comes oxidized and dispersed, the gold remaining untouched. Another method is to fuse the gold alloy with a mixture of oxide of lead and sulphur, and to add to the fused mass char- coal in the state of fine powder; the gold is thus obtained alloyed with the lead only, from which it may be separated by cupellation. In this operation it frequently happens that some of the antimony remains alloyed with the gold, a plan for removing 430 QUANTITATIVE ANALYSIS. which has been suggested by Mr. Warrington (Jour. Chem. Soc.,vol. xiii. p. 33). It consists in employing about ten per cent, of oxide of copper, and a small quantity of borax, with which the alloyed gold must be kept in a well-fused state for half an hour. The result is a perfectly malleable gold, containing a small per- centage of copper and well fitted for the purpose of coinage. 187. SELENIUM. When existing in solution in the form of selenious acid, it is reduced by sulphite of ammonia, which separates selenium in the form of a cinnabar-red powder, which, if boiled, becomes black ; it is collected on a filter, and dried at a very gentle heat : nitric acid, if present in the solution, must be decomposed and expelled by hydrochloric acid, previous to the addition of the sulphite of ammonia. When, however, selenium is contained in a solution, as selenic acid, it must first be reduced to seleni- ous acid by boiling with hydrochloric acid, and then precipi- tated by sulphite of' ammonia. As selenium is precipitated from its solution when in the form of selenious acid by sulphuretted hydrogen, it may thus be separated from all those substances which are not precipi- tated by that reagent. Selenic acid does not, however, possess the property of being precipitated by sulphuretted hydrogen ; when selenium, therefore, exists in this state, and has to be separated from other substances, it must be reduced to sele- nious acid. Sulphide of selenium has the following composition : One equivalent of Se . . 39'75 . . 55-40 Two ditto of S . . . JJ2-00 . . 44-GO One ditto of SeS .... 7175 . . 100'OU According to Oppenheim, selenium may be estimated by fusing the body containing' it with cyanide of potassium, dis- solving the fused mass in water, and then supersaturating the solution with hydrochloric acid, which effects the complete separation of the selenium at the end of a few hours. The fusion ought to be performed in an atmosphere of hydrogen ; it is advisable when the substance contains free selenious acid, previously to saturate this acid with an alkaline carbonate, so as to avoid volatilizing small portions before the cyanide of po- tassium has had time to react upon it. The solution obtained by treating the fused mass with water, contains seleno-cyanide SELENIUM. 431 of potassium, together with a small quantity of selenide. It is necessary on this account to boil the liquid for some time be- fore the addition of hydrochloric acid, to convert the seleuide into seleno-cyanide. Without this precaution, a portion of the selenium might be disengaged in the form of seleniuretted hydrogen. The ordinary process for estimating selenic acid consists in precipitating it as a baryta salt. According to Rose, however (Poggendorff's 'Annalen der Physik und Chemie,' cxiii. 472 and 624), seleniate of baryta is much more soluble than sulphate, and it possesses in a high degree the property of carrying down the soluble salts which are contained in the liquid ; it is always better, therefore, to reduce the selenic acid to selenious acid by hydrochloric acid, and then to precipitate the selenium with sul- phurous acid. Selenium cannot be separated from the metals with which it is combined when the sulphides of these metals are insoluble in sulphide of ammonium, by making use of the solubility of sele- nium in this reagent, because the insoluble metallic sulphide is almost always mixed with selenium. Most frequently sele- nium may be separated from metals by heating the mixture in a current of chlorine ; the chlorides of selenium are sufficiently volatile to render the separation generally easy. In acid solu- tions of the selenites of metals not precipitable by sulphuretted hydrogen, the selenium may be precipitated by this gas in the state of sulphide. To estimate the alkalies and alkaline earths combined with selenium, it is sufficient to fuse them with chloride of ammonium. The alkali or alkaline earth remains in the state of chloride ; one single fusion, or two at the most, are sufficient to drive off all the selenium. When it is desired to estimate the selenic acid in an inso- luble combination, particularly in seleniate of baryta, the com- pound is decomposed by an alkaline carbonate, the transfor- mation into alkaline seleniate takes place even in the cold ; it is then easy to reduce the selenic into selenious acid by means of hydrochloric acid. Analysis of Seleniate of Lead. It is reduced to a fine state of division, and diffused in water, through which a current of sulphuretted hydrogen is passed ; the liquor filtered from the sulphide of lead is boiled, to expel the excess of sulphuretted hydrogen, and the selenic acid is precipitated as seleniate of 432 QUANTITATIVE ANALYSIS. baryta, which is decomposed by an alkaline carbonate as above directed. 188. TELLTJRITJM. According to B-OSE (Poggendorff's 'Annaleu,' vol. cxii. p. 307), the best reagent to precipitate tellurium from solutions of tellurous acid is sulphurous acid. The solution should be strongly acidulated with hydrochloric acid. Phosphorous acid may also be used, but not so effectively. To analyse compounds of telluric acid, Eose directs to fuse the compound at a gentle heat with cyanide of potassium in a long-necked flask through which a current of hydrogen is passed. After fusing for about ten minutes, it is allowed to cool in the current of hydrogen ; the flask is then filled up with water, and turned upside down in a beaker filled with the same fluid. The telluride of potassium dissolves, colouring the water red ; it decomposes in the air, depositing tellurium ; this decomposition is hastened by pass- ing a current of air through the solution. The deposited tellu- rium is collected on a filter, and the liquid after the addition of hydrochloric acid is treated with sulphurous acid to reduce a small quantity of tellurous acid which may have been formed. A boiling solution of cyanide of potassium dissolves very little tellurium, telluro-cyanide of potassium is formed in this case; from this solution, tellurium is precipitated by hydrochloric acid. The results obtained by fusing tellurates with alkaline car- bonates are less accurate than when cyanide of potassium is used. Estimation as Sulphide. Tellurous acid is completely preci- pitated from its hydrochloric solution by sulphuretted hydrogen ; it may be dried at 212, and weighed as sulphide, the composi- tion of which is One equivalent of Te . . . 64-5 . . 66'g4 Two ditto of 8 . . . 32-0 . . 33'16 One ditto of TeS . . "OG^ . . lOO'OO Tellurous acid may also be estimated by simple evaporation on the water-bath ; when it is dissolved in other acids, especially in nitric acid, the dry mass is heated in a platinum crucible at not too great a heat. If there be hydrochloric acid in the solu- tion nitric acid must be added. Tellurium, like Selenium, cannot be separated from other metals by sulphide of ammonium, because a portion of tellurium invariably remains in the precipitated sulphides. TUNGSTEN. 433 Separation of Tellurium from Selenium. The substance is fused with ten times its weight of cyanide of potassium in an atmosphere of hydrogen, in w r hich the fused mass is allowed to cool. This is dissolved in abundance of water, and a current of atmospheric air passed through the solution. At the expira- tion of twelve hours the deposited tellurium is separated by nitration, and the selenium is estimated in the liquid in the manner already described. Separation of Selenium, Tellurium, and Sulphur. The sub- stance is fused with cyanide of potassium in an atmosphere of hydrogen, the fused mass is dissolved in water, and the tellu- rium precipitated by a current of air ; the filtered liquid is boiled, the selenium is precipitated with hydrochloric acid, and the sulphur in the filtrate is changed into sulphuric acid by the action of chlorine, after supersaturating with potassa. 189. TUNGSTEN. The method proposed by Berzelius for quantitatively estima- ting tungstic acid, was to precipitate it from its neutral or al- kaline solution in the form of sulphide of tungsten by sulphide of ammonium, and then to add excess of the reagent, by which the sulphide is dissolved ; from this solution it is reprecipitated by nitric acid, and having been washed and dried, it is roasted by a gentle heat, whereby it becomes converted into tungstic acid, in which state it is weighed. The composition of tungstic acid is One equivalent of W (Wolfram) 92 . . 79'31 Three ditto of . . . . _24 . . 20-69 One ditto of WO 3 . . .116 . . lOO'OO A method of determining the quantity of tuugstic acid in neutral salts by nitrate of mercury, has also been described by Berzelius. The following is the plan proposed by M. Marguerite (' Comptes Kendus,' Feb. 3rd, 1845). He recommends it as being very simple and perfectly exact. The salt to be analysed is placed in a small platinum crucible, and several times its weight of pure concentrated sulphuric acid added. At first a gentle heat is employed, and then the temperature is gradually raised to a red-heat. After calcination the residue is composed of an acid sulphate and of free tungstic acid. It is thrown on a filter, washed with w r ater, charged with 434 QUANTITATIVE ANALYSIS. sal-ammoniac, which prevents the tungstic acid from combining with water, and from passing through the filter, which fre- quently happens, even when it has been ignited. When the last washings no longer precipitate chloride of barium, the residue is ignited to drive away the sal-ammoniac, arid a few drops of nitric acid added to obviate any error that might arise from a slight reduction which the tungstic acid may have undergone, and to ensure the perfect combustion of the last traces of char- coal of the filter. In this way the tungstic acid may be directly determined. Analysis of Insoluble Tunff states. They are fused with car- bonate of potassa in a platinum crucible, by which operation tungstate of potassa is formed, which salt, being soluble in water, may be treated with sulphide of ammonium as above directed. Analysis of Wolfram. This mineral, which is a double tnng- state of iron and manganese, is best analysed according to Wohler, by mixing it in a finely divided state with two parts of dry chloride of calcium, and fusing in a platinum crucible. The melted mass is treated with water, which dissolves out the chlorides of manganese and iron, and leaves the tungsten in the form of insoluble tungstate of lijne. Another method { Wohler). The levigated mineral is di- gested with a mixture of four parts of concentrated hydrochloric acid, and one part of nitric acid, until it is completely decom- posed, the solution is then evaporated to dryness on the water- bath, the chloride of manganese and sesquicbloride of iron are dissolved out, the tungstic acid filtered off, washed with alcohol, dissolved in ammonia, again filtered, the solution evaporated, the residual ammonia-salt ignited with access of air, and the tungstic acid weighed ; the filtrate containing al- cohol is evaporated to expel the latter, diluted with water, and the oxides of manganese and iron separated in the usual manner, 190. MOLYBDENUM, This metal is weighed in the form of the grey bisulphide. It is precipitated from its solution in sulphide of ammonium by hydrochloric or acetic acid. The compound thus precipitated is the tersulphide, which may indeed be converted by ignition into molybdic acid (MO 3 ), but it would not be safe to estimate tbe metal in this form, because, at the temperature necessary MOLYBDENUM. 435 to decompose the tersuipbide, a portion of molybdic acid would be volatilized. Rose, therefore, directs the precipitate occa- sioned bv adding hydrochloric or acetic acid to the solution of tersulphide of molybdenum in sulphide of ammonium, to be gently ignited in a small weighed retort, by which it loses one of its atoms of sulphur, leaving the bisulphide, which has the following composition : One equivalent of Mo .... 48 .. 60 Two ditto of 8 32 . . 40 One ditto of Mo S 2 ... 80 .. 100 From all the metallic oxides, the sulphides corresponding to which are not soluble in sulphide of ammonium, molybdenum, like tungsten, may be separated by that reagent. For this pur- pose, the compound is dissolved in excess of hydrochloric acid; should lead or silver be present, the chlorides thus formed are removed by filtration, excess of sulphide of ammonium contain- ing excess of sulphur is added, and the mixture digested for an hour. It is then filtered, the filtrate containing the tersul- phide of molybdenum is supersaturated and digested with ace- tic or hydrochloric acid, and the precipitate treated as above directed. Analysis of Molyldate of Lead (Wohler). The finely pow- dered mineral is completely decomposed by digestion with nitric acid, the solution is neutralized with ammonia, and digested with excess of sulphide of ammonium. The sulphide of lead which is thus formed is filtered off from the dissolved molybdate, washed with dilute sulphide of ammonium, dried at 212, and weighed. Prom the solution, the sulphide of molybdenum is precipitated by dilute nitric acid, collected upon a filter, washed, dried at 212, and weighed. A weighed portion of it is then introduced into a bulb-tube, and heated in a stream of hydrogen until it loses no more sulphur ; from the weight of residual bisulphide calculated for the total amount of the precipitate, that of the molybdic acid is ascertained. 191. TITANIUM. This metal when it exists in solution in the form of titanic acid may be completely precipitated by ammonia, taking care not to use a large excess of the precipitant ; it shrinks very much in drying. When ignited it has a light brown tinge ; its composition is 436 QUANTITATIVE ANALYSIS. One equivalent of Titanium . 25 . . 60-97 Two ditto of Oxygen . JL6 . . 39-Q3 One ditto ofTiO 2 . . 41 .. lOO'OO Acid solutions of titanic acid are precipitated by long con- tinued boiling, but the precipitation is not complete, unless it exist in solution in sulphuric acid. The precipitated titanic acid must be washed with acidulated water, but as this always dissolves a little titanic acid, it cannot be completely separated from other substances in this way. If pure water be employed to wash the precipitated titanic acid, it- gradually carries the whole of it through the filter. Titanic acid after ignition is .insoluble in all acids but concentrated and boiling sulphuric. When fused with pure or carbonated alkalies, it enters into combination with them. The fused mass loses a portion of its alkali by washing with water, and an acid titanate of the alkali is left which is soluble in warm hydrochloric acid. When the solution is diluted and boiled for a long time, nearly the whole of the titanic acid is separated in the form of a heavy white precipitate. If a piece of zinc, iron, or tin be introduced into the hydrochloric solution of an alkaline titanate, the liquor assumes a blue colour, and a blue precipitate is formed which gradually becomes white ; a similar precipitate is formed on add- ing excess of potassa or ammonia to the hydrochloric solution, having previously removed the zinc, tin, or iron. When titanic acid or titanates of bases which do not impart colours to fluxes are heated in contact with a fragment of pure tin, or still better of pure zinc, with microcosmic salt in the inner blowpipe flame, a blue or violet-coloured bead is produced which disappears in the outer flame ; by this reaction, titanic acid is well distin- guished from all other substances. Estimation of Titanic Acid in Silicates Titanic acid exists in many silicates in which its presence may easily be overlooked. Scheerer ('Annalen der Chemie und Pharmacie,' Band cxi. 372) gives the following method for separating and estimating it : When the silicate is decomposed in the ordinary way by acids or by carbonate of soda, the greater part of the titanic acid is found in the precipitate obtained by adding ammonia to the so- lution filtered from the silicic acid but some remains with the latter. To separate this it is heated with hydrofluoric and sul- phuric acids, and the residuum is added to the precipitate which has been previously ignited. The united mass in which besides TITANIUM. 437 titanic acid, alumina, oxide of iron, manganese, and some lime and magnesia may be present, is fused with bisulphate of po- tassa, the temperature being gradually raised, until the greater part of the sulphuric acid in excess, is dissipated. The fused mass is then dissolved in a large quantity of water, and sulphu- retted hydrogen is passed through until the peroxide of iron is reduced to protoxide ; after this, the solution is boiled, and the fluid being kept hot, carbonic acid is passed through ; the bases remain in solution, while the titanic acid is precipitated. Separation of Titanic Acid from the metals of the Fifth Group. This is easily effected by sulphuretted hydrogen, which in an acid solution, does not precipitate titanic acid, whilst all the oxides comprised in this group are thrown down by this rja^^j 1 ^^. agent as sulphides. Separation of Titanic Acid from Oxide of Iron. Roj8& Di- rects to add to the solution containing these two oxide%,firsi;, ^fi, pure tartaric acid in excess, and afterwards ammonia, anditjjjen to precipitate the iron as sulphide, by sulphide of ammonium. If +> * no other fixed substance besides titanic acid be present in^he filtrate from the sulphide of iron, it may be evaporated to cbsy- 3j>^ ness and ignited till all the volatile substances, and the carbon^ ' of the tartaric acid have been burnt off; titanic acid remains. Mosander places the finely divided compound in a porcelain tube, and ignites it in a current of dry hydrogen gas, until the oxide of iron is completely reduced, which is known by aqueous vapour ceasing to be condensed in a glass tube connected with the porcelain tube in which the ignition has taken place. The residuum is allowed to cool in the stream of hydrogen ; it is then weighed, and digested in strong hydrochloric acid which dis- solves the iron and leaves the titanic acid. To determine the titanium which frequently exists in pig- iron, Eiley (Journ. Chem. Soc., vol. xvi. p. 391) proceeds in the following manner : 200 grains of the pig are dissolved in dilute hydrochloric acid ; when nearly the whole is dissolved, and the action of the acid has ceased, more hydrochloric acid is added, and the solution boiled, so as thoroughly to extract the iron. The solution is then thrown on dried counterpoised filters encircling each other, and the residue on the filters well washed to remove all the iron. It is then treated with a dilute solution of caustic potassa, washed once, and then again treated with the alkali so as entirely to remove the silica. The potassa is thoroughly washed out, and the filters washed, first with hydrochloric acid, and then 2 a 438 QUANTITATIVE ANALYSIS. with boiling water ; it is then dried at 250 until the weight is constant. This gives the graphite, on burning which, a residue of a dirty light-brown colour is left. This is fused with bi sul- phate of potassa ; the fused mass is dissolved in cold water and then boiled for some hours ; on standing, the titanic acid sub- sides, it is collected on a filter, washed with dilute hydrochloric acid, dried, ignited, and weighed. II. ACIDS. 192. ACIDS OP SULPHUE. (1.) Sulphuric Acid. This acid is nearly always estimated by converting it into sulphate of baryta. If the substance to be examined be dissolved in water, it must be acidified with pure hydrochloric acid, heated nearly to ebullition, and chloride of barium or nitrate of baryta added till no further precipi- tation occurs. The sulphate of baryta formed, usually subsides quickly and completely, but it should be allowed to settle thoroughly before it is filtered ; it is washed on the filter re- peatedly with boiling distilled water, dried, and ignited, the weight of the ash left by the filter, which has been determined by a previous experiment, being deducted : for the composition of sulphate of baryta, see p. 277. If the substance analysed be dissolved in nitric acid, it must be well diluted with distilled water before it is precipitated by a baryta salt, otherwise nitrate of baryta may be thrown down, which would afterwards require long-continued washing to remove ; in all cases, the operator must be careful to wash the precipitate on the filter till the washings no longer give any cloudiness when dropped into di- lute sulphuric acid. Should the substance under examination be insoluble in water and in hydrochloric acid (sulphates of baryta, strontia, lime, or lead, for example), it must be fused in a platinum crucible (porcelain in the case of lead) with about four times its weight of a mixture of pure carbonate of soda and car- bonate of potassa, the whole having been previously well mixed by trituration in a mortar ; the mass should be kept in a state of fusion for some time, to ensure complete decomposition, and when cool, transferred to a dish, and digested with distilled water repeatedly, till all the soluble matter is removed ; this contains the whole of the sulphuric acid of the compound ana- lysed, in combination with soda and potassa, together with the excess of the alkaline carbonates ; the base or bases with which SULPHUR. 439 the acid had been previously in combination, remain in the form of insoluble carbonates, and are to be determined in the man- ner directed in the preceding pages. The sulphuric acid is pre- cipitated from the solution by nitrate of baryta or chloride of barium, the liquid having been previously supersaturated with hydrochloric acid, or with nitric acid, if lead be present ; the precipitant in that case being nitrate of baryta. Volumetric estimation of Sulphuric Acid (Schwartz). A standard solution of nitrate of lead is prepared, containing 33' 100 grammes in each litre, which solution corresponds to the standard solution of bichromate of potassa employed by the author for the volumetric estimation of lead (see LEAD). This lead solution is added to the fluid containing the sulphuric acid in such proportion as to allow a slight excess of the lead, after which, the precipitate is filtered and washed, and the filtrate being mixed with acetate of soda, the excess of lead is estimated by the standard solution of bichromate. Every cubic centimetre of the lead solution requires O'OOS grammes of sulphuric acid for precipitation ; should the base (as in sul- phate of lime) occasion some inconvenience, it may be removed by boiling with carbonate of soda ; and hydrochloric acid, if pre- sent, must be expelled, by adding nitric acid and evaporating to dryness. Phosphoric and arsenic acids also occasion embar- rassment, as in the presence of acetate of soda, tribasic phos- phate and arsenate of lead are both thrown down ; this me- thod of estimating sulphuric acid would seem therefore to have somewhat limited applications. (2.) Sulphurous Acid. The precipitate which this acid forms with barytic salts is soluble in hydrochloric acid, by which it is at once distinguished from sulphate of baryta ; when a soluble sulphite, or a salt of any of the other lower acids of sulphur is presented for analysis, it is converted into sulphate by careful and protracted digestion with nitric acid, or with aqua-regia : dry sulphites, hyposulphites, and hyposulphates may also be converted into sulphates by fuming nitric acid. "When a mixture of a sulphate and a sulphite has to be examined, two different portions of the compound are analysed; in one the quantity of sulphuric acid is determined in the ordinary man- ner, namely, by precipitating it from its hydrochloric solution by chloride of barium or nitrate of baryta ; the other portion is digested with nitric acid or with aqua-regia, and, the sulphite being thus converted into sulphate, the solution is also precipi- 2 (*2 440 QUANTITATIVE ANALYSIS. tated by a barytic salt. The difference in the two quantities of sulphuric acid found, furnishes the data for calculating the pro- portion of sulphite in the compound. The same method may be adopted for analysing mixtures of sulphates with salts of any of the other lower acids of sulphur. Analysis of a mixture of the different compounds of Sulphur with Oxygen* (Fordos and Grelis). The process is founded on the different action which chlorine and iodine exert on the acids of sulphur. Thus, neither of these bodies exerts any ac- tion on sulphuric or hyposulphuric acids, while they rapidly convert sulphurous into sulphuric acid. On passing a current of chlorine into a dilute solution of a hyposulphite, sulphur is precipitated, and sulphurous acid disengaged ; and on pass- ing the gas into a solution of the sulphyposulphuric acid of Langlois, or the bisulphyposulphuric acid of Fordos and Grelis, the whole of the sulphur becomes converted into sulphuric acid ; again, iodine, though without any action on the sulpho- hyposulphates, exerts an action of a peculiar nature on so- lutions of the hyposulphites. MM. Fordos and Gelis apply these facts to the analysis of mixtures of salts of the different acids of sulphur. Thus, suppose the solution to contain a sul- phate, a sulphite, a hyposulphite, a hyposulphate, and a bisulpho- hyposulphate. The liquid is divided into four equal portions. i. The sulphuric acid is determined by chloride of barium, the precipitated sulphate of baryta being washed first for some time with boiling distilled water, and then with water rendered slightly acid with hydrochloric acid ; the solution must not be acid previous to adding the chloride of barium, or a partial oxi- dation of the sulphurous acid may take place, thus occasioning too large an amount of sulphate of baryta. ii. Another portion is mixed with a considerable quantity * Six of these compounds have been well made out, and the existence of several others is probable; the six well-defined acids have the following com- position: 803,800,8.02,8205,8305,8405, and they have been thus classi- fied by Berzelius : Monothionic SO 3 ,SO 2 . Sulphuric and sulphurous acids. Dithionic . . S 2 O3,S 2 O 5 Hyposulpliurous and hyposulphuric acids. Trithionic . S 3 O 5 . . Sulphyposulphuric acid of Langlois. Tetrathioriic . S 4 5 . . Bisulpho-hyposulphuric acid of Fordos and Gelis. Besides these, a new series of acids of sulphur has been announced by Plessy (' Comptes Rendus,' Aug. 1845), to which he assigns the folio wing compo- sition : S 5 6 ,S 5 O 7 ,S fi O r ,S 8 O 10 ; and M. Wackeuroder, and MM. Fordos and Gelis, have each d\33cribed an acid having the composition S 5 5 . SULPHUR. 441 of carbonate of magnesia, and saturated with iodine, the quan- tity required being carefully noted ; the amount of sulphuric acid contained in the liquid is again determined by chloride of barium ; the quantity of sulphate of baryta obtained will be greater than in the first experiment, and the increase will indi- cate the quantity of sulphurous acid, and the weight of iodine which it was necessary to employ in order to convert it into sulphuric acid ; when this point is attained, it is easy, without having recourse to any other experiment, and by a simple sub- traction, to procure all the elements necessary for the determi- nation of the amount of hyposulphurous acid. The quantity of iodine which had been required to convert the sulphurous acid into sulphuric, is subtracted from the total amount em- ployed ; the difference will have been absorbed by the hyposul- phurous acid ; two equivalents of this acid absorb one of iodine. In treating the liquid with iodine, an alcoholic solution of known strength is employed, or small fragments of iodine are added by degrees to the liquid from a flask, the weight of which has been previously determined. The solution is very rapid, and the point of saturation is easily ascertained. As soon as the liquor acquires a yellow colour, no more should be added. The change of colour is very striking, and it is quite useless to add starch or any other foreign body to the solution. The necessity of adding carbonate of magnesia to the liquid previous to the addition of the iodine, arises from the fact, that water is decomposed by the sulphite, which is also in solution ; it becomes sulphate, taking oxygen from the water ; at the same time the iodine takes the hydrogen, and hydriodic acid is formed ; but this acid would, if at the moment of its production it did not find a base to satu- rate it, act on the undecomposed portion of the sulphite, or on the hyposulphite, which the liquid likewise contains, and there would be a loss of sulphurous acid and a deposit of sulphur. in. This portion of the liquid is employed to determine the amount of bisulpho-hyposulphuric acid. It is saturated with iodine, without, however, keeping any account of the quantity employed. The iodine forms as before, a sulphate at the expense of the sulphite, and a bisulpho-hyposulphate at the expense of the hyposulphite. This quantity will go to in- crease that already contained in the solution. This being done, about 100 parts of water are added to the liquid under examination, through which a current of chlorine is then passed. This gas will bring to the state of sulphate, the whole 442 QUANTITATIVE ANALYSIS. of the sulphur of the bisulpho-hyposulphate, without attacking that of the ordinary hyposulphite. When the saturation is complete, chloride of barium is added to the solution ; the weight of the sulphate of baryta obtained will represent the sulphur of the sulphate, of the sulphite, of the hyposulphite, and of the bisulpho-hyposulphite. As the operations made with the first and second portions of the liquid will have indicated the quantity of sulphur contained in the three first, the dif- ference between the two weights will serve to determine the amount of sulphur contained in the latter, and consequently its entire amount. iv. There will now only remain the hyposulphuric acid. To determine this it will suffice to know the total amount of sulphur in the mixture analysed ; for then, after having assigned to the four other acids the amount pertaining to them, the dif- ference must of course belong to the hyposulphuric acid. The solution is evaporated to dryness, a small quantity of caustic soda being added, to retain the sulphurous gases; the solid residue is treated in the usual manner with fuming nitric acid, and the amount of sulphur determined by chloride of barium. (3.) HydrosulpTiuric Acid (Sulphuretted Hydrogen) : i. Estimation of, in a free state, by a Solution of Iodine in Iodide of Potassium. This method is based on the fact that one equivalent of iodine decomposes one equivalent of hydrosul- phuric acid, the results being hydriodic acid and sulphur : 5 grammes of pure iodine (perfectly dry) are dissolved in a concentrated solution of pure iodide of potassium, and water added till the whole occupies exactly one Utre=lQQQ cubic centimetres ; each cubic centimetre containing therefore 0*005 gramme of iodine. A known quantity of the sulphuretted water to be tested is diluted with distilled water that has been boiled, and cooled out of contact with air, some thin starch paste is mixed with it, and the ioduretted iodide dropped in from a burette until a permanent blue colour is obtained ; now, as one equivalent of iodine =127, corresponds to one equivalent of sulphuretted hydrogen = 17, the quantity of iodine contained in 1000 cubic centimetres of the standard solution of 5 grammes, is equivalent to 0'669 gramme of sulphuretted hydrogen, hence the calculation is easily made. This method was first proposed by Dupasquier. In the analysis of sulphuretted mineral wa- ters, Fresenius recommends that the iodine solution should STJLPHUB. 443 be diluted five times, that is, until each cubic centimetre con- tains O'OOl gramme of iodine. The presence of hyposulphites renders this process inexact, as the iodine is changed into hydriodic acid by hyposulphurous acid as well as by sulphuretted hydrogen. Lyte (' Comptes Rendus,' Oct. 20, 1856) proposes to precipitate the sulphur as sulphide of silver by means of the double hyposulphite of silver and soda dissolved in excess of hyposulphite of soda. This reagent is prepared by dissolving chloride of silver in a solution of hyposulphite of soda : it may be kept for a long time, especially with the addition of one or two drops of am- monia. The sulphuretted hydrogen contained in a mineral water, may be removed by passing a current of pure hydrogen gas through it, and receiving the removed gas in a solution of acetate of lead ; sulphide of lead is precipitated, which is con- verted into sulphate, from the amount of which, that of the sulphuretted hydrogen is calculated. IT. As Sulphide of Arsenic. If the acid be in solution in water, as, for example, in certain mineral springs, the water is mixed with a clear solution of arsenious acid in hydrochlo* ric acid, and the precipitate which forms allowed to subside ; it is then collected on a filter, washed with cold water, dried at 212, and weighed : for the composition of sulphide of arsenic, see page 407. If the hydrosulphuric acid be evolved in the gaseous form, the substance from which it is expelled is heated in a flask, and the gas received into a vessel con- taining a solution of arsenious acid in caustic potassa : the last traces of the gas are removed from the flask by pouring into it, solution of bicarbonate of potassa. The operation being completed, the alkaline ley is supersaturated with hydrochloric acid, and the sulphide of arsenic thereby precipitated, collected, dried, and weighed. Separation of Sulphur from the Alkalies and Alkaline Earths. If the compound be soluble in water, its solution may be mixed with a solution of chloride of copper, upon which sul- phide of the latter metal is precipitated. The precipitate is collected on a filter as quickly as possible, to prevent the oxi- dation of a portion of the sulphide, and then treated with red fuming nitric acid, by which the sulphur is brought into the state of sulphuric acid, and may then be precipitated with the proper precautions, by a soluble salt of baryta. The weight of the sulphate of baryta obtained furnishes data for calculating 444 QUANTITATIVE ANALYSIS. the amount of sulphur present in the substance analysed. If the sulphide to be examined be not readily soluble in water, it may be decomposed in a flask by dilute sulphuric or hydro- chloric acid, the hydrosulphuric acid evolved being conducted through a series of Wolfe's bottles, about two-thirds filled with a neutral solution of chloride of copper ; three bottles may be employed, the latter containing, in addition to the metallic chloride, a little caustic ammonia, to ensure the complete ab- sorption of the hydrosulphuric acid gas. The gas remaining in the flask at the end of the operation must be driven forward into the bottles, by a current of carbonic acid gas. This is ef- fected by pouring into the flask through its funnel, a solution of carbonate of ammonia, not, however, sufficient to neutralize the acid, because it may happen that a portion of sulphur has been deposited in the generating flask during the decomposi- tion, which it would of course be necessary to collect on a tared filter, wash, dry, and weigh; but if the liquor were rendered alkaline (supposing the sulphide of an alkaline earth to have been under examination), an insoluble carbonate would be precipitated, which would render the collection of the sul- phur difficult. The following method for the prompt estimation of the soluble sulphides contained in crude sodas, has been proposed by Lestelle (' Comptes Rendus'). It is based on the insolu- bility of sulphide of silver, and the solubility of all other argen- tiferous salts, in the presence of ammonia. A standard solution of ammonio-nitrate of silver is prepared by dissolving 27'69 grammes of silver in pure nitric acid, adding 250 cubic centi- metres of ammonia, and then diluting with water to the volume 1 Utre=1000 cubic centimetres ; each cubic centimetre of this solution corresponds to O010 gramme of monosulphide of so- dium. The substance to be analysed having been dissolved in water, ammonia is added, and it is then boiled. The ammo- niacal silver liquid is then added, drop by drop, by means of a burette divided into tenths of a cubic centimetre. A black precipitate of sulphide of silver takes place. "When nearly all the sulphur is precipitated, the liquor is filtered, and a fresh quantity of the silver solution is poured in, until after repeated nitrations, a drop of this liquid produces only a slight opacity. The estimation is then at an end, and it is only necessary to read the divisions indicated by the burette, and to compare this number with that of the weight. To estimate very small SULPHITE. 445 quantities of sulphide, the argentiferous liquid must be 0*005 more diluted, so that each cubic centimetre corresponds to gramme of sulphide. Analysis of Compounds of Sulphur with Bases. Sulphur is readily detected in sulphides and in sulphates by fusing the sample with soda upon charcoal before the blowpipe ; a hepar is thus obtained, which, on being moistened, is easily recognized by means of silver ; as, however, it may still be uncertain whether the sulphur existed in the compound in the form of a sulphate or of a sulphide, Yon Kobell has suggested the following simple method of solving the question : The finely pulverized sample is boiled with caustic potassa in a porcelain crucible, and heated till the potassa begins to melt ; or the sample is fused with hydrate of potassa in a platinum spoon before the blowpipe. The mass is then dissolved in a little water, and filtered ; a bright strip of silver is placed in the so- lution ; and, if the sample contain a sulphide, the hepatic re- action is frequently immediately perceptible, but sometimes only after a few hours. With very small quantities, the plati- num spoon, with the flux, may be placed in a small glass with some water, and the silver introduced. The silver is easily restored to brightness by rubbing it with leather and a little caustic lime : sulphates after such treatment have no reaction on the silver. Nitro-prusside of Sodium as a Test for SulpJiur. Any sub- stance containing sulphur, when heated with carbonate of soda, either with or without the addition of carbonaceous matter, according as a deoxidizing action is, or is not required, will yield an alkaline sulphide. On adding to the fused mass a drop of the solution of nitro-prusside of sodium, a magnificent purple colour is produced. This test is so delicate that, ac- cording to Bailey (' Silliman's Journal,' May, 1851), the pre- sence of sulphur may be detected in the smallest particles of coagulated albumen, horn, nails, feathers, mustard seed, etc., which can be conveniently supported on a platinum wire for blowpipe experiments. Dana (' Silliman's Journal,' Nov. 1851) gives the following more minute directions for the use of this elegant test : Heat by the blowpipe, any sulphide or sulphate (or anything containing sulphur) upon charcoal, with carbonate of soda ; put the fused mass into a watch-glass with a drop of water, and a particle not larger than a pin's head, of the nitro-prusside of sodium ; there will be a magnificent purple 446 QUANTITATIVE ANALYSIS. at once. When this test is tried on parings of nails, hair, al- bumen, etc., Dana advises that the carbonate of soda be mixed with a little starch, which appears to prevent the loss of any sulphur by oxidation. If a piece of hair four inches long be coiled round one point of a platinum support, moistened, dipped into a mixture of carbonate of soda with starch, and then heated by the blowpipe, the fused mass will give with the nitro-prusside the unmistakable reaction indicative of sulphur. By this test the presence of sulphur in hops, wines, silk, etc., is easily detected. The substance under examination, is placed in a flask with a piece of sheet zinc, diluted hydrochloric acid is poured over it, and the gas conducted into the solution of nitro-prusside. If the gas contain but a minimum of sulphu- retted hydrogen, the first bubble causes a violet cloud in the solution : after passing the gas for a short time it assumes the magnificent colour of the solution of permanganate of potassa. The vapour of hydrochloric acid passing over with the gas may be arrested by filtering it through cotton-wool, though it does not affect the reaction unless it be continued too long. Metallic sulphides may be analysed in the dry way, by mixing them with thrice their weight of pure dry carbonate of soda, and an equal weight of nitre, to which must be added a con- siderable quantity of pure dry chloride of sodium. If the sul- phide loses sulphur by heat, the mixture is kept for some time in fusion in a porcelain crucible ; by this operation, the sulphur is separated from the metal with which it was originally com- bined, and is transferred in the form of sulphuric acid to the alkali; a soluble sulphate is thus produced, which may be pre- cipitated by chloride of barium, and the quantity of sulphur calculated from the weight of sulphate of baryta formed. This process is applicable to the determination of sulphur in coal and coke. A quantity not exceeding 20 grains of the fuel is intimately mixed with 500 grains of a mixture of 4 parts of pure chloride of sodium, 2 parts of pure nitre, and 1 part of pure carbonate of soda, the mixture is gradually heated in a platinum dish till the whole is in tranquil fusion, and the heat is con- tinued till all blackness has disappeared. When cold, the fused mass is washed out of the dish with boiling water into a beaker, the solution is supersaturated with pure hydrochloric acid, filtered (if necessary), and precipitated at a boiling heat by chloride of barium. MMv v Cloez and Gruignet (' Comptes Eendus,' June 7th, i SULPHUR. 447 1858), suggest the employment of permanganate of potassa as an agent of oxidation for the determination of sulphur in sul- phuretted compounds, and particularly in gunpowder. About one gramme of the material to be analysed, is put into a small ma- trass with a saturated solution of pure permanganate of po- tassa; and the liquid is boiled, with the addition of perman- ganate from time to time, until the mixture has a persistent violet tint ; all the sulphur is hereby converted into sulphuric acid. The oxide of manganese which the liquid holds in sus- pension is dissolved by hydrochloric acid, the solution is concen- trated, if necessary, by evaporation, a little nitric acid added, and then precipitated by chloride of barium. Estimation of Sulphur in Iron and Copper Pyrites. The fol- lowing alkalimetrical method has been proposed by Pelouze (Ann. de Chim. et de Phys., 3rd series, vol. Ixiii.). It is based upon the property which chlorate of potassa possesses in the presence of an alkaline carbonate, of transforming into sulphuric acid, the sulphur contained in metallic sulphides. The sulphuric acid formed neutralizes a portion of the alkaline carbonate, it is only necessary therefore to ascertain how much uncombined alkali exists in the mixture after fusion, to learn the quantity that has been saturated by the sulphuric acid formed from the sulphur of the sulphide, and this is done by a standard solu- tion of sulphuric acid. The normal acid employed by Pelouze is of such a strength that 92'4 cubic centimetres are exactly neutralized by 10 grammes of pure and dry carbonate of soda ; these numbers correspond to equal equivalents of carbonate of soda, and monohydrated sulphuric acid. A litre of normal acid contains 100 grammes of monohydrated acid, or 32*653 of sul- phur. The process is conducted as follows : 1 gramme of the mineral is intimately mixed in a porcelain mortar with 5 grammes of pure and dry carbonate of soda, 7 grammes of chlorate of potassa, and 7 grammes of pure fused or decrepitated common salt, the mixture is heated to dull redness for eight or ten mi- nutes iu a platinum crucible. AVhen nearly cold, the fused mass is washed into a beaker, boiled and filtered, the residue on the filter being well washed with boiling water. The filtrate and washings being well mixed, are now neutralized with the normal sulphuric acid. Example : Suppose 34 cubic centimetres have been required, subtract this from 46'2 (the quantity that would have been required to neutralize 5 grammes of carbonate of soda), and there will remain 12'2 cubic centi- 448 QUANTITATIVE ANALYSIS. metres, which represent the sulphuric acid formed by the py- rites ; this number multiplied by 32-653, and divided by 100, gives the weight of the sulphur sought, viz. '0398, or 39'8 per cent. The process does not occupy more than thirty or forty minutes, and the errors involved in it, do not, according to the author, exceed 1-J- per cent, of the weight of the sulphur to be determined. Determination of Sulphur in Sulphide of Lead. The sulphur is converted into sulphuric acid by digesting the ore in fuming nitric acid : the sulphate of lead produced is dissolved in hot caustic potassa, the alkaline solution is filtered from the insolu- ble matters, and the lead again precipitated as sulphide by sul- phuretted hydrogen, it is received on a filter, washed, dried, and digested in a weighed platinum capsule with fuming red nitric acid ; it is thus again converted into sulphate, in the form of which salt it is weighed. Sulphide of silver and sulphide of bismuth must be decom- posed by pure nitric acid ; the substances should be reduced to as fine a state of division as possible, and projected into a flask capable of being closed with a stopper, and containing the very strongest fuming nitric acid; the bottle must be closed the moment the sulphide is introduced, and occasionally agitated ; when the action slackens a gentle heat may be applied. In the case of sulphide of silver, should ary of the sulphur have escaped oxidation, it may be dissolved by adding, in small portions, chlorate of potassa ; but in the case of sulphide of bismuth, the unoxidized sulphur must be collected on a tared filter, washed with acetic acid, dried, and weighed; from the clear filtrate, the bismuth is precipitated by sulphuretted hydrogen, and from the filtrate from the precipitated sulphide of bismuth, the sul- phuric acid is precipitated by a salt of baryta, hydrochloric acid having been previously added. Separation of Sulphur from metals by Chlorine Gas. This method is applicable to the analysis of compound sulphides of those metals, the chlorides of which are not volatile at a low red-heat. The substance is introduced into a bulbed tube con- nected with an apparatus by means of which a stream of dry chlorine gas may be passed through it. The disposition of the apparatus is seen in Fig. 76 : a is the flask containing the ma- terials for generating chlorine gas, which is dried by passing first through the Wolfe's bottle b, containing concentrated sulphuric acid, and then through the siphon-shaped tube c, containing SULPHUR. 449 chloride of calcium; the sulphide to be. analysed is weighed, and carefully introduced into the bulbed tube d, bent at a right angle, and dipping into water contained in the bottle e. If the process cannot be conducted under a flue, it may be advisable to close the mouth of the bottle e with a cork, through which a tube is inserted, which may convey the excess of chlorine into another bottle, containing milk of lime, or potassa, thus prevent- ing annoyance to the operator by the escape of the irritating gas into the apartment. The whole apparatus is first filled with chlorine, which is made evident by the yellow-green colour of the space above the water in the bottle e ; a very gentle heat is then applied to the bulb, the sulphide is decomposed, its me- tallic base or bases combining with the chlorine, while the sul- phur is driven forward into the bottle e in the form of chloride of sulphur, which is immediately decomposed by the water into hyposulphurous acid, hydrochloric acid, and sulphur; the former, by the action of the chlorine, passes into sulphurous acid, and finally into sulphuric acid, so that the last result of decomposi- tion is sulphuric acid, hydrochloric acid, and a greater or less quantity of free sulphur. When the action is over, the bulb is heated towards the bend, so as to force the whole of the chloride of sulphur into the bottle e ; the tube is then cut off under the bend, and the separated end thrown into the bottle. The ex- cess of chlorine is expelled from e by the application of a gentle heat, it is allowed to remain at rest for some hours till the sul- 450 QUANTITATIVE ANALYSIS. phur has completely solidified, it is then collected on a tared filter, washed, dried, and weighed, and the sulphuric acid is lastly precipitated by nitrate of baryta or chloride of barium. Rose applies the above method to the analysis of mixed sul- phides of those metals the chlorides of which are either partly or wholly volatile at the temperature employed. For example, in the analysis of the mixed sulphides of antimony, iron, and zinc ; in such a case, the water in the bottle e must be mixed with tartaric acid and a small quantity of hydrochloric acid ; and the chloride of antimony which distils over into the bottle e is precipitated by sulphuretted hydrogen, after the free sul- phur has been removed by filtration, and the sulphuric acid by a soluble salt of baryta. If the heat employed has not been too great, the whole of the chloride of zinc, and the greater part of the perchloride of iron, will be retained in the bulb ; a certain portion of the latter, however, will have been volatilized, and will have found its way into the liquid in the bottle e, from which, together with any chloride of zinc which may also have distilled over, it must be precipitated after the removal of the antimony, by sulphide of ammonium. Determination of Sulphur in Nitrogenous organic com- pounds (Fleitmann). Sulphur exists in nitrogenous organic substances, such as caseine, albumen, and fibrine, in two dif- ferent forms ; one portion is removed by heating the sub- stance in a concentrated solution of caustic potassa ; the other portion remains combined with the organic substance. Fleit- mann determines the amount of this combined sulphur, by first dissolving the compound in caustic potassa, and then di- gesting the solution for some hours, with frequent agitation, with a sufficient quantity of recently-precipitated hydrated oxide of bismuth ; when cold, the liquid is supersaturated and digested for some time with excess of acetic acid ; it is then filtered, and the washed sulphide of bismuth is oxidized with the filter in a silver crucible, by fusion with potassa and nitre, and the sulphur thus converted into sulphuric acid, is deter- mined in the usual way as sulphate of baryta. Estimation of Sulphur in volatile organic compounds (Debus, Liebig's ' Annalen,' Oct. 1850). One equivalent of bichro- mate of potassa which has been purified by recrystallization, is dissolved with two equivalents of carbonate of potassa, or soda, in water, and the liquid evaporated to dryness ; in this way, a lemon-coloured mass, consisting of one equivalent of neutral SULPHUR. 451 cliromate of potassa and half an equivalent of carbonate of potassa or soda, is obtained. It is powdered, and, in order to remove the last portions of water, exposed to a good red-heat in a Hessian crucible, and then filled while hot, into a glass tube sealed at one end and drawn out somewhat at the other, so that when wanted, the chromate of potassa can be conve- niently poured into the combustion-tube without loss. This precaution is necessary with the use of carbonate of potassa, as it absorbs water from the air, and, on being subsequently heated in the combustion-tube, again parts with it, and cracks the tube. After the chromate of potassa has cooled in the above-mentioned corked glass tube, a layer of it of about three to four inches in length is poured into the combustion-tube, such as is used in organic analysis ; the substance is then added, and then again a few inches of the mixture of chromate and carbonate of potassa. When the substance to be analysed is a solid, it is intimately mixed with the chromate of potassa in the combustion-tube, by means of a wire, twisted at one end like a corkscrew. When this is done, the empty portion of the combustion-tube is filled with a mixture of chromate and car- bonate of potassa, placed in a furnace, and heated as for an organic analysis. When the whole is incandescent, oxygen is disengaged from a small retort, and passed in a slow current over the red-hot mixture for about half an hour or an hour. When cold, the tube is cleansed of adherent ash, and cut into several pieces over a sheet of paper; these are placed in a beaker, and so much water poured over them as may be necessary to dissolve the calcined mass. The liquid thus obtained is strongly acidified with hydrochloric acid, the chromic acid reduced with alcohol, and boiled till it becomes green. It is now fil- tered from the insoluble oxide of chromium, which is washed, first with water acidified with hydrochloric acid, afterwards with alcohol, and then dried. The filtered green liquid, which contains nearly the whole amount of sulphuric acid, is placed aside for' a time, and the oxide of chromium (which always most tenaciously retains a small quantity, frequently one or two per cent., of sulphuric acid, in the form of a basic salt) is collected on a filter and dried. The oxide of chromium is then emptied into a platinum crucible, and the filter burnt. It is mixed with one part of chlorate and two of carbonate of potassa, and heated to redness until the whole of the oxide of chromium is converted into chromate of potassa. The fused mass is dis- 452 QUANTITATIVE ANALYSIS. solved in dilute muriatic acid, the chromic acid reduced by alcohol, and the liquid added to that previously obtained. The united solutions are now heated to boiling, the sulphuric acid precipitated by chloride of barium, and the amount of sulphur calculated from the sulphate of baryta formed. 193. ACIDS or PHOSPHOBUS. Phosphoric Acid. In a solution which contains nothing but phosphoric acid, the amount of that acid may be esti- mated, by evaporating to dryness, with the addition of a known weight of recently ignited protoxide of lead ; the residue after being heated to redness is weighed ; the increase of weight in- dicates the amount of phosphoric acid. The presence of nitric acid does not interfere with the accuracy of this method. The quantitative determination of phosphoric acid, when in combination with bases, and in the presence of other acids, is frequently attended with considerable difficulties, especially when it exists in small quantities only. The following are the principal methods at present in use. (1.) Estimation as Pyrophosphate of Magnesia. Applicable when the phosphoric acid is in combination with an alkali. Chloride of ammonium, ammonia, and sulphate of magnesia are added to the solution, which is then Avell agitated, or stirred with a feather ; a glass rod should not be used for the purpose, as the crystalline double salt which is precipitated, adheres so firmly to those parts of the sides of the beaker which have been touched by the glass rod, that it cannot afterwards be readily removed. The mixture is allowed to remain at rest for several hours ; it is then filtered, and the precipitate washed with water containing a little ammonia, in which it is almost insoluble ; the washing must not be continued too long, but stopped immediately it is found that a few drops of the filtrate, when evaporated on pla- tinum foil, leave no residue ; it is then ignited, first gently, and afterwards strongly : for the composition of the ignited salt, see page 281. This method does not apply to the analysis of phosphates after they have been ignited, without a previous preparation, as has been shown by Weber (Chem. Graz., vol. v. p. 450); the reason is, that they are converted by heat from tribasic into bibasic salts, and bibasic or pyrophosphoric acid forms with magnesia a salt which is much more soluble in water than tribasic or common phosphoric acid. By the ac- tion of acids it is true that pyrophosphoric acid is converted PHOSPHORUS. 453 into the tribasic acid, but Weber found that the transformation was not perfect, even when the compound was treated for a considerable time with concentrated sulphuric acid. In order to ascertain which of these two phosphoric acids we have under examination, the following test may be applied. The compound is dissolved in nitric acid, and precipitated by ammonia ; the precipitate, after long standing, is collected on a filter ; to the filtrate, nitrate of silver is added, and the ammoniacal liquid is then carefully neutralized with nitric acid ; the appearance of a white turbidness, or the formation of white flakes, indi- cates the presence of pyrophosphoric acid. If the phosphoric be in combination with an alkali, it will, if the acid be tribasic, give a yellow precipitate with nitrate of silver (see p. 146). When the phosphoric acid in a phosphate is to be separated and determined by magnesia, and is not contained in the solution as the tribasic compound, the phosphate must be fused with excess of carbonate of soda. This can be done in all cases when the phosphoric acid is combined with the alkalies, or those metals which are completely decomposed by fusion with carbo- nate of soda ; but it does not succeed in converting the phos- phoric acid into the tribasic compound, when in combination with the alkaline earths, as these are only imperfectly decom- posed. The only way to reprecipitate the pyrophosphate of magnesia from its solutions by ammonia, is to heat it for some time with concentrated sulphuric acid, when, on solution, it can be almost entirely precipitated by ammonia. The precipitates which the tribasic and bibasic phosphoric acids produce with magnesia, differ essentially from each other in their external properties. The first is granular, crystalline, and soon subsides, especially when warmed ; the latter is vo- luminous and flocculent, and remains long suspended in the liquid. Small quantities of pyrophosphoric acid cannot be se- parated, or even detected, by magnesia. (2.) Estimation by Nitrate of Silver (Chancel). This method is founded on the insolubility of the yellow phosphate of silver (3AgO,PO 5 ) in a neutral liquid. The weighed substance is dissolved in nitric acid ; the solution is diluted with distilled water, and to the clear liquid is added, first a sufficient quantity of nitrate of silver, and then a slight excess of carbonate of silver. The saturation of the free nitric acid takes place ra- pidly, without the application of heat ; in a few moments the yellow phosphate of silver is thrown down. When the super- PAET II. 2 H 454 QUANTITATIVE ANALYSIS. natant liquid no longer reddens litmus-paper, the precipita- tion is complete. The precipitate is collected on a filter and washed. When the washing is finished, the filter is per- forated by a platinum wire, and the whole of the precipitate washed out into a flask, and dissolved in nitric acid. The silver is then precipitated by hydrochloric acid, and from the filtrate, the phosphoric acid is thrown down as ammonio-mag- nesian phosphate, by the addition of an ammoniacal solution of sulphate of magnesia. (3.) Estimation as Phosphate of Sesquioxide of Uranium. This method is well adapted for the separation of phosphoric acid from the alkaline earths, particularly magnesia, a separation usually attended with much difficulty. It does not succeed so well in the presence .of iron and alumina, these substances being in part precipitated with the phosphate. To the solu- tion of the phosphate in hydrochloric acid, or in water, nitrate of uranium is added; it is then supersaturated with ammonia, and afterwards acidified with acetic acid : phosphate of ura- nium is precipitated. The precipitation is known to be com- plete, by the supernatant fluid having a yellow tint: if in- complete, a fresh quantity of nitrate of uranium is added, and afterwards ammonia and acetic acid. The precipitate is al- lowed to deposit from the warm solution, and is washed several times by decantation with boiling water : the settling of the precipitate is* favoured by the addition of chloride of ammonium. It is collected on a filter, washed with hot water containing chloride of ammonium, and dried : the dry preci- pitate is removed as completely as possible from the filter, cal- cined in a platinum crucible, the filter being burnt separately. The composition of phosphate of sesquioxide of uranium is Two equivalents of U 2 O 3 . . . 288'0 . . 80*22 One ditto of PO 5 . . . . 71'0 . 19'78 One ditto of 2(U 2 3 )P0 5 . sljSK) . . 100-00 By the following modification, suggested by K. Arendt and "VV. Kopp (Chem. Centralb. 1857), phosphoric acid may be es- timated by acetate of sesquioxide of uranium, in the presence of iron. The substance is dissolved in the smallest possible quantity of hydrochloric acid, and the sesquioxide of iron is reduced to protoxide, by means of protochloride of uranium. This salt is somewhat troublesome to prepare. The authors direct to dissolve ammonio-carbonate of uranium in double PHOSPHORUS. 455 the quantity of hydrochloric acid required to effect solution : to add a few drops of solution of bichloride of platinum, and to throw into the mixture some fine copper turnings. The mixture is boiled for ten or fifteen minutes, or until the liquid assumes a distinct green colour. To remove the subchloride of copper, the boiling is continued until water produces a co~ pious precipitate in a sample of it. When this is attained, the whole is diluted with water and allowed to cool, it is then filtered, and sulphuretted hydrogen in excess is transmitted through the filtrate : the precipitated sulphide of copper is fil- tered off, a considerable quantity of chloride of ammonium is added, and the solution is boiled until every trace of sulphu- retted hydrogen has disappeared. The solution of protochloride of uranium thus obtained is a powerful reducing agent. The protochloride of uranium thus prepared is added to the solu- tion containing phosphoric acid and peroxide of iron, until a distinct green colour is produced, or until a drop of sul- pho-cyanide of potassium does not produce a red colour. Am- monia is then added to alkaline reaction, then acetate of ses- quioxide of uranium, and free acetic acid, together with a few idrops of acetate of protoxide of uranium, (obtained by preci- pitating the protochloride with ammonia, and redissolving in warm acetic acid,) and the mixture is heated to boiling. The liquor must have a distinct greenish colour, not a dirty tinge, which shows the presence of undissolved protoxide of uranium. It is set aside till the precipitate has settled, the supernatant liquid is then decanted on a filter, the precipitate is boiled with water containing chloride of ammonium, and again de- canted ; this operation is repeated once or twice, and finally the precipitate is thrown on the filter, washed, dried, and ignited in the manner above directed. (4.) Volumetric determination of Phosphoric Acid ly Ace- tate of Uranium (Pincus, Journ. fur Prakt. Chemie, Ixxvi.) A solution of phosphate of soda, the amount of phosphoric acid in which is accurately known, must be prepared ; its strength should be such that each cubic centimetre exactly represents O01 of phosphoric acid. The acetate of uranium is prepared by dissolving pure ammonio-carboriate of uranium in acetic acid; it must be perfectly free from protoxide, to ensure which, the solution must be protected from the light. The strength of the solution is determined by pouring 5 to 10 cubic centimetres of the solution of phosphate of soda into a 2 H 2 456 QUANTITATIVE ANALYSIS. beaker, adding ammonia, and an excess of acetic acid, and then pouring the solution of uranium from a burette with divisions of ^th to gVth cub. centim., stirring the mixture frequently during the process. The peculiar slimy precipitate of phos- phate of uranium and ammonia is produced. From time to time a drop of the mixture is put upon a white porcelain plate, close to a drop of solution of ferrocyanide of potassium, and the two drops are allowed to flow together ; as long as the phosphoric acid is not all precipitated, only a bluish-green colour is produced : but when the smallest excess of oxide of uranium is present, the spot which is at first bluish-green, be- comes distinctly surrounded with a darker or lighter choco- late-coloured border : if the solution be very dilute, the reac- tion disappears again in a few minutes, after strong stirring ; a further quantity of the solution of uranium must then be added. If the brownish-red colour has remained permanent for ten minutes, it does not again disappear, but makes its ap- pearance still more distinctly, without any admixture of yellow and green, when, after the settlement of the slimy precipitate, a drop of the clear supernatant fluid is employed for the re- action. "When it has been clearly ascertained by repeated ex- periments, how many cubic centimetres of the solution of ura- nium are necessary for the precipitation of 0'05 or O'lO gramme of phosphoric acid, the solution may by calculation, and by adding water, be brought nearly to such a state of dilution that each cubic centimetre thereof may precipitate the phos- phoric acid contained in one cubic centimetre of that solution, namely, O'lO gramme. By repeated experiments, and by the ad- dition of small quantities of water, or of a concentrated solu- tion of oxide of uranium, it may be easily managed that the two solutions may be equivalent within the limits of y^th to a^th cubic centimetre, for quantities of 10 to 20 cubic cen- timetres ; so that the highest possible error may lie within the limits of 0*0005 and O'OOl gramme of phosphoric acid. To apply this method, the solution containing the phos- phoric acid to be determined is mixed, whether neutral or acid, with ammonia, acetate of soda, and then with acetic acid in excess. Every T Vth cubic centimetre of solution of oxide of uranium employed until the production of the reaction, repre- sents O'OOl gramme of phosphoric acid ; the deposition of the precipitate is facilitated by heating the solution. The solu- tion must not contain either protoxide of iron or alumina. PHOSPHORUS. 457 (5.) Estimation as Phosphate of Bismuth (Chancel, ' Comptes Hendus,' t. i.) One part of pure crystalline subnitrate of bismuth (Bi0 3 ,N0 5 -f-aq.) is dissolved in 4 parts of nitric acid, sp. gr. 1'36, adding to the solution 30 parts of distilled water, and then boiling and filtering, if necessary. Each cubic centimetre of such solution will precipitate from 7 to 8 mil- ligrammes of phosphoric acid. The substance to be analysed is dissolved in nitric acid, avoiding too great an excess ; it is then diluted with distilled water, and the acid nitrate of. bis- muth added, until nothing more is precipitated. It is then boiled, filtered, and washed with boiling water, until the wash- ings are no longer coloured by sulphuretted hydrogen. The precipitate is now carefully dried, and removed from the filter as completely as possible ; the filter is incinerated by itself, and the residue added to the dried precipitate, which is then calcined. The composition of phosphate of bismuth is : One equivalent of BiO 3 . 234'0 . . 7672 One ditto of P0 5 . 71-Q . . 23-28 305-0 100-00 Consequently the weight of the precipitate multiplied by 0'2328 will give the weight of the phosphoric acid in the sub- stance analysed. If either chlorine or sulphuric acid be present, they must be removed before the nitrate of bismuth is added. (6.) Estimation of Phosphoric Acid, founded on the insolubility of Phosphate of Tin in Nitric Acid in the presence of Stannic Acid, and on the solubility of that compound in Sulphide of Ammonium. (Grirard, ' Bulletin de la Societe Chimique de Paris.) The substance is dissolved in nitric acid ; when there are any difficulties in getting it into solution with this acid, it may be dissolved, first in any convenient reagent, and precipitated by ammonia ; the well-washed precipitate then dissolves readily in nitric acid. Into the solution is thrown an indefinite quan- tity of pure tin, about four or five times the supposed weight of the phosphoric acid present, may be employed. The tin passes into the state of stannic acid, carrying down with it the whole of the phosphoric acid, as well as a great portion of the iron, and a little alumina ; it is washed, first by decantation, and then on a filter ; the filtrate and washings contain all the bases minus a portion of iron and alumina. The precipitate containing the phosphoric acid, is redissolved 458 QUANTITATIVE ANALYSIS. in n small quantity of aqua-regia, the solution supersaturated with ammonia, and then with excess of sulphide of ammonium. The stannic acid and phosphate of tin immediately dissolve, leaving a black precipitate of sulphide of iron, containing alumina. It is allowed to stand for an hour or two, and then filtered, taking care to wash the precipitate with sulphide ol ammonium, to remove the last traces of tin. It is then only necessary to add sulphate of magnesia to the filtered liquid, to obtain the characteristic precipitate of ammonio-phosphate ol' magnesia. (7.) Volumetric estimation of Phosphoric Acid ~by PercTiloride of Iron (Liebig's method modified by E. "W". Davy. L. and E. Phil. Mag. and 'Chemical News,' vol. i. p. 181). A standard solution of perchloride of iron is made by dissolving clean pianoforte wire in pure hydrochloric acid, and adding suffi- cient nitric acid to convert the protochloride into perchloride. Any free hydrochloric acid is then carefully neutralized with caustic ammonia, which is added until a little peroxide re- mains undissolved on shaking the mixture. Acetic acid is now added to dissolve the precipitated oxide, and when the solu- tion is effected, it is largely diluted with distilled water, and graduated in the ordinary way, so that the amount of iron in a given quantity may bo known. This solution may be kept a considerable time without undergoing any change. The phos- phate to be analysed is dissolved in an acid, ammonia is added until the solution is decidedly alkaline, but not in large excess, and then a sufficient quantity of acetic acid to dissolve com- pletely the precipitated phosphate, and leave a slight excess. The standard solution of iron is then added carefully from the burette, till the iron begins to be in slight excess, which is as- certained by taking a drop of the mixture (after it has remained a few minutes with occasional stirring) on a glass rod and touching with it a piece of thick filtering-paper placed over another piece which has been soaked in a solution of gallic acid and dried. The insoluble phosphate of iron is retained on the filtering-paper, and the solution which passes down to the lower paper, shows at once by the purple stain when sufficient iron has been added, and a minute excess exists in the mixture. If this excess be very minute, the stain will become more visible when the gallic paper is dried.* The results should be controlled by * Instead of employing gallic acid to indicate the precise moment when the iron solution is in excess, a few drops of solution of sulphocyanide of po- PHOSPHORUS. 459 repeating this experiment a second and a third time, the phos- phate being dissolved in a given quantity of solution, and a certain amount of it being taken for each determination. The compound of iron and phosphoric acid has the invariable composition (Fe 3 O 3 ,PO 5 ). In all cases, bibasic salts must be converted into tribasic, by heating with hydrochloric acid, and the solution must be allowed to cool before the estimation is made. This method may be usefully applied to the determination of phosphoric acid in urine, in superphosphate of lime, and in manures generally. (8.) Estimation of Phosphoric Acid .in Soils (Schulze, ' Annalen der Chemie und Pharmacie,' vol. cix.). About 50 grammes of the earth are ignited to destroy the organic matters, and then digested for a long time with hydrochloric acid. The filtered liquid is nearly neutralized with dilute ammonia, care being taken to avoid a permanent precipitation. To the solution, which should amount to about a pint and a half, 35 or 45 drops of terchloride of antimony are added, and it is set aside for 24 hours. During this time, there is deposited a yellowish-white flocculent precipitate which contains all the phosphoric acid, but is principally composed of antimonic acid carrying with it some oxide of iron and alumina ; it contains, besides, a quantity of ammonia in proportion to that of the phosphoric acid. The precipitate, well washed with distilled water, is boiled with soda ley containing a certain amount of silicate. After boiling, the liquor is filtered ; the oxide of antimony changed into antimoniate of soda remains on the filter with the alumina and oxide of ir5n. The filtered alkaline solution which contains the phosphoric acid with a little alumina and silica, is super- saturated with hydrochloric acid and ammonia, concentrated by evaporation, again treated by ammonia, and filtered. The pre- cipitate of alumina and silica thus obtained, carries with it a small quantity of phosphoric acid, to remove which, it is re- dissolved in a drop or two of hydrochloric acid, the solution evaporated to dryness, and the residue wanned with a little acidulated water. After having filtered again to separate the silica, a small quantity of tartaric acid is added to the filtered tassium may be added to the solution of the phosphate before commencing the experiment. The solution will remain colourless until the whole of the phosphate is precipitated. But the blood-red colour characteristic of sul- phocyanide of iron will make its appearance on the addition of a drop of the iron solution in excess. 460 QUANTITATIVE ANALYSIS. liquor, which is then mixed with the preceding ammoniacal solu- tion, containing the greater part of the phosphoric acid. From these solutions the phosphoric acid is separated as ammonio- magnesian phosphate in the usual manner. (9.) Estimation of Phosphoric Acid ly Molyldate of Ammonia. According to Lipowitz (Pogg. Annal. der Physik uud Chem. vol. clix.), to obtain accurate results by this method, the molyb- date should be prepared by dissolving at a gentle heat, 2 parts of pure molybdic acid and 1 part of tartarie acid in 15 parts of water, and afterwards adding 10 parts of ammonia, sp. gr. 0'97, and 15 parts of nitric acid. The whole is heated to ebullition in a porcelain basin, whereupon about -^h of the molybdic acid is deposited, which is separated by filtration. In employing the above solution, the quantity of it thought necessary is placed in a capsule and heated to boiling, and then the liquid containing the phosphoric acid is added. The preci- pitate is washed with water acidulated with -j^tti of nitric acid, and the filter is dried at 80 or 90 F. between folds of blotting- paper, or better still, over sulphuric acid. The composition of phospho-molybdate of ammonia is Molybdic Acid 90' 744 Phosphoric Acid 3'142 Ammonia 3'570 Water 2-544 100-000 (10.) Separation of Phosphoric Acid from Bases ly means of Ni- tric Acid and metallic 3Iercury(H. Rose). The phosphatic com- pound is dissolved in a sufficient quantity of nitric acid, the so- lution is conveyed to a porcelain dish, and so much metallic mercury added that a small portion is left undissolved by the free acid. It is then evaporated to dryness on the water-bath the dry mass again moistened with water, and again evaporated, this operation being repeated until the dry mass no longer smells of nitric acid while warm. It is next mixed well with water, filtered through a small filter, and washed till a few drops of the wash- water leave a residue, which disappears entirely on ignition. The filtrate contains all the bases combined with nitric acid together with nitrate of mercury, which is sepa- rated by hydrochloric acid ; the solution is again filtered, and quickly washed, some chloride of ammonium having been pre- viously added. If no iron or alumina are present, the alkalies, PHOSPHORUS. 461 lime, magnesia, and other metallic oxides existing as nitrates in the filtered liquid, are separated from each other in the usual manner; or the nitrate of mercury may be removed from the solution of the nitrates, by evaporating the solution in a porcelain dish, and exposing the residue to a red-heat in a pla- tinum crucible. It should however be observed, that when alkaline nitrates are present, the dry residue on ignition must be mixed with a small quantity of dry carbonate of ammonia, in order to convert into carbonates, the free alkalies produced by the decomposition of the nitrates. If this precaution be neg- lected, the platinum crucible is very much acted upon. The calcined residue is dissolved in dilute hydrochloric acid, and the bases separated in the usual manner. The phosphoric acid remains on the filter in combination with the mercury. There are likewise present nitrate and me- tallic mercury. The whole is well dried, removed from the fil- ter, and mixed in a platinum crucible with an excess of carbo- nate of soda, the filter is rolled up and thrust into the mixture, which is then covered with a layer of carbonate of soda, and exposed for half an hour under a chimney to a moderate but not red heat, that the contents may not fuse. At this temperature the whole of the metallic mercury and the mercurial salts, with the exception of the phosphate, are expelled. The mass is now fused, and the fused mass treated with hot water, in which it dissolves entirely, if the operation has been conducted with care, and no iron were present. It is supersaturated with hy- drochloric acid, and the phosphoric acid estimated as ammonio- magnesian phosphate. When peroxide of iron is present, the greater portion is left on treating the dry residue with water, with the phosphate of mercury, and only a small portion dissolves with the nitrates ; it is filtered and washed in the usual way, the peroxide of iron in the solution determined with the other bases, and the insoluble portion fused with carbonate of soda. On treating the fused mass with water, the peroxide of iron is obtained perfectly free from phosphoric acid, whilst the entire amount of phosphoric acid in combination with soda is dissolved. When alumina is present this method becomes inapplicable. Separation of Phosphoric Acid from Phosphorous and Hypo- phosphorous Acids (Rose). The solution is poured gradually and in small quantities into a solution of chloride of mercury which must be fully saturated; a precipitate of subchloride o'f 462 QUANTITATIVE ANALYSIS. mercury, exhibiting the splendour of mother-of-pearl, is deter- mined, but its complete separation requires gentle digestion for several days : the subchloride of mercury is collected on a filter, dried by exposure to a gentle heat, and weighed. From its weight it is easy to calculate how much oxygen has been taken up by the phosphorous or hypophosphorous acid to effect its conversion into phosphoric acid : hence the quan- tity of either of these acids is readily inferred. This quantity of oxygen is equivalent to the quantity of chlorine contained in the subchloride of mercury, since the chloride of that metal, on being converted into subchloride, loses exactly one- half of its chlorine. Another quantity of the solution to be analysed is treated with nitric acid, and a weighed quantity of oxide of lead, in the manner already described, the object being to determine the whole quantity of phosphoric acid, both that originally contained in the solution, and that formed from the phosphorous or hypophosphorous acid, at the expense of the oxygen of the nitric acid. Now, it is learnt by the first operation how much phosphoric acid is produced by the phos- phorous or hypophosphorous acid of the solution, and, by de- ducting this from the whole quantity obtained by the second operation, the actual proportion of phosphoric acid in the so- lution analysed is known. According to Wohler, phosphorous acid is converted into phosphoric acid by being heated with sulphurous acid, sulphu- retted hydrogen being formed and sulphur separated. In this manner he states that phosphorous acid may easily be detected in phosphoric acid : when arsenious acid is present at the same time, the sulphide of arsenic is immediately formed on treat- ment with sulphurous acid. Phosphorous acid in phosphoric acid may also be detected by Marsh's apparatus, by the phos- phuretted hydrogen gas which is disengaged. The gas burns with a whitish flame, and when held close to a surface of porce- lain, a green ring of light is observed in the expanded flame similar to that afforded by phosphorus when burnt in chlorine gas, or with an insufficient access of atmospheric air. Analysis of Bone Earth. Dissolve in hydrochloric acid, filter (if necessary), add oxalate of ammonia, and boil; then add excess of acetate of ammonia, and filter to separate the oxalate of lime ; add tartaric acid to the filtrate, then ammoni- acal sulphate of magnesia, and determine the phosphoric acid as pyrophosphate of magnesia. BORON. 463 194. BORON. Quantitative determination of Boracic Acid (Stromeyer, Liebig Annalen, C. p. 80). The boracic acid, if free, is com- bined with a sufficient quantity of potassa, pure hydrofluoric acid is then added in excess, and the whole is evaporated to dryness. Gelatinous borofluoride of potassium is produced, which subsequently forms hard, transparent crystals. The dry saline mass is then stirred up at the ordinary temperature with a solution of acetate of potassa of 20 per cent., and left to stand for some hours, when the fluid is passed through a weighed filter, and washed with the solution as long as the filtrate gives a precipitate with chloride of calcium. The acetate of potassa is then washed away with alcohol of specific gravity 0'851, and the salt is dried at 212 F. The presence of salts of the alkalies does not interfere with the method, but other bases must be separated, which is easily done in all cases by fusing the borates with carbonate of potassa. The borofluoride of potassium should be filtered through funnels made of gutta-percha or india-rubber, because the fluoride of calcium running away, attacks glass. The evaporation of the borate of potassa with hydrofluoric acid is effected in vessels of silver or platinum. The composition of borofluoride of potassium is One equivalent of K .... 39'0 . . 30'95 One ditto of B .... 11-0 . . 873 Pour ditto of F .... 76-0 . . 60*32 One ditto of (KF,BF 3 ) . 126^ . . 100-00 Boracic acid may also be determined by adding to the solu- tion a known quantity of pure oxide of lead, evaporating to dryness, igniting and weighing the residue: the amount of boracic acid is calculated from the increase of weight. Boracic acid, even when existing alone in a solution, cannot be esti- mated by evaporating to dryness, because although it is one of the most fixed of all substances at an intense heat, it pos- sesses the property of volatilizing with the vapour of water or alcohol. Separation of Boracic Acid from those fixed Bases with which it forms compounds decomposable ~by Sulphuric Acid. The following method was proposed by Arfwedson : A known weight of the compound is reduced to a fine powder, and 464 QUANTITATIVE ANALYSTS. mixed in a platinum crucible with four parts of pure pulverized fluor spar. It is then moistened with sulphuric acid, and heated, first gently, and then gradually to redness, at which tempe- rature it is maintained as long as fumes continue to be evolved. The whole of the boracic acid is thus driven off in the form of fluoride of boron (B0 3 + 3F1H=BF] 3 + 3HO). The bases, together with gypsum, remain behind, in combination with sulphuric acid ; they are separated from each other according to the directions which have been given in the preceding pages, and the amount of boracic acid is calculated from the loss sus- tained by the original substance, after deducting from it the quantities of the different bases found ; these determinations must therefore be made with great care, since the whole loss falls on the boracic acid. When the bases with which boracic acid is combined are precipitable by sulphuretted hydrogen, or by sulphide of am- monium, these reagents are employed to separate them ; their quantities must be accurately determined, and deducted from the weight of the compound submitted to analysis; the dif- ference of weight is the quantity of boracic acid. Prom the alkaline earths boracic acid is separated in the same manner as phosphoric acid, the amount of acid being calculated from the loss. Determination of Boracic Acid in Silicates undecomposable l>y Acids (Axinite, Tourmalines}. The following process adopted by Gmelin, and given by Hose, is probably the best. The mineral is finely levigated, and heated strongly with carbonate of baryta. The ignited mass is treated with a sufficient quan- tity of hydrochloric acid to decompose it, and evaporated to dryness on the water-bath. The silica is separated from the residual mass in the usual manner, and the baryta precipitated from the filtered solution by carbonate of ammonia. The clear liquid is again evaporated to dryness and gently ignited ; it is then weighed, and treated with alcohol and hydrochloric acid. The alcohol is inflamed, and in this way the boracic acid is got rid of; the residue is once more ignited and weighed, and the quantity of boracic acid is indicated by the loss of weight. Another method, recommended by the same chemist, is to ignite the pulverized mineral with carbonate of soda, and to treat the ignited mass with water. The solu- tion having been digested with carbonate of ammonia, to remove the small portions of silica and alumina which the BORON. 465 water had dissolved, is evaporated to dryness ; the dry mass treated with sulphuric acid, and then the boracic acid dissolved out by alcohol. The solution is finally saturated with am- monia, and the residue, which consists of boracic acid, is ignited and weighed. Separation of Boracic Acid from Phosphoric Acid. The phosphoric acid may be precipitated by sulphate of magnesia, and determined as pyrophosphate, or by Yon Kobel's method, by the addition, first, of perchloride of iron, and then of car- bonate of lime ; not a particle of boracic acid is precipitated, but the separation of the phosphoric acid is complete. 195. SILICON. Silicic Acid. This acid is always converted from its soluble to its insoluble condition, for the purpose of weighing. The conversion is effected by evaporating the solution to dryness with hydrochloric acid, or any other volatile acid. The silicic acid is generally first obtained in a gelatinous state ; but, by continuing the heat, it loses the whole of its water, and is brought to the condition of a dry powder. The evaporating mass must be constantly stirred with a glass rod during its exsiccation, to prevent spirting. The bases with which the silicic acid was combined, are separated by dilute acid, and the silicic acid, having been thoroughly washed and dried, is ignited in a platinum crucible, and weighed, when cold, with the cover on, the dry powder absorbing moisture rapidly from the air. The composition of silicic acid is One equivalent of Si . . . 14*0 . . 46'67 Two ditto of O . . . 16-0 . . 53'33 One ditto of Si0 2 . . 3OO . . 100-00 Separation of Silicic Acid from Bases : Quantitative Analysis of Natural Silicates. The great natural family of silicates may be arranged in three classes : 1st, those easily decomposable by hydrochloric acid ; 2nd, those only decomposable by prolonged digestion with con- centrated acid; and 3rd, those which absolutely resist the action of acids, even after prolonged digestion. The following list of the names and composition of a series of interesting minerals belonging to each of these classes is taken from Eose's 'Treatise on Analysis' (English edition, translated by A. Ncrmandy) : 460 QUANTITATIVE ANALYSIS. 1. Silicates decomposable by Hydrochloric Acid in the Cold. 1. Apophyllite . . 2. Natrolite . . . 3. Scolezite . . . 4. Mesolite . . . 5. Mesole . . . . 6. Analcime . . . 7. Laumonite . . . 8. Potassa harmotone 9. Leucite . . . . 10. Eleolite .... 11. Brewsterite . . 12. Sodalite . . . 13. Cronstedtite . . 14. llvaite . . . . 15. G-ehlenite . . . 16. Wernerite . . . 17. Tabular spar . . 18. Nepheline . . . 19. Cancrinite . . . 20. Mellinite . . . 21. Chabasite . . . 22. Pectolite . . . 23. Okenite . . . 24. Davyne . . . . 25. Sodolenite . . 26. Allophane . . . 27. Helvine . . . , 28. Datholite . . . 29. Botryolite . . . Combination of Potassa, silica, lime, and water. Silica, soda, and water. Silica, alumina, lime, soda, and water. Ditto ditto ditto. Ditto ditto ditto. Silica, alumina, soda, and water. Silica, alumina, lime, and water. Silica, alumina, baryta, potassa, and water ; sometimes lime. Potassa, silica, and alumina. Silica, alumina, lime, potassa, soda, and water. Silica, alumina, strontia, baryta, lime, and water. Soda, silica, and alumina, with a small quantity of hydrochloric acid. Silica, oxide of iron, and water. Silica, protoxide of iron, and lime. Silica, alumina, lime, and oxide of iron. Silica, alumina, lime, and soda. Silica and lime. Soda, silica, and alumina. Soda, silica, and lime. Silica, magnesia, lime, and oxide of iron. Silica, alumina, lime, and water, with a little potassa. Silica, lime, soda, potassa, oxide of iron, and water. Silica, lime, soda, potassa, oxide of iron, oxide of manganese, and water. Silica, alumina, lime, iron, and water. Yttria, silica, glucina, oxides of cerium, and iron. Alumina, silica, and water. Silica, glucina, alumina, protoxide of iron and of manganese. Silica, boracic acid, lime, and water. Differs from the preceding in contain- ing one atom more water. SILICON. 467 30. Haiiyne . ' .^-:i 31. Nosian . . . . 32. Lazulite . . . . 33. Eudialite . . . 84. Orthite . 35. Electric Calamine 36. Sideroschisolite 37. Hisingei'ite -v -, 38. Dioptase . . 39. Meerschaum 40. Copper, Malachite Combination of Potassa or soda, silica, alumina, lime, and sulphuric acid. Sesquisilicate of alumina and soda (Klaproth). Silica, alumina, lime, oxide of iron, mag- nesia, soda, and sulphuric acid. Silica, soda, zirconia, lime, oxides of iron and manganese, hydrochloric acid, and water. Silica, alumina, oxides of iron, cerium, lanthanum, and manganese ; lime, yttria, magnesia, and a small quantity of water. Oxide of zinc, silica, and water. Protoxide of iron, silica, alumina, and water. Silicate of protoxide and sesquioxide of iron + six atoms of water. Silicate of copper and water. Magnesia, carbonic acid, silica, water, a little alumina, and traces of manga- nese and lime. Carbonate and silicate of copper and water. The above minerals, when finely pulverized, form a jelly when hydrochloric acid is poured upon them. The next nine do not form a jelly, and some of them are only decomposed by protracted digestion with hot hydrochloric acid. Nearly all these minerals resist altogether the action of acids after being ignited. 1. Stilbite .... Silicate of alumina, lime, and water. 2. Epistilbite . . . Ditto ditto. 3. Heulandite . . . Tersilicate of alumina and lime. 4. Anorthite . . . Silica, alumina, lime, and magnesia. 5. Sphene or Pitanite Tersilicate of lime and titanate of lime. 6. Pyrosmalite . . . Lime, tersilicate of oxide of iron and of manganese. 7. Cerite Hydrated silicate of peroxide of ce- rium. 468 QUANTITATIVE ANALYSIS. Combination of 8. Cerine or Allanite . Silicate of alumina and of cerium, of iron and of lime. 9. Pitchblende . . . Uranium ore, containing from two to five per cent, of silica, probably in a state of mechanical, not chemical, combination. Silicates only decomposable by Fusion with Carbonate of Potassa or Soda. 1. Felspar. 12. Mica. 26. Dichroite. 2. Albite. 13. Lepidolite. 27. Emerald. 3. Rhiacolite. 14. Talc. 28. Euclase. 4. Petalite. 15. Chlorite. 29. Phenakite. 5. Spodumene 16. Pinite. 30. Tourmaline. (Soda Spodu- 17. Achmite. 31. Axinite. mene). 18. Amphibole. 32. Topaz. 6. Oligoclase. 19. Anthrophyllite. 33. Chondrodite. 7. Labradorite. 20. Pyroxene. 34. Picrosmine. 8. Barytic harino- 21. Diallage. 35. Carpholite. tone. 22. Chatoyant or 36. Steatite. 9. Olivine. Schiller Spar. 37. Serpentine. 10. Prehuite. 23. Epidote. 38. Pumice stone. 11. Carbonated 24. Idocrase. 39. Obsidian. Manganese. 25. Grarnet, 40. Pitch stone. To this class belong also the different species of false gems and artificial glass. Silicates which resist the action of Acids and fusion with Alka- line Carbonates, but which are decomposed by ignition with . pure Potash. \. Zircon. 3. Cymophane. 5. Andalusite. 2. Cyanite. 4. Staurolite. Decomposition of Silicates by Acids. The mineral is reduced to the finest possible powder in an agate mortar. It is dried at 212, avoiding a higher temperature, because many of these compounds contain water and other volatile matters which must not be expelled ; and because, moreover, an elevated tempera- ture might interfere with the readiness with which the silicate is decomposed by acids. The pulverized mineral, having been deprived of its hygrometric water, is weighed in a platinum or porcelain dish, and digested at a gentle heat until complete de- SILICON. 469 composition has taken place, which is known by rubbing a glass rod against the bottom and sides of the vessel; if it produce a grating noise it shows that the decomposition is not complete, and the digestion must be continued till the end of the rod glides smoothly over the bottom of the dish. The separated silicic acid does not, in all cases, assume the same appearance, as was remarked above ; sometimes it separates in a bulky gelatinous state, in other cases no jelly is formed, the acid being liberated in the form of a fine powder : some silicates, again, are decomposed almost instantaneously, while others re- quire long-continued digestion. The operator must bear all these facts in mind ; and, above all, he must assure himself of the complete decomposition of the mineral, or all his time and labour will have been expended in vain. The decomposition being complete, the mass is evaporated to dryness on the water- bath, and the residue heated until all moisture is expelled ; the dry mass is then digested with hydrochloric acid, which din- solves out all the bases, leaving the silicic acid in its insoluble modification: it is collected on a filter, washed with boiling dis- tilled water, dried, ignited, and weighed. The filtrate and the washings are quantitatively examined for bases, according to the instructions given in the last chapter. Rose objects to the evapo- ration of the silicate to dryness after its decomposition by hydro- chloric acid, partly on the grounds that the whole of the silica is not thereby rendered insoluble, and that portions have still to be separated in the course of the analysis, but chiefly because it is possible that during the evaporation, certain volatile consti- tuents of the compound may be expelled, and thus escape de- tection, which, he says, has happened with many chemists who have pursued this method. Compounds which are easily de- composed by hydrochloric acid, should never, according to this eminent chemist, be exposed to heat, but digested with cold acid, only employing heat when the decomposition will not pro- ceed without it. In cases where the silicic acid is separated in the gelatinous form, Rose treats the mass with water, and thus separates the silicic acid in light flocks. Most chemists, however, prefer the method of evaporating the whole mass at once to dryuess ; it is the simplest, and on the whole the least troublesome, and if the mass has been thoroughly exsiccated, there is very r little fear of any of the silicic acid finding its way again into the solution. In the case of those minerals which contain volatile constituents, their nature will generally be PART II. 2 I 470 QUANTITATIVE ANALYSIS. made known by the preliminary examination, and the course of analysis will be shaped accordingly ; it is evident, however, that these particular cases cannot come within the limits of a de- scription of a general method, to which the above instructions must be understood to apply. (2.) Decomposition of Silicates l>y fusion with Alkaline Carbo- nate. The mineral reduced to the finest possible. stateof division, (by elutriation if necessary,) is mixed in a platinum crucible with from three to four times its weight of pure anhydrous carbonate of soda, or with an equal quantity of well-dried and previously pulverized carbonate of potassa, or, which perhaps is still better, with a mixture of the two carbonates in the proportion of single equivalents, such as is obtained by calcining Eochelle salt (tar- trate of potassa and soda), lixiviating the calcined mass in water, and evaporating the solution to dryness. The mixing of the pulverized mineral with the alkaline carbonate is made by a rounded glass rod ; it must be very intimate, and all portions ad- hering to the rod must be carefully wiped off into the crucible. The cover is placed loosely on the crucible, and it is gradually raised to intense ignition, at which it must be maintained from thirty minutes to an hour ; at the end of this time the mass will be either completely fused, or in a state of semi-fusion, ac- cording as the mineral contains more or less silicic acid. The contents of the crucible when cold, are transferred to a beaker, which can be covered with a glass plate. When the mass has only been in a state of semi-fusion, it is generally easily removed by bending and pressing the sides of the crucible ; but when the fusion has been complete, it is sometimes not very easy to remove it without having recourse to hydrochloric acid : in the event of this being necessary, the crucible should be placed in a. large beaker before the acid is poured on it, in order to avoid loss by the sudden evolution of carbonic acid. The contents of the crucible being safely lodged in the beaker, the mass is drenched with water, and hydrochloric acid added cautiously, to prevent too violent an effervescence ; it is then covered with its plate, and exposed to a gentle heat until perfect solution has taken place, or until only slight flakes of silica are seen floating in it, by which it is known that the decomposition of the mine- ral has been complete. If, on the other hand, a heavy gritty powder should subside to the bottom of the vessel, the ope- rator hereby learns that, either from not having been in suffi- ciently fine a state of division, or from the heat not having SILICON. 471 been sufficiently intense or long continued, or from some other cause, the mineral has not been entirely decomposed ; and the powder must be carefully collected, washed, dried, and weighed, and the weight deducted from the quantity originally taken ; or, which is always best, the whole experiment must be reoom- nieuced with a fresh portion of substance. Supposing, however, the decomposition to have been successfully performed, the silica is separated by evaporation to dry ness on the water-bath, and the analysis proceeded with precisely in the same manner as in the case of a silicate, decomposable by simple digestion with hydrochloric acid. It must be observed that the silicic acid, as separated from minerals of difficult decomposition, may contain a considerable quantity, 12 or 15 per cent, of alumina, and still form a trans- parent glass when fused with soda. It is always necessary, therefore, to examine the silica obtained from such minerals very carefully for that earth, with which view it is fused with carbonate of potassa, the melted mass treated with hydrochloric acid, and the solution evaporated to dryness. The dry mass is moistened with hydrochloric acid, and afterwards treated with water. The acid solution, filtered from the undissolved silicic acid, is now supersaturated with ammonia. If the solu- ion give no precipitate, the silicic acid may be considered to have been pure ; but, if a precipitate be formed, this can only arise from an impurity, which, in most cases, will prove to be alumina. (3.) Decomposition of Silicates by Hydrofluoric Acid. The finely powdered mineral is placed in a shallow platinum dish b, supported on a lead tripod, standing in the centre of a lead dish a, Fig. 77, about six inches in diameter. The bottom of this lead ves- sel is covered to the depth of from one-eighth to one- j,. HH fourth of an inch, with a mixture, made to the consistence of paste, of powdered fluor- spar and sulphuric acid. The dish is provided with a flat cover, made of the same metal. The mineral to be decomposed is moistened with a little water, and the cover having been placed on the dish, the latter is gently heated over the sand- bath underneath a flue ; hydrofluoric acid is liberated, by means 2 i2 472 QUANTITATIVE ANALYSIS. of which from 20 to 30 grains of siliceous mineral may be completely decomposed in the course of an hour and a half. During the process, the powder must be twice moistened with a few drops of water ; if it be well spread out on the platinum dish, it is unnecessary to stir it. When the operation is finished, concentrated sulphuric acid is added, drop by drop, to the powder, as long as hydrofluosilicic acid is given off; u gentle heat is at the same time applied, and finally the excess of sulphuric acid is driven off by continued heat and evapo- ration to dryness. The dry residue, after being moistened with hydrochloric acid, is boiled in water, and further examined in the usual way. This method is particularly applicable to the examination of such minerals as contain alkalies : or the finely pulverized silicate may be digested with rather concentrated hydrofluoric acid and a gentle heat, sulphuric acid added cau- tiously in excess, and the mixture evaporated to dryness, first on the water- bath and afterwards at a higher temperature, until the excess of sulphuric acid is expelled. If, on the addition of water aided by heat, the dry residue be completely dissolved, the decomposition of the silicate is complete ; if, on the, con- trary, a residue remain, this must be again treated with hydro- fluoric acid. The evaporation must be conducted under a hood so as to carry off the fumes of hydrofluoric acid, which are very corrosive and injurious. Another method of analysing such silicates, proposed by Abich, is to fuse them with hydrate of baryta or with carbonate of baryta, using from 4 to 6 parts of the carbonate to 1 of the mineral ; the most intense heat is required, no action taking place till a state of fusion has been induced ; but, when once the action has commenced, it goes on rapidly and energetically, and the operation is finished in a quarter of an hour. J\ o silicate has yet been found to withstand the action of" this agent, but the most powerful heat attainable in a wind furnace of the best construction is required. When hydrate of baryta is employed, the operation may be conducted in a silver crucible over a good spirit lamp ; from 4 to 5 parts of the hydrate, deprived of its water of crystallization, are mixed intimately with L of the powdered mineral, and the mixture is then covered with a layer of carbonate of baryta. The decomposition being complete, the analysis is proceeded with in the usual manner. Fluoride of ammonium may also be employed for the decom- position of silicates ; it is made by saturating hydrofluoric acid SILICON. 473 with ammonia. The impure acid which generally contains some fluoride of silicon, lead, and iron, will answer for the pur- 3ose. A small quantity of carbonate of ammonia and sulphide of ammonium must be added to the solution, which after a time s decanted and evaporated in a platinum crucible. During the evaporation, a small piece of carbonate of ammonia must be added from time to time ; when the mass becomes pasty, it must ?e stirred with a spatula ; when well dried, the salt may be kept In vessels of platinum, silver, or gutta-percha. To apply this salt to the analysis of silicates, Hose directs ;hus (Poggendorff's Annalen der Physik und Chem., vol. cviii.). Mix the finely pulverized mineral with six times its weight of ;he fluoride in a platinum crucible ; add a little water, and heat at first gently, and then to bright redness, and continue the leat as long as vapours are disengaged. Treat the residue with sulphuric acid, and drive off the excess of this acid by evapora- ;ion. When the sulphates formed are not completely dissolved 3V water acidulated with hydrochloric acid, the insoluble resi- due must be treated with a fresh portion of the fluoride. The :emperature should not be raised too high during the operation, ? or in that case, if the silicate contain alumina, a fluoride of alu- riinum will be formed difficult to decompose by sulphuric acid. (4.) .Decomposition of Silicates by fusion with Carbonate of Lime and Chloride of Ammonium (Lawrence Smith, ' Silliman's Journal,' July, 1853). Mix intimately in a glazed mortar, one part of the finely pul- verized mineral with five or six parts of carbonate of lime, and lalf to three- fourths of pure chloride of ammonium ; introduce the mixture into a platinum crucible, and heat to bright red- ness from thirty to forty minutes. There is no silicate which after having undergone this process is not easily dissolved by lydrochloric acid. For the action of the lime to have been complete, it is not necessary that the mass should have settled down to perfect fusion. The contents of the crucible are dis- solved in dilute hydrochloric acid, and the solution evaporated :o dryness, the operation may be completed over a lamp with- out the danger of the spitting, which occurs when the fusion s made with soda. To the dry mass hydrochloric acid is added, t is then diluted and boiled ; when cold, carbonate of ammonia s slowly added in excess, and the precipitate which occurs is filtered off. The filtrate contains chloride of ammonium, chlo- rides of calcium, and perhaps of magnesium, and the alkalies 474 QUANTITATIVE ANALYSIS. of the mineral. It is concentrated by evaporation, a'nd nitric acid added in the proportion of about three parts to every one of chloride of ammonium, the hitter is completely decom- posed by the nitric acid at a low temperature, into gaseous pro- ducts, and much trouble is avoided, the evaporation proceed- ing rapidly. When dry, the residue is dissolved in water, lime water is added, and the liquid is boiled and filtered. To the filtrate, a sufficient quantity of carbonate of ammonia is added to separate the lime, which is filtered off, and the filtrate is again evaporated to dryness. Dilute sulphuric acid is now added, by which the nitric and hydrochloric acids are expelled, and the alkalies remain in the form of sulphates. For the con- version of the alkaline sulphates into chlorides, acetate of lead is added in slight excess, in order to precipitate the sulphuric acid ; the sulphate of lead is filtered off; the lead is removed from the filtrate by sulphuretted hydrogen, the sulphide of lead is filtered off', excess of hydrochloric acid is added to the filtrate, by which the acetates of the alkalies are converted into chlorides, and the liquid is evaporated to dryness. "When the object is the quantitative estimation of the alkalies only, in the siliceous compound, (analysis of soils, cinders, etc.) the above operations may be considerably curtailed. After the mineral or product under examination is perfectly decomposed by fusion with carbo- nate of lime and chloride of ammonium, the mass is heated with water alone, until thoroughly disintegrated, when it is thrown upon a filter, and thoroughly washed. The filtrate contains merely the chlorides of the alkalies, a little chloride of calcium, and caustic lime. A small quantity of carbonate of ammonia is added to precipitate the two latter substances, the carbonate of lime is separated by filtration, the filtrate is evaporated and gently ignited, and the residue consists of the alkalies in the state of chlorides. 196. CARBONIC ACID. (1.) Quantitative determination of Carbonic Acid in a free state in aqueous solution. This is very conveniently effected by adding to a weighed or measured quantity of the solution, a clear mix- ture of chloride of barium and excess of ammonia ; the latter absorbs the carbonic acid, and carbonate of ammonia is formed which, reacting with the chloride of barium gives rise to chloride of ammonium and carbonate of baryta : the latter is precipitated. It is collected on a filter and washed with water, to which am- CARBONIC ACID. 475 monia has been added ; the funnel must be carefully protected from the air during nitration, otherwise carbonic acid might be absorbed, which would increase the amount of carbonate of baryta. The precipitate having been well washed, is dried and gently ignited, together with the filter, it may then be weighed, and the amount of carbonic acid calculated from the weight of the carbonate of baryta; but should the original solution contain any substances which might also be precipitated by ammonia, and chloride of barium, the amount of carbonic acid in the pre- cipitate should be determined in the apparatus of Will and Fresenius (Fig. 69, p. 268). A. process for determining the amount of free carbonic acid in a water, apart from that which is combined with lime and magnesia in the form of soluble bicarbonate?, has been suggested by M. Graultier de Claubry (Comptes Rendus, vol. xlviii.). It consists in passing through the water a current of atmospheric air which has been freed from carbonic acid ; the free carbonic acid in the water only is re- moved, the bicarbonates remaining in solution; after passing through the water, the air is dried and then passed through a weighed Liebig's potassa-bulb. (2.) Estimation of Carbonic Acid in the gaseous state. The volume of gas is first accurately measured in a graduated tube standing over mercury ; a small piece of moistened caustic po- tassa, attached to a piece of ignited iron wire, is then passed through the mercury into the tube ; when the absorption has ceased, the potassa is withdrawn, and the volume of the unab- sorbed gas accurately determined : the difference in the two measurements indicates the quantity of carbonic acid. When the volume of the gas is large, and contained in a receiver which is not graduated, it is sometimes determined by the increase in weight of a piece of caustic potassa passed into the receiver, and allowed to remain for some time in contact with the gas. (3.) Estimation of Carbonic Acid in Aerated Waters. A piece of india-rubber tubing a (Fig. 78) is attached to a cork borer 5 or 6 inches long, having small holes pierced in its sides ; a bent tube leading into a flask containing pure am- monia, is securely connected with the other end of the vulca- nized tube. On boring the cork the plug carried down stops the tube until the lateral holes come below the lower surface of the cork, when effervescence commences, the liberated car- 476 QUANTITATIVE ANALYSIS. bonic acid being absorbed by ammonia. As soon as all visible action has ceased, the vessels are disconnected, and the contents of the bottle emptied into the flask containing the solution of ammonia; chloride of barium is then added, and the whole is boiled. The amount of carbonic acid in the precipitate is de- Fig. 78. termined in the apparatus of Will and Fresenius. This repre- sents the total quantity of carbonic acid, free and combined. To ascertain the quantity in the latter state, the contents of a similar bottle of the water are evaporated to dryness, the resi- due is very gently ignited, and the amount of carbonic acid de- termined ; this is deducted from the total quantity found by the first experiment ; the difference is the weight of the carbonic acid existing in the water in a gaseous state. Analysis of Carbonates. We have already described at some length the method which is easiest of execution, and which, on the whole, gives the most satisfactory results, viz. that of Fresenius and Will. Fig. 79 shows a little apparatus that mav sometimes be conveniently employed. It consists of a small thin-bottomed flask, capable of containing about four CARBONIC ACID. 477 ounces of water. It is fitted with a cork, through which is in- serted a bent tube adapted to another wide tube, containing frag- ments of dried chloride of calcium ; the extremity of this tube is drawn out to a capillary orifice. The small tube seen inside the flask is intended to hold the sulphuric or hydrochloric acid, with which it is proposed to de- compose the carbonate. It should rest against the side of the flask at an angle of about 45 from the bot- tom, so that by inclining the flask the whole of the acid can be made to flow out. The carbonate to be ana- lysed is accurately weighed, and pro- jected into the flask, which should Kg. 79. contain about an ounce of water ; and, the chloride of calcium tube having been fixed air-tight into its place, the whole appa- ratus is accurately weighed. Great care must be taken that none of the acid in the tube comes into contact with the carbonate before the operation is commenced. Everything being ready, the flask is inclined so as to allow a portion of the acid to flow out of the tube ; effervescence immediately takes place, carbonic acid is expelled, and escapes through the orifice of the chloride of calcium tube, being thoroughly dried during its passage through the lime salt. As soon as the effervescence has ceased, a fresh portion of acid is caused to flow out of the tube, and the operation is repeated until no further escape of carbonic acid gas is perceived. The flask is now placed in a vessel of boiling water, where it is allowed to remain for some time in order to expel all the carbonic acid in the liquor, and in the upper part of the flask, aqueous vapour is prevented from escaping by the chloride of calcium. On cooling, atmospheric air finds its way into the flask, which is thus brought to the same condition it was in, previous to commencing the experiment. It is now weighed ; the loss of weight indicates the amount of carbonic acid. This is a very convenient little apparatus for the analysis of limestones and marls. Many other forms have been given to this apparatus ; that by F. Mohr, shown in Fig. 80, is neat, light, and convenient. The acid with which the carbonate is to be decomposed is con- 478 QUANTITATIVE ANALYSIS. tained in the bulb, it is introduced into the flask, containing the carbonate to be decomposed, by relaxing the india-rubber tube adapt- ed to the upper end, by the compres- sion of the clamp ; the exit tube is filled with fragments of chloride of calcium. Another form of apparatus is shown in Pig. 81. The acid is contained in the test-tube, which is connected with the flask containing the carbonate to be analysed by the siphon b, a; by applying suction at d the acid is intro- duced into the flask, and the liberated carbonic acid gas escapes through the chloride of calcium tube c, by which it becomes dried. Mohr's apparatus is represented in Fig. 82 ; it is especially adapted for de- termining the amount of carbonic acid Fig. 80. in the carbonates of the metallic oxides, and in mortars, cements, etc. The weighed substance is intro- duced into the flask 5, together with some water, and a little tincture of litmus; d\ contains strong hy- drochloric acid. The] flask a is about one- eighth filled with so- lution of ammonia, free from carbonic acid, and the tube c contains fragments of coarsely pound- ed glass, moistened with the same liquid. Everything being ar- ranged, and the appa- ratus having been proved to be tight, Fig. 81. the clip of the tube d is relaxed so as to allow the acid to flow CARBOLIC ACID. 479 gradually into I, which should then be gradually heated till the liquid boils ; it is allowed to cool, and again boiled, and the same operation is several times repeated, until it is certain that Fig. 82. the whole of the carbonic acid has been carried into the flask a. The liquid in b, at the end of the operation, must be intensely red, and the end of the siphon-tube must not reach the surface of the ammonia solution in a. When the apparatus is cold, it is taken apart ; solution of chloride of barium is added to the contents of a, which are then boiled, rapidly filtered, and the amount of carbonic acid in the precipitate determined in the usual manner. The analysis of a great many solid carbonates may be effected by simply igniting them, and estimating the amount of carbonic acid from, the loss of weight; it is, of course, indispensable that the substance shall contain no other volatile 480 QUANTITATIVE ANALYSIS. constituents, which may be expelled by the ignition, together with the carbonic acid. It is also necessary that the sub- stance from which the carbonic acid is expelled shall not itself undergo any alteration, whereby it may be either increased or diminished in weight. The carbonates of certain metallic oxides are in this condition ; as, for example, the carbonates of protoxide of iron and of manganese, and of oxide of cobalt. The analysis of these salts may be performed by igniting them in a bulbed tube of hard glass, through which a stream of dry carbonic acid gas is passing. Carbonates which con- tain water, are analysed by the same process which is adopted in determining the carbon and hydrogen in organic substances, and which will be fully described hereafter. The decomposition of carbonates by heat is greatly facili- tated by mixing them with about four times their weight of finely powdered borax, as was first pointed out by Count F. Schaffgotsch (Poggendorff's 'Annalen'). He did not find anhydrous boracic acid so convenient for the purpose, since by long melting this acid decreases slowly but constantly in weight. The anhydrous biborate of soda, or glass of borax, is, however, absolutely fixed, and is well adapted for the pur- pose, from the ease with which it decomposes the carbonates in fusing, and from its causing 110 ebullition on account of its thick fluidity. Separation of Carbonic Acid from Sulphurous A.cid. For this purpose Persoz introduces into a graduated tube con- taining the mixed gases, a glass rod covered with starch paste, and coated with powdered iodate of potassa or soda ; the sul- phurous acid is hereby converted into sulphuric acid, and as soon as no further decrease in the quantity of the confined gas is perceptible, the glass rod is removed, and the residual gas immediately measured. Before, however, the carbonic acid can be considered as free from sulphurous acid, another glass rod must be introduced, which is also covered with the paste made with starch and a dilute solution of iodate of potassa ; if the least trace of sulphurous acid exist in the gas, the rod is in- stantly coloured blue. This method is therefore well adapted for the detection of minute traces of sulphurous acid. Separation of Carbonic Acid from all other Acids. As car- bonic acid is readily removed from its combinations by heating with stronger acids, its separation from most other electro- negative bodies is attended with but little difficulty. The CARBONIC ACID. 481 carbonic acid is determined in one portion of the compound, and the other acids in another portion. If the substance under examination contain fluorides, the carbonic acid is expelled by one of the weak non-volatile acids, such as tartaric or citric acid ; since, if sulphuric or hydrochloric acid were employed, hydrofluoric acid would, at the same time, be discharged. When there occurs in the course of an analysis, a mixed pre- cipitate of fluoride of calcium and carbonate of lime, Fresenius directs the separation of the two substances to be effected by pouring acetic acid over the mixed precipitate, adding alcohol, filtering, and finally washing the residue with alcohol. Separation of Carbon from Nitre and Sulphur : Analysis of Gunpowder. The sample intended for analysis (from 20 to 30 grains) is finely pulverized and dried, either in the water bath, or in vacuo over sulphuric acid, in the manner shown in Fig. 59, p. 247. It is then digested for some time with boiling distilled water, and the insoluble portion thrown on a filter, the weight of which is known. It is repeatedly washed while on the filter with hot water. The filtrate and the washings are evaporated to dryness, and the dry mass estimated as nitre ; or, after having weighed the residue, it may be redissolved in water, and tested for common salt by nitrate of silver, and, if any notable quan- tity is thereby indicated, the silver salt must be added as long as any precipitation takes place, and the resulting chloride of silver calculated into its equivalent of chloride of sodium, and the amount deducted from the dry salt. The substance re- maining on the filter is dried and weighed ; it consists of car- bon and sulphur. To determine the proportion of these two elements, a fresh quantity, (from 15 to 20 grains,) of the finely powdered sample, is intimately mixed with an equal amount of pure anhydrous carbonate of soda ; the mass is then mixed with about eight parts of nitre, and six of common salt. The whole is strongly heated in a platinum crucible : the sulphur becomes hereby oxidized into sulphuric acid at the expense of the oxygen of the nitric acid, and the acid thus formed unites with the potassa of the decomposed nitre, forming sulphate of potassa. As soon as the mass has become white, it is dissolved in water, acidified with hydrochloric acid, and the sulphuric acid precipitated by chloride of barium ; from, the weight of sul- phate of baryta obtained, the amount of sulphur is calculated. Marchand* modifies the above (Gay-Lussac's) method of esti- "mating the sulphur in gunpowder in the following manner : A 482 QUANTITATIVE ANALYSIS. mixture of 1 part of nitrate and 3 parts of carbonate of baryta is intimately mixed with a twelfth part of the powder, and heated in a tube closed at one end. A layer from 3 to 4 inches in length of baryta salt is inserted in front of the mixture, and the whole is heated in the combustion furnace, beginning at the anterior portion. The mixture, which does not fuse, is easily removed from the tube, which is rinsed with dilute hydrochloric acid, in which the ignited mass is then dissolved. The liquid is retained in a beaker for several hours near 212?, and the sul- phate of baryta is then collected on a filter. The amount of sulphur may likewise be determined accurately in the moist way. From 30 to 40 grains of the powder are treated in a flask with concentrated nitric acid, with the addition of a few grains of chlorate of potassa ; the mass is kept gently boiling, until at last, a colourless liquid is obtained, which is diluted with a large quantity of water, and precipitated while hot with chloride of barium. M. Cloez and Gruignet employ as an oxidizing agent, pure crystallized permanganate of potassa. About one gramme of the well-dried powder is boiled with a saturated solution of the permanganate, fresh portions being added from time to time, until the mixture has a persistent violet colour : concentrated hydrochloric acid is then added, and the whole boiled until the oxide of manganese is completely dissolved ; the solution is filtered, concentrated (if necessary), a little nitric acid added, and the sulphuric acid precipitated by chloride of barium. The charcoal is estimated indirectly by deducting the weights of the moisture, nitre, and sulphur, from the original weight of the powder analysed. The amount of nitre in gunpowder may, according to Mar- chand, be ascertained with great precision, by determining the quantity of nitrogen in the powder. For this purpose a por- tion of the sample is weighed off, reduced to a fine powder, and intimately mixed with oxide of copper ; the mixture is trans- ferred to a combustion tube, the sealed end of which is charged with carbonate of lead, and then with oxide of copper; oxide of copper and metallic copper are placed in front. The details of the process will be given hereafter. Uchatius employs the following ingenious process for deter- mining the amount of nitre. 20 grammes of the powder are intro- duced into a flaskwithabout 50 grains of lead shot; 200 grammes of well water are then added by means of a graduated pipette ; CARBONIC ACID. 483 the flask is well closed and shaken for eight minutes, by which the solution of the nitrate of potash in the powder is completely effected. The solution is filtered, and 172 grammes of the filtrate measured off in a second pipette, and placed in a narrow beaker ; it is brought to the normal temperature, and a glass bubble is introduced. This bubble is so constructed that when the powder contains 75 per cent. (= 15 grammes of nitrate of potassa) it will rise to the surface, while the addition of three or four drops of water will cause it to sink to the bottom. By means of a gra- duated pipette, fixed quantities of one of two solutions is added, so as to cause an increase or diminution in the quantity of nitrate of potash contained in the solution, equal to one or more tenths percent, until the bubble attains the surface of the solution. The percentage quantity of nitre in the gunpowder is now ascertained by adding to, or deducting from, 75, the amount of diminution or increase in the strength of the solu- tion effected by the addition of the test fluid. Separation of the Sulphur from the Charcoal. For this pur- pose Marchand adopts Wohler's method, which is as follows : A weighed quantity (about 20 grains of the dried mixture of sulphur and charcoal) is introduced into a bulb blown out about 2 inches from the extremity of a tube of hard glass, about 8 inches in length. An asbestos plug is thrust into the longer part of the tube to within half an inch or so of the bulb, and the remainder of the tube is then filled with metallic copper ; the tube is then weighed, a current of dried carbonic acid gas. is passed through the tube, and the part containing the copper having been raised to a full red-heat, the sulphur is distilled through it ; the sulphur vapours combine with, and are retained by, the copper ; when the operation is over, the tube is allowed to cool, it is cut off between the asbestos plug and the bulb, and the latter containing the carbon is rapidly weighed ; the loss of weight expresses the amount of sulphur. The heat which must be applied to remove the whole of the sulphur from the char- coal is, in general, greater than that at which the charcoal had been prepared, and there is observed during the expul- sion of the sulphur a strong odour of humic acid ; carbonic acid, carbonic oxide, and water are also expelled. Instead of distilling the sulphur from the charcoal, it may be removed by solvents. Marchand employs for this purpose sulphide of carbon. After exhausting the gunpowder with water, which he effects in a displacement apparatus of the form of an ordinary chloride of 481 QUANTITATIVE ANALYSIS, calcium tube, from 7 to 8 inches long, the bulb being stopped with asbestos, he treats it with absolute alcohol, which displaces the water, upon which sulphide of carbon, which has been recti- fied over oxide oflead,is poured over it until what passes through, leaves no sulphur on evaporation; the powder is then finally treated with alcohol. It is still better to use the sulphide of carbon mixed with absolute alcohol. As soon as the charcoal is washed, a current of dry air is drawn th rough the tube by means of an aspirator, the tube itself being confined in an air bath at 248. The dry charcoal is weighed accurately in the tube. The sand, etc., contained in the powder may be determined with tolerable accuracy by trituration and suspension. 197. OXALIC ACID. The quantitative estimation of this acid may be effected by precipitating it from its solutions in the form of oxalate of lime, and determining the latter as carbonate or sulphate ; or by as- certaining the amount of carbonic acid yielded by its decom- position. Quantitative Estimation of Oxalic Acid by precipitation. The acid, if in a free state, is neutralized as exactly as possible with ammonia, diluted with water, and chloride of calcium mixed with acetate of soda added ; the precipitated oxalate is well washed, dried, and ignited, with the precautions prescribed .page 279, by which it is converted into carbonate ; and the quan- tity of oxalic acid is calculated from the weight of the latter, every equivalent of oxalic acid yielding two of carbonic acid, thus : C 2 03 + = 2C0 2 . There are two ways of estimating the amount of carbonic acid yielded by the decomposition of oxalic acid ; one is by burning it with oxide of copper, according to the ordinary me- thod of organic analysis ; and the other, which is by far the most convenient and expeditious, is by neutralizing it with am- inonia, and bringing it into contact with peroxide of manga- rese and sulphuric acid, in the manner fully described, under the head of MANGANESE, page 338. 9 parts of oxalic acid require theoretically 11 parts of manganese, but an excess of the latter does no harm. Both of these latter methods are applicable to the determination of oxalic acid in the oxalates. When the oxalate is soluble in water, the amount of oxalic acid may be CHLORINE. 485 determined by precipitation, as carbonate of lime, and the base or bases with which it was combined, determined in the filtrate by appropriate methods. Many of the insoluble oxalates may be analysed by simple ignition, the bases being left either as pure metals, or as oxides, or as carbonates. Oxalates, the bases of which are precipitated by carbonate of potassa, may be decom- posed by boiling for some time with that alkali ; oxalic acid is thus brought into a soluble state, in the form of oxalate of potassa, and may be precipitated by a neutral solution of lime. Oxalic acid, as met with in commerce, frequently contains tartaric acid and sulphate of potassa. The former is detected by digesting a portion of the acid in a test tube with concentrated sulphuric acid ; it should remain colourless ; but if tartaric acid be present, it turns black, that acid being decomposed. The presence of sulphate of potassa is indicated by a precipitate in- soluble in nitric acid, produced on adding to a solution of the suspected acid a few drops of chloride of barium. CHLOEINE. 198. HYDROCHLORIC ACID. The quantitative determination of this acid is generally ef- fected by precipitating its chlorine in the form of chloride of silver ; for the composition of this salt, and for the precautions to be observed in its precipitation, the student is referred to section 174. If hydrochloric acid alone be present in aqueous solution, it may be estimated in the form of sal-ammoniac, by adding ammonia in excess, evaporating the solution to dryness, and performing the exsiccation of the residue on the water- bath. Soluble chlorides are analysed by first precipitating the chlorine by nitrate of silver, and then, after removing the ex- cess of the silver salt from the filtrate by hydrochloric acid, esti- mating the amount of base or bases by the ordinary processes. The analysis of the compounds of chlorine with tin and anti- mony, is performed in a somewhat different manner. If a solu- tion of nitrate of silver were added to a solution of perchloride of tin, there would be precipitated, besides chloride of silver, a compound of peroxide of tin with oxide of silver, which would vitiate the result. It is necessary, therefore, to remove the base from the solution before attempting to precipitate the electro-negative element; this is done by passing through the PART II. 2 K 486 QUANTITATIVE ANALYSIS. solution, contained in a flask which can be corked, a stream of sulphuretted hydrogen, and, when it is saturated, allowing it to repose for a considerable time ; the sulphide of tin is filtered off as soon as it has completely subsided, and the excess of sul- phuretted hydrogen removed from the filtrate by mixing it with a solution of sulphate of copper; this is indispensable ; other- wise, on the addition of nitrate of silver, a precipitate both of chloride and of sulphide of silver would take place. The ex- cess of sulphuretted hydrogen being removed, the solution is precipitated by .nitrate of silver in the usual manner. In the case of chloride of antimony, it is unnecessary to remove the base previous to determining the chlorine ; but the precaution must be taken of mixing the solution with a sufficient quantity of tartaric acid to prevent the formation and precipitation of a basic salt. Such chlorides as are insoluble, or only sparingly soluble in water, but soluble in nitric acid, are dissolved in the latter, the solution diluted with water, and precipitated by nitrate of silver in the usual manner. Analysis of Chlorides insoluble in Water and in Nitric Acid. This maybe performed in three ways: 1st, by reducing them to the metallic state by ignition in a stream of dry hydro- gen gas: 2nd, by fusion with alkalies; and 3rd, by digestion with caustic potassa. (1.) By reduction. The apparatus, Fig. 71, p. 303, may be employed. The chloride to be analysed (chloride of lead for instance) *s accurately weighed in the little bulb d, a stream of dry hydrogen gas from the bottle a is conducted over it, and the bulb is then heated to redness by means of a spirit lamp, the heat required being somewhat greater than in the reduction of oxides. When hydrochloric acid gas ceases to be disengaged, which is known by fumes ceasing to be produced when- a feather or glass rod moistened with ammonia is held near the aperture of the tube, the bulb is allowed to cool and again weighed ; the loss of course indicates the amount of chlorine. (2.) By fusion with Alkalies. The chloride to be analysed (chloride of silver for instance) is mixed in a porcelain cruci- ble with three parts of a mixture of carbonates of potassa and soda, and heated till the mass enters into fusion and efferves- cence ceases ; it is then allowed to cool, and the mass treated with water. The silver remains undissolved in a state of very fine division. It is filtered, washed, ignited, and weighed. The CHLORINE. 487 chlorine in the filtrate, representing that in the compound ana- lysed, is determined as chloride of silver. Gmelin recommends the following process for decomposing chloride of silver: a cru- cible is filled almost to the top with an intimate mixture of three parts of the chloride and seven of colophony ; then ex- posed to a gentle heat, when the resin burns with a green flame, owing to the hydrochloric acid generated from the chlo- rine of the chloride arid the hydrogen of the resin ; a stronger heat is then given to fuse the silver, a little borax is added, and the crucible tapped gently once or twice to facilitate the union of the silver. This process was first proposed by Mohr. (3.) By digestion with Caustic Potassa. This method is par- ticularly applicable to the analysis of subcliloride of mercury. The solution filtered from the metallic suboxide is acidulated with nitric acid, and precipitated with nitrate of silver. The suboxide of mercury may be dissolved in aqua-regia, and deter- mined by one of the methods described in section 175 ; or a fresh portion of the subchloride may be taken and treated with hydrochloric acid and protochloride of tin, as directed p. 369. Chloride of silver may be decomposed by digestion with caustic potassa ; a useful method of preparing pure oxide of silver by this decomposition was proposed by Dr. Gregory (Proc. Chem. Soc. vol. i.). The moist chloride is mixed with a suffi- cient quantity of caustic potassa, sp. gr. 1*25, and boiled for about ten minutes, or until the chloride has become converted into a jet-black powder. If any white specks be observed, the mixture must be rubbed in a mortar, and again boiled. When the decomposition appears complete, the oxide is to be carefully washed by decantation with hot water until all the saline matter has been removed; it is thus obtained as a heavy black powder. It is worthy of remark, that if the chloride of silver has once been dried, it is only decomposed by boiling for a long time with caustic potassa, and then with great difficulty. According to Dr. Meurer, the above process does not succeed when the experiment is made on several ounces. Levol modifies the process thus : he places the chloride in a solution of caustic potassa, in which some sugar is dissolved, and then boils. The silver is quickly reduced by the sugar, carbonic acid being evolved, and is brought to the state of fine powder, which is easily washed. The amount of chlorine in many metallic chlorides may be determined indirectly by converting them into sulphates, for 2 K2 488 QUANTITATIVE ANALYSIS. which purpose the compound is treated with concentrated sul- phuric acid, and heated till all the hydrochloric acid gas and the excess of sulphuric acid are expelled ; the amount of chlorine is inferred from the loss of weight. Volumetric determination of Chlorine in its combinations. As a solution of chloride of sodium of known strength may be used to estimate the amount of silver in a fluid, so by adding to a liquid containing hydrochloric acid, or a combi- nation of chlorine with a metal, a solution of silver of known strength, the amount of hydrochloric acid or chlorine may be ascertained. In order to indicate with greater precision the exact point of complete precipitation, Levol employed phosphate of soda, which in a neutral liquid indicates the presence of an excess of silver by the production of a yellow precipitate. Mohr (Liebig's 'Annalen,' xcvii.) found that by this means the results are always too high, in consequence of the very light colour of the phosphate of silver, which re- quires to be present in considerable quantity to become visible in the presence of chloride of silver. Arsenate of soda gave better results, but with chromate of potassa results of very great accuracy are attainable. If a drop of silver solution be pre- sent above the quantity of the chloride, the blood-red colour of chromate of silver makes its appearance distinctly. If O2 cubic centimetre too much of the silver solution be added, the mixture is distinctly red. The solution must not be acid, as then chromate of silver is not formed, or only in very small quantities ; moreover, bichromate of potassa has a rather red colour. A slight excess of pure carbonate of potassa, on the contrary, is not injurious, because then only the pale yellow colour of chromate of potassa can be exhibited. The standard solution of silver is made by dissolving 108 grains of the pure metal in pure nitric acid, evaporating the solution to perfect dryness on the water-bath, and redissolving the residue in 10,000 grain measures of distilled water. The standard solu- tion of chloride of sodium is made by dissolving 58'5 grains of the pure salt in 10,000 of distilled water. If both solutions have been properly prepared, they will correspond with each other, drop for drop ; this, however, must be tested by mixing together equal volumes of the two solutions, shaking the mix- ture, applying heat, and then allowing the precipitate to sub- side, the supernatant liquid is decanted, and to one portion of the clear liquid a drop of nitrate of silver is added, and to an- CHLORINE. 489 other a drop of solution of chloride of sodium. Both solutions should remain clear ; should however a cloudiness be produced in the portion to which chloride of sodium has been added, it indicates that a small loss has occurred in the evaporation of the silver solution, and its correct strength must be ascertained by repeated experiments ; the proportion found forms the basis of the calculations in the determinations of chlorine by the so- lution. Mohr found by repeated experiments that the excess of the silver solution required to produce a distinct red colour was exactly -^ cubic centimetre, this quantity must therefore be subtracted from the quantity of the silver solution employed. Alkaline chlorides may be determined with great accuracy by this method, which may be employed with concordant results with urine, ivell water -, saltpetre, potashes, etc. Quantitative estimation of free Chlorine. I. Since chlorine gas is absorbed both by water and mercury, it cannot be estima- ted by the ordinary methods of measuring gases without very great difficulty. The determination of the gas may be effected by leading it into ammonia, a portion of which it decomposes, chloride of ammonium being formed, and nitrogen gas set free. The operation is conducted as follows : the substance from which the gas is to be liberated, is introduced into the flask A, Fig. 83. Fig. 83. This flask is provided with a well-fitting cork, through which two tubes are inserted, one of which, b, communicates 490 QUANTITATIVE ANALYSIS. with a Wolfe's bottle, B, containing a sufficient quantity of ammonia to absorb all the chlorine that can be evolved ; the other tube, , is a safety-tube, and is surmounted by a funnel, through which any required liquid may be poured into the flask. The Wolfe's bottle, B, is connected with a second bottle, c, also containing ammonia, further to prevent the possibility of any chlorine escaping absorption. This second bottle is pro- vided with a long tube, surmounted by a funnel, in which, du- ring the operation, there is placed some cotton moistened with ammonia. Heat is applied to the flask A, and when the disen- gagement of chlorine has ceased, all the gas with which the flask is filled, as well as that in the connecting tubes, is expelled, and conducted into the ammonia, by pouring through the funnel of the safety-tube (which during the experiment is filled with a strong solution of common salt) a concentrated solution of bicarbonate of potassa until it is filled. The extremity of the tubes leading from the flask to the first Wolfe's bottle, and from the first Wolfe's bottle to the second, must not quite reach the surface of the ammonia, to avoid the danger of a sudden absorption, and the consequent rushing of the ammonia into the evolution flask. The operation being ended, the ammonia and the moistened cotton are transferred to a be.'iker, acidulated with nitric acid, and the amount of chlorine condensed, deter- mined by nitrate of silver. The disengagement of the chlo- rine throughout the experiment must be very slow, other\Tise a complete decomposition of the ammonia might not take place, and a portion of chlorine may be liberated in company with the nitrogen. If the chlorine be in aqueous solution, it may be de- termined by the same decomposition, for which purpose it is only necessary to add excess of ammonia, and then to proceed with the solution in the usual manner. According to Buchner, the amount of chlorine in chlorine water may be determined by agitating it with a weighed amount of metallic mercury, taking care to use such an excess, that a portion of the fluid metal separates in the free state, in order that the whole of the chlorine shall be in the form of the insoluble subchloride ; the quantity of chlorine is found by collecting, drying, and weighing the calomel and unaltered mercury, and comparing the weight with that of the metal when introduced into the solution. The same method may be employed to test a solution of chlorine for hydrochloric acid, the chlorine being so entirely removed that the presence of hydrochloric acid may be ascertained by CHLORIMETRY. 491 the acid reaction of the liquid, and its amount determined in the usual manner. IT. Chlorine may also be determined by the use of what is called the residual method. The liquid containing it is mixed with a solution containing excess of iodide of potassium, and as soon as the chlorine has been completely absorbed, hyposulphite of soda is added from the burette until the browncolourarising from free iodine has entirely disappeared. Solution of starch paste is then added, and the analysis completed by the addition of the solution of iodine until the blue colour appears. If, then, we note the amount of iodine requisite to produce this reaction, by calculating the equivalent quantity of hyposulphite of soda, and subtracting it from the total amount originally employed, we find by the difference, the quantity of hyposulphite of soda required to saturate the amount of chlorine in the solution. (Scott, 'Handbook of Volumetric Analysis.') (See IODINE.) in. Pisani's method ('Comptes Kendus,' Feb. 16th, 1857). After the chloride to be analysed has been slightly acidified by pure nitric acid, a known quantity of silver is added to it from a normal solution of nitrate of silver : this should be in slight excess. It is then filtered to separate the chloride of silver, which is carefully washed, the excess of silver is determined in the filtrate, and the difference gives that which is combined with the chlorine, and consequently the chlorine itself. The determination of the silver is effected by means of iodide of starch, which allows the appreciation of a very minute excess of silver employed. The standard solution of iodide of starch is made by preparing a solution of it in water, and ascertaining its strength by its action on a solution of neutral nitrate of silver ; upon adding the iodide, the fluid acquires a yellow colour owing to the formation of iodide of silver; the moment the reaction is complete, the fluid becomes bluish-green. A small quantity of pure precipitated carbonate of lime previously added, not only neutralizes any free acid, but enables the operator to hit off with greater exactness the precise time when the reaction is complete. CHLORIMETRY, OB, DETERMINATION OF THE VALUE OF CHLORIDE OF LIME. (1.) Modification of the method of Gay-Lussac. This is founded on the conversion of calomel into chloride of mercury by the chlorine set free from the bleaching-powder. A fluid is 492 QUANTITATIVE ANALYSIS. prepared, holding in suspension a known quantity of subchloride of mercury, into this, a measured quantity of the solution of chloride of lime to be tested is poured, until the liquid becomes perfectly clear; the quantity of chlorine consumed, and conse- quently the value of the bleaching powder, is calculated from the number of divisions of the burette which have been required to render the turbid mixture clear. Preparation of the standard Mercurial Test Liquor. It is based on the following considerations: Every 236 '5 grains of subchloride of mercury require 35 '5 grains of chlorine to convert them into chloride ; every 58'5 grains of common salt (which is the chloride most convenient for the purpose) contain 35'5 grains of chlorine, or 164'7 grains contain 100 grains of chlorine. Let 50 measures of a solution of a subsalt of mercury, the subnitrate, for example, be introduced into the graduated burette, and then poured off into a beaker, and diluted with 4 or 5 ounces of water, then let 82'4 grains of common salfe be dissolved in water and introduced into the burette, so as to exactly fill 100 measures. Let this solution be poured gra- dually into the solution of the subuitrate, (which during the time should be kept warm by immersion in a basin of hot water.) until the last drop added no longer produces a precipitate. Let us suppose that 20 measures have been required ; now, as 100 measures of the saline solution contain 82'4 grains of chloride of sodium = 50 grains of chlorine, the 20 measures must con- tain 16*48 grains = 10 of chlorine. We consequently arrive at the conclusion that 50 measures of the subnitrate of mercury require 10 grains of chlorine to convert the salt into subchloride, and if the 50 measures are increased to 100 by dilution with water a solution is obtained, every 10 divisions of which are equivalent to 1 grain of chlorine, and every single division of which represents y^th of a grain ; any quantity of the test liquor may thus be easily prepared. Method of testing the Bleaching Powder. 100 grains of the specimen, selected from different parts of the sample, and well mixed together, are rubbed in a mortar with a little water, more water is gradually added, the mixture stirred, and the heavier particles allowed to subside; the milky fluid is then poured off, and the precipitate again agitated with water, allowed to stand, and the supernatant fluid poured off, mixed with the first, and transferred to a tube graduated into 200 equal parts, each part CHLOR1METRY. 493 corresponding with a division of the burette; the mortar is rinsed out with a little more water, the washings transferred to the graduated tube, the whole well agitated, and the mark at which the solution stands in the tube is accurately noted. Suppose the whole to occupy 150 measures. The burette is now filled to 100 with the mixture. Next. 100 measures of the standard subnitrate of mercury are poured into a beaker and diluted with water, excess of common salt is added to the solu- tion, which is then acidified with pure hydrochloric acid. The well-agitated solution of chloride of lime is then dropped gradu- ally into the newly-formed calomel in the beaker, until it is perfectly clear, by which it is known that the whole of the calo- mel has been converted into corrosive sublimate. It must be constantly stirred while the bleaching solution is being added, and it is essential that it should remain acid during the whole time. Now, from what has been explained above, it is evident that the number of divisions of the bleaching solution that have been consumed in rendering the turbid mercurial salt clear, must correspond with 10 grains of chlorine. Suppose 65 measures were required, then the value of the specimen is found very simply by the following proportion : As 65 : 10 : : 150 : x, x 23-07 = the quantity of chlorine in 100 grains of the bleaching powder. It is evident that any other proportions than those here given may be employed, both in making the standard mercurial solu- tion, and in performing the analytical process. The above must be regarded as merely an illustration of the principles of the operation. This method is found to yield very accurate results ; there is perhaps one disadvantage attending it, which is, that the solution of subnitrate of mercury is somewhat liable to change, and to give rise to the formation of an insoluble basic salt, which would, of course, interfere with its accuracy. (2.) Modification of Gay-Lussacs method by Arsenious Acid. When arsenious acid, water, and chlorine are brought into con- tact with each other, a reaction takes place, the result being the formation of hydrochloric and arsenic acids, thus Every equivalent (or 99 parts) of arsenious acid require two equivalents (or 71 parts) of chlorine ; consequently 100 parts of chlorine are required to convert 139*45 parts of arsenious acid into arsenic acid. A standard solution of arsenious acid 494 QUANTITATIVE ANALYSIS. is prepared by dissolving 139'45 grains of the pure acid in a solution of potassa, diluting the solution, and then adding hydrochloric acid in considerable excess; the volume of the solution is then brought to 10,000 grain measures, and the whole well mixed ; every 100 grain measures of the solution thus prepared, contain 1*395 grains of arsenious acid = l grain of chlorine. When the chloriraetric assay is about to be made, 50 grains of the bleaching powder are prepared in the manner already described, and the burette is filled to the 100th division with the solution ; 1000 grain measures of the arsenious acid are then placed in a beaker, diluted with water, and coloured blue by the addition of a few drops of solution of indigo. The bleaching liquid is then gradually and carefully dropped into the arsenical solution, which is all the time well stirred until the blue tinge is entirely destroyed. This shows that the conversion of the arsenious acid into arsenic acid is com- plete, and consequently that 10 grains of chlorine have been consumed, and as before, the value of the bleaching powder is ascertained by the rule of simple proportion : thus, supposing 90 measures have been consumed, then As 90: 10:: 100 : #, ,r= 11*11 = the quantity of chlorine in 50 grains of the bleach- ing powder, or 22*22 per cent.; or still more simply, as each mea- sure of the burette corresponds with half a grain of bleaching powder, and as the number of measures consumed contain 10 grains of chlorine, we find at once the percentage amount of chlorine, by dividing 2000 by the number of measures required ; thus : 9nnn ?^= 22*22, 90 (3.) Otto's method by Protosulphate of Iron. When proto- sulphate of iron, water, chlorine, and free sulphuric acid are brought into contact, a reaction takes place, resulting in the formation of hydrochloric acid and persulphate of iron : thus : every two equivalents, or 152 parts of anhydrous, or 278 parts of crystallized protosulphate of iron, require one equivalent, or 35*5 parts of chlorine ; 100 grains of chlorine are therefore required to convert 783*1 grains of crystallized protosulphate of iron into persulphate. To prepare the iron salt for these experiments, Fresenius directs that iron nails free from rust be dissolved in dilute sulphuric acid, finally with the ap plica- CHLOR1METRY. 495 tion of heat ; the solution is filtered while still warm, into about twice its volume of spirits of wine ; the precipitate which is produced consists of protosulphate of iron with seven equivalents of water. It is collected on a filter, edulcorated with spirits of wine, spread upon a sheet of blotting-paper, and allowed to dry in the air until it has completely lost the smell of spirits of wine : the dry salt must be kept in a well-closed bottle. To perform the chlorimetric assay, 7S - 3 grains of the protosulphate are dissolved in about two ounces of water, and the solution strongly acidulated with sulphuric acid ; 50 grains of the bleachiug-powder are then prepared as before directed, and the burette filled with the solution, which, after being well agitated, is added drop by drop to the solution of the iron salt until complete peroxidation has taken place ; this is known by solution of red prussiate of potassa no longer striking a blue precipitate. The best way of testing the solution from time to time, is to cover a large white plate with drops of the red precipitate, and, as the operation draws towards an end, to con- vey a minute drop from the solution, at the end of a small stirring-rod, to one of the drops ; the process is complete when prussian blue ceases to be formed, a green colour being pro- duced in its stead. The number of divisions of the burette which have been consumed is now noted ; this corresponds with 10 grains of chlorine, that being the quantity required for the peroxidation of 78'3 grains of crystallized protosulphate of iron. The calculation is precisely the same as in the last method. (4.) By Yellow Prussiate of Potasli. When ferrocyanide of potassium is brought into contact with free chlorine, a reaction takes place, resulting in the production of ferricyanide of po- tassium and chloride of potassium ; thus : _ 2 (K 2 Cfy ) + Cl = K 3 Cfy 2 + K Cl. Every two equivalents, or 368 parts of anhydrous, or 422 parts of crystallized ferrocyanide of potassium require one equiva- lent, or 35'5 parts of chlorine; 100 grains of chlorine, there- fore, are required to convert IISS'7 grains of crystallized yellow prussiate of potassa into red prussiate, or 118'87 grains corre- spond to 10 grains of chlorine. The experiment is conducted in precisely the same manner as before, and as soon as the whole of the ferrocyanide is converted into ferricyanide, the mixture ceases to give prussian blue with a solution of per- oxide of iron. 406 QUANTITATIVE ANALYSIS. IODINE. 199. HYDRIODIC ACID. I. Estimation as Iodide of Silver. Iodine when present in a solution in the form of hydriodic acid, and in the absence of bromine and chlorine, is best precipitated in the form of iodide of silver, by solution of nitrate of silver. This salt, which when newly precipitated is of a pale yellow colour, has many proper- ties in common with the corresponding chlorine and bromine compounds. Thus, it is insoluble in water and in dilute nitric acid, and soluble, though with difficulty, in ammonia.* Like bromide of silver, it is decomposed by chlorine ; it also blackens when exposed to the light, and, when heated, it fuses into a horny mass, without undergoing decomposition. Its composi- tion is One equivalent of I ... 127 .. 54'04 One ditto ofAg . . 108 . . 45-96 One ditto of Agl . . 235 . . lOO'OO II. Estimation as Protiodide of Palladium. In the pre- sence of hydrobromic and hydrochloric acids, iodine may be very accurately estimated in the form of protiodide of palladium, by means of protochloride of palladium. The precipitate, which is of a deep brown-black colour, is allowed to settle for some hours ; it is then collected on a tared filter, dried at a tempera- ture below the boiling-point of water, or, what is still better, MJ vacuo, over sulphuric acid, and weighed. Its composition is One equivalent of I . . 126'36 . . 70 34 One ditto of Pd . . 53-27 . . 29-66 One ditto ofPdl. . 179'63 . . 100 '00 in. Estimation as Protiodide of Copper. Iodide may also be estimated by the process recommended by M. Duflos, t viz. by precipitating it in the form of protiodide of copper !. by a solution of sulphate of copper in concentrated sulphu- y rous acid ; the precipitate may either be collected on a filter, ? washed, and determined as protiodide of copper, or it may [ be dissolved in dilute nitric acid, precipitated by nitrate of * According to Wallace and Lament (Cliem. Gaz., vol. xvii. p. 140) 1 part of iodide of silver is soluble in 2793 parts (measure) of liquid ammonia, sp. gr. '890, and 1 part of chloride of silver in 12*38 parts of the same solution. IODINE. 497 silver, and weighed as iodide of silver. This method is well adapted to the separation of iodine from chlorine, in cases where the presence of chlorine is immaterial. M. Sarphati (Ann. der Chem. und Pharrn., xxxix. p. 254) recommends a solution of chloride of copper in dilute hydrochloric acid ; or a solution composed of 1 part of crystallized sulphate of copper, and 1J of crystallized sulphate of iron. iv. Pisani's method. Iodine, in alkaline iodides, may also be estimated by the following method, recommended by Pisani (' Comptes Eendus,' February 16, 1857): A few drops of so- luble iodide of starch are added to the liquid containing the iodide to be analysed : the quantity must be sufficient to give it a distinct blue tint. A normal solution of silver is then dropped in until the iodide of starch is decolorized, which only takes place after the complete precipitation of all the iodine contained in the alkaline iodide. The number of divisions em- ployed is then read off, and the corresponding quantity of iodine calculated. y. Indirect method. Iodine is sometimes determined by the indirect method, viz. by converting the base or bases w r ith which it is combined into sulphates, by digestion with sulphuric acid in a porcelain capsule, and calculating the quantity of iodine from the loss. Metallic iodides, which are insoluble in water, may be decomposed by ignition with carbonate of soda ; iodide of sodium is obtained on treating the fused mass with water, from which the iodine may subsequently be precipitated by nitrate of silver. The base or bases are found in the insoluble residue, either as carbonates or oxides. Metallic iodides may also be decomposed by boiling with caustic potassa. YI. ^Estimation of free Iodine. Iodine in a free state, in aqueous solution, cannot be determined in the same manner as chlorine and bromine, viz. by the decomposition of ammonia, since the contact of iodine with ammonia gives rise to the for- mation of iodide of nitrogen, which is not decomposed by excess of ammonia. In order to estimate the iodine under these cir- cumstances, it may be dissolved in excess of caustic potassa. the solution, which must be colourless, treated with nitric acid, and precipitated by nitrate of silver ; the washed and dried pre- cipitate is ignited, by which the small quantity of iodate of silver formed, is converted into iodide. Separation of Iodine from Chlorine. Of all the methods which have been proposed, none are so convenient or easy of 498 QUANTITATIVE ANALYSIS. execution as that of precipitating the iodine in the form of protiodide of palladium by protonitrate of palladium. The separation is complete, and the chlorine in the filtrate may be precipitated by nitrate of silver, after having removed the ex- cess of palladium by sulphuretted hydrogen, and afterwards destroying the excess of the latter by a solution of sulphate of sesquioxide of iron. It is better however to operate on two different portions of the solution, determining the iodine in one by protochloride of palladium, and precipitating the chlo- rine and iodine together in the other by nitrate of silver ; from the weight of the precipitate is subtracted that of the iodide of silver corresponding with the quantity of iodide of palladium found. Estimation of Iodine in Kelp. The following process has been employed by Wallace and Lamorit (Chem. Gaz., vol. xvii. p. 137). It is based upon two well-established facts, 1. that- when nitrate of silver is added to a solution containing iodine and chlorine, the iodine is thrown down first ; and 2. that; iodide of silver is only sparingly soluble in liquid ammonia, while chloride of silver is freely soluble in that liquid. 500 grains of the pulverized and well-mixed sample are boiled with water, and the liquor and washings are nearly but not quite neutralized with nitric acid, and evaporated to a small bulk. The residue is then fused by the application of a stronger heat, and kept in this condition for a few minutes, when all the sul- phur compounds are converted into sulphates without danger of decomposing the iodide. The fused mass is dissolved in water, the solution filtered (if necessary), and nitrate of silver added in small successive portions with brisk agitation, until the preci- pitate produced on a further addition is perfectly colourless ; nitric acid is now added in excess, and the mixture is well washed by decantation. The last washing is decanted as com- j. pletely as possible, half an ounce of the strongest liquid am- 1 monia is added, and the mixture vigorously shaken. The di-H gestion is continued for several hours, after which the iodide of silver is collected on a small tared filter, and washed first i.l with liquid ammonia, and subsequently with hot water, it is then f dried. The weight of the iodide of silver + *1 grain (dissolved-: by the ammonia) divided by 5 and multiplied by '54, gives the percentage amount of iodine in the sample. A method of detecting the presence of iodine in waters: has been proposed by S. de Luca (' Comptes llendus,' Aug. 1, IODINE. 499 1859). It is founded on the power possessed by bromine of decomposing iodides, and setting free iodine. The water is precipitated by nitrate of silver, the precipitate, washed and dried, is placed at the bottom of a glass tube, and on it is dropped a small glass bubble filled with vapour of bromine. The air in the tube is replaced by carbonic acid, and the tube is then hermetically sealed. By shaking the tube two or three : times the little bubble is broken, when the vapour of bromine comes in contact with the iodide, which it decomposes, setting free iodine in the form of violet fumes, which condense on the cold part of the tube, and may be dissolved out by alcohol. The amount of iodine may be estimated by a standard solution of sulphurous acid, and the hydriodic acid may afterwards be reconverted into iodide of silver, and weighed. Another method has been employed by MM. 0. Henry and Humbert (* Comptes Rendus,' March 23, 1857). The precipi- tate by nitrate of silver is washed and dried ; it is then mixed intimately with cyanide of silver, and introduced into a tube, at one end of which it is fixed between two plugs of asbestos. A current of very dry chlorine is then passed slowly over the mixture, whilst the tube is slightly heated. The iodine and cyanogen are displaced, combine, and condense in the colder parts of the tube, in the form of a white crystalline ring of iodide of cyanogen. To detect iodine in the presence of nitric acid Stein (Polyt. Centra Ibl., 1858) employs tin as a reducing agent and bisul- phide of carbon as a solvent for the iodine. A quantity of the acid to be tested is poured into a test tube, and a rod of tin im- mersed in it until red fumes are distinctly evolved. The rod of tin is then taken out, and a small quantity of sulphide of carbon added to the liquid, the mixture is shaken and left quiet for a few moments. The stratum of sulphide of carbon which usu- ally collects over the acid appears of a red colour, unless the amount of iodine be too small. With traces of iodine, the colour is only deep yellow. In this case however, it becomes red, when the sulphide of carbon is drawn off and evaporated in a small porcelain capsule by blowing upon it. It was discovered by Dupasquier that when iodine and sul- phurous acid are brought into contact, in the presence of water, sulphuric and hydriodic acids are formed thus : and on this he founded a volumetric process for the estimation 500 QUANTITATIVE ANALYSIS. of iodine. Bunsen found the results unsatisfactory in conse- quence of a reverse action which ensues when the sulphurous acid is present beyond a certain quantity thus : He ascertained, however, that the former reaction alone takes place if the quantity of sulphurous acid does not exceed from 0*04 to 05 per cent, of anhydrous acid, and that under these circumstances one equivalent of iodine converts one equivalent of sulphurous acid into sulphuric acid; on this he has based a volumetric method applicable to a great number of analytical operations depending on oxidation and reduction, the oxidizing substance being iodine. For details of his process we must refer to the author's original memoir (Annal. der Chem. und Pharm. 86, 205),* or to works specially devoted to volumetric analysis (Scott, Button). BROMINE. 200. HYDKOBHOMIC ACID. "When present in a solution in the form of hydrobromicncid, bromine is determined by precipitating it in the form of bro- mide of silver, precisely in the same manner as chlorine. Like the chloride, the bromide of silver is insoluble in water and in nitric acid : it is yellowish-white when precipitated, but it changes colour when exposed to the light. It is decomposed by chlorine, bromine being separated. When heated, it fuses, solidifying, on cooling, into a yellow horny mass. Its compo- sition is One equivalent of Br .... 80'0 . . 42-55 One ditto of Ag .... 108'Q . . 57'45 One ditto of AgBr . . . ISS'O . . ICHX) The analysis of bromides may be effected in the same man- ner as that of the corresponding chlorides. The indirect me- thod is, in many cases, the most convenient : a weighed quan- tity of the compound is decomposed by heating it in a porce- ' lain capsule, wdth concentrated sulphuric acid ; the amount of' metal is calculated from the weight of the resulting sulphate, and that of the bromine from the loss sustained. The analysis of those bromides, the bases of which are capable of being pre- cipitated by sulphuretted hydrogen, may be performed by that * Translated in full in the Journal of the Chemical Society, vol. viii. p. 219J BROMINE. 501 reagent. Free bromine, in aqueous solution, is determined in exactly the same manner as free chlorine. Separation of Bromine from Chlorine. Two methods of performing this analysis are known, a direct and an indirect method. i. Direct method. This is based on the solubility of bro- mide of barium in absolute alcohol, and the insolubility of chloride of barium in the same menstruum. The two substances are precipitated together out of their solution, by nitrate of sil- ver ; the precipitates, after having been washed and dried, are reduced by being brought into contact with zinc and dilute sulphuric acid, there are thus formed bromide and chloride of zinc, which are dissolved in the acid liquor ; on the addition of excess of baryta water, the zinc is precipitated together with sulphate of baryta in the form of hydrated oxide ; the filtered solution contains bromide and chloride of barium, it is eva- porated to dryness, and the dry mass treated with absolute alcohol, which, as above stated, dissolves only the bromide of barium. The bromine is separated from the barium in the alcoholic solution, and the chlorine from the barium in the so- lution obtained on treating the residual mass with water, by one of the methods already given. This method does not give absolutely accurate results. ii. Indirect method. This is based on the decomposition of bromide of silver by chlorine gas. The bromine and chlorine are, as before, precipitated together, by a solution of nitrate of silver, and the precipitate having been washed, dried, fused, and very accurately weighed, is introduced into a bulbed tube of hard glass, the exact weight of which has also been noted. The tube is connected with an apparatus for evolving chlorine, and a cur- rent of that gas dried by passing through chloride of calcium, is caused to traverse the tube, to which heat is at the same time , applied sufficient to keep the mixed silver salts in a state of fusion. After the operation has been continued for about half an hour, the tube is allowed to cool, and the chlorine in its in- terior having been replaced by atmospheric air, it is again weighed ; the process is again repeated, and the tube a second time weighed : should the second weight correspond with the first, the analysis is complete ; if not, the process must be again repeated, till the two last weighings are absolutely alike. The amount of bromine originally present is calculated from PAllT II. 2 L 502 QUANTITATIVE ANALYSIS. the diminution in the weight of the mixture, the bromide of; silver being now replaced by chloride of silver. The calcula- i tion is made thus : Suppose the weight of the mixed chloride ! and bromide to have been 100 grains, and, after the experiment, ! suppose the weight to have decreased 5 grains. The equivalent of bromide of silver is 188'0, that of chloride of silver 143'5 ;,' then, as 44'5 (the difference between the respective equivalents | of bromide and chloride of silver) : 188'0 (the equivalent of 1 , bromide of silver) : : 5 (the difference in weight found in the ; experiment) : x. x = 2-112 = the quantity of bromide of silver* in the mixture analysed. Separation of Chlorine, Uromine, and Iodine. i. Pisani's method (' Comptes Eendus,' Feb. 16, 1857). The liquid is divided into two parts, to one of which a known; slight excess of silver is added, the precipitate is filtered, washed, and weighed, and the excess of silver in the nitrate is determined by iodide of starch. The difference gives thd quantity of silver combined with the chlorine, bromine, and iodine. From the other portion of the liquid the iodine is pre- cipitated by nitrate of palladium ; it is filtered and estimated, and the weight of the silver salt which it would form being deducted from the joint weights of chloride, iodide, and bro- mide of silver found, the difference gives the joint weights of the bromide and chloride of silver, from which the quantity of silver being known, the weights of the chlorine and bromine in the mixture are calculated. ii. Lyte and Field's method (Quart. Journ. Chem. Soc.. vol. x. p. 234). This is founded on the fact that chloride oi silver is decomposed by bromide of potassium, and both chlo- ride and bromide of silver by iodide of potassium. AVhen mixed solution of bromide of potassium and chloride of sodiuix is added gradually to a solution of nitrate of silver not in ex< cess, no trace of chloride of silver is precipitated as long as anj bromide remains in solution. If, to a similar solution, iodide anc bromide of potassium and chloride of sodium be added, iodide of silver and nitrate of potassa are formed, bromide of potas sium and chloride of sodium remaining undecomposed ; whei bromide of potassium is poured upon chloride of silver, ai entire decomposition ensues, bromide of silver and chloride o potassium being produced. "When iodide of potassium is adder to chloride of silver, iodide of silver and chloride of potassiun CYANOGEN. 503 are formed; and when iodide of potassium is added to bromide of silver there is a similar decomposition, the iodine replacing the bromine. When chloride of silver in excess is agitated with a solution of iodide of potassium, and warmed for some hours, no trace of iodine can be detected in the solution ; when however chloride of sodium is poured upon iodide of silver, no decomposition occurs, neither is there any action upon bromide of silver with the same salt : and w r hen bromide of potassium is added to iodide of silver, there is no alteration in the union of the elements. Hence the following simple method of ana- lysing a mixture containing bromides, iodides, and chlorides. The liquid is divided into three portions. To the first is added slight excess of nitrate of silver, and the precipitate with the requisite analytical precautions, is weighed. To the second is added nitrate of silver as before, the precipitate is filtered off, w r ashed, returned to the flask and digested in a weak solution of bromide of potassium. A similar operation is performed upon the third, the precipitate in this instance however being digested with a weak solution of iodide of potassium instead of bromide. In the first precipitate, we obtain the combined weights of the chloride, iodide, and bromide of silver ; in the second the combined weights of the bromide and iodide of silver, and in the third, that simply of the iodide of silver. Bromide of potassium having no action upon iodide of silver has con- verted the chloride into bromide, leaving the iodide untouched ; while iodide of potassium having the power to decompose both chloride and bromide, the whole is converted into the iodide. .The calculation is very simple : as the difference between the equivalents of bromide of silver and chloride of silver (44'5) is to the equivalent of chloride of silver (14S*6), so is the differ- ence in weight after digesting in bromide of potassium to the quantity of chloride of silver originally present ; and as the dif- ference of the equivalents of bromide and iodide of silver (47'1) is to the equivalent of bromide of silver (188), so is the excess in weight of the precipitate after digestion in iodide of potas- sium to the weight of bromide of silver originally present. Eresenius (' Quantitative Analysis,' Eng. Trans, p. 393) objects to this method, which he says may give rise to very considerable mistakes, on account of the solubility of iodide of silver in iodide of potassium solution, and of bromide of silver in bro- mide of potassium solution. Field has however shown (Chem. ' News, vol. ii. p. 324) that there is no force in this objection, 2 L 2 504 QUANTITATIVE ANALYSIS. provided the solutions of the solvents, iodide and bromide of potassium, be sufficiently weak ; and that if this be carefully attended to, the results are reliable. CYANOGEN. 201. HYDROCYANIC ACID. This acid, when in a free state, is best estimated by preci- pitating it in the form of cyanide of silver. It falls as a white curdy mass, which is insoluble in water, and in dilute nitric acid, but soluble in ammonia. Unlike the chloride, bromide, and iodide of silver, it does not change colour when exposed to light, neither will it bear ignition without being decom- posed. It may, however, be dried at 212, without alteration.] For its composition, see page 853. Soluble cyanides may also j be precipitated by nitrate of silver; insoluble cyanides, which; do not readily dissolve in dilute nitric acid, may be analysed by j determining the amount of carbon and nitrogen which they contain by the methods of organic elementary analysis ; or the j amount of cyanogen may be determined in the indirect man- ner, by igniting the compound, and weighing the residue. | According to Rose (PoggendorfPs 'Annalen') it is better to treat the compound with concentrated sulphuric acid and a! little water, and to heat the mixture long enough to get rid of I nearly all the excess of sulphuric acid. All cyanides, simple! and double, even cyanide of silver, as well as nitro-prusside of j sodium and the sulpho-cyanides, are transformed by this pro- cess into sulphates, which present no difficulty in analysis. Double cyanides containing metals whose cyanides are so- luble in nitric acid, those of nickel, copper, and zinc, for in- stance, may be analysed thus : Cover the solid cyanide with I a solution of nitrate of silver, and then with water, and a cer- tain quantity of nitric acid, taking care not to add too large j an excess, cyanide of silver being more soluble in nitric acid; than chloride. Weigh the cyanide of silver dried at 212 on; a weighed filter, or, still better, calcine it at a moderate tem-i perature so as to avoid melting the silver until it ceases to' lose weight. The cyanogen in mercuric cyanide cannot be precipitated in the form of cyanide of silver, a crystalline compound of nitrate with cyanide being formed, which is soluble in water. .For analysing this salt, E.ose dissolves it in twenty-five or thirty times its weight of water, and leaves the solution for thirty-six CYANOGEN. 505 hours in contact with an equal weight of cadmium filings in a well-stoppered bottle ; he then adds nitric acid and nitrate of silver, and obtains almost the theoretical quantity of cyanide of silver. A large number of simple and compound cyanides are easily transformed into cyanide of mercury by boiling them with ex- cess of oxide of mercury; the metal of the cyanide separates in the state of oxide, and the cyanogen combines with the mercury. Smith proposes to decompose and estimate hydro- cyanic acid by means of chloride of lime, a process which oc- cupies but a few seconds ; the products are nitrogen gas and carbonate of lime. In some cases he prefers the employment of chloride of soda to chloride of lime, on account of the solu- bility of the compounds that are formed. The same method is applicable to the analysis of the salts of cyanogen, and also to that of the ferrocyanides. The usual method of analysing the latter, when free from alkali, is to fuse them, in small portions at a time, at an intense heat, in a porcelain crucible, with nitre. The bases which re- main are separated from each other by the ordinary processes, and the cyanogen is estimated in a fresh portion by the direct determination of the carbon and nitrogen. Alkaline ferro- cyanides are decomposed by boiling sulphuric acid, nitric acid, or aqua-regia, and the evaporated mass treated as usual for the bases. The reason why it is inadmissible to decompose them by fusion with nitre is, that an explosion would certainly re- sult from the contact of the nitre with the fused salt. Those alkaline double cyanides, which resist decomposition by boil- ing concentrated acids, are kept for a long time ignited in the air, or in contact with oxide of copper ; and, should the ferro- cyanide under examination contain a volatile metal, the latter must be determined by dissolving a separate portion of the compound in caustic potassa, or in hydrochloric acid, and pre- cipitating the metal with sulphuretted hydrogen. A method of analysing compounds of cyanogen, founded on the action of iodine, has been proposed by M. V. Gerdy ('Comptes Rendus,' Jan. 22, 1842). It was found by this che- mist that in many of the soluble combinations of cyanides, the cyanogen may be substituted, atom for atom, by iodine ; so that, from the amount of solution of iodine employed and de- colorized before its action on starch becomes evident, it is pos- sible to calculate with the greatest ease, and in a few moments, 508 QUANTITATIVE ANALYSIS. the amount of the cyanogen contained in the liquid, and the quantity of metal with which it was combined. Gerdy has thus analysed cyanide of potassium with great accuracy. 'He states, also, that the method is equally applicable to the analysis of cyanide of mercury; of the double compound of cyanide of potassium and cyanide of silver, which contains eight atoms of cyanogen to one of silver ; of cyanide of silver, dissolved in ferrocyanide of potassium ; of double cyanide of potassium and copper ; of cyanide of gold dissolved in cyanide of potassium, etc. The analysis of cyanide of gold is, he says, not so easy of execution, according to this method, as is the case with the other compounds above mentioned, and requires some care. The reaction is here of a different description ; for, while eight atoms of cyanogen are required to dissolve one atom of gold, viz. four atoms combined w r ith the gold, and four with the po- tassium, four atoms of iodine alone suffice to effect the decom- position. This peculiarity he considers to be owing to the cyanide of gold not being acted on by the iodine, and being only precipitated entirely when the iodine has decomposed the whole of the cyanide of potassium. Determination of the amount of Hydrocyanic Acid in the medicinal Hydrocyanic Acid, Sitter Almond, and Cherry -Laurel\ waters (Liebig). This process is founded on the fact that[ oxide and chloride of silver are soluble in cyanide of potassium! up to a certain point, viz. until the double compound consisting of equal equivalents of cyanide of potassium and cyanide of* silver is formed, which compound is not decomposed by excess of potash. To the liquid containing the prussic acid, caustic potassa is added until it has a strong alkaline reaction, and then a few drops of chloride of sodium. A solution of nitrate of; silver of known strength is then added, drop by drop, from; a burette until a slight milkiness is perceptible. The quantity!- used is noted ; every equivalent of silver consumed exactly cor-; responds to two equivalents of prussic acid. The presence of formic or hydrochloric acid has no influence on the result. The, method may be employed advantageously to test the purity ofii commercial cyanide of potassium. (Liebig' s Annaleu, Jan. 1851.) Separation of Cyanogen from Chlorine, Bromine, and Iodine. The best method is to precipitate all three of the electro- negative elements with nitrate of silver, and, having carefully collected, washed and dried the precipitate in the water-bath, to determine the amount of carbon and nitrogen yielded by the FLUORINE. 507 mixture by the ordinary methods of organic analysis. Prom the numbers obtained, the amount of cyanide of silver is calcu- lated, and the proportion of iodine, bromine, and chlorine is inferred from the difference. Fresenius gives the following method of determining the relative proportions of hydrocyanic and hydrochloric acids, when together present in a solution. It is founded on the circum- stance, that percyanide of mercury is not decomposed by nitrate of silver. The solution is divided into two equal parts. One portion is precipitated with nitrate of silver, and the mixed Drecipitate of chloride and cyanide of silver weighed ; the other portion is mixed with oxide of mercury, and the mixture agitated until perfectly inodorous. The fluid is then filtered off, the nitrate subsequently precipitated with nitrate of silver, and the precipitate weighed. The difference between the re- spective weights of the two precipitates indicates the amount of cyanide of silver contained in the precipitate in the first portion. FLUORINE. 202. HYDROFLUORIC ACID. When this acid exists in aqueous solution, it is quantita- tively estimated by rendering the liquid alkaline by ammonia, 'and precipitating the hydrofluoric acid in the form of fluoride of calcium by the addition of chloride of calcium ; the preci- pitate, which is gelatinous, is washed on the filter with hot distilled water, and then with acetic acid, to remove any car- bonate of lime which may have formed during the process of filtration. The composition of the ignited salt is One equivalent of Ca . . . 20 . . 51'28 One ditto of Fl . . . 19 . . 48-72 One ditto of CaPl . . 39 . . lOO'OO The fluorine in soluble fluorides may be determined in the same manner, the bases being estimated in the filtrate according to the usual rules. To detect fluorine in water, Mene treats the solid residue of a large quantity of the water with an excess of concentrated sulphuric acid, and passes the gaseous pro- ducts into water slightly ammoniated, in which, if fluorine were present in the water, some gelatinous silica is obtained, from the decomposition of the fluoride of silicon which was disen- gaged from the residue. 508 QUANTITATIVE ANALYSIS. Insoluble fluorides may be analysed in two ways. (1.) Decomposition by Sulphuric Acid. The fluorine in this process is estimated indirectly. A weighed quantity of the fluoride, finely pulverized, is heated in a platinum capsule with concentrated sulphuric acid ; the fluorine escapes in the form of hydrofluoric acid gas, and the excess of sulphuric acid is finally expelled by raising the heat to ignition. From the weight of the resulting sulphate, the quantity of metal in the base is calculated ; that of the fluorine is indicated by the loss of weight. Should the substance analysed contain more than one base, the resulting sulphates must be submitted to further analysis, before a calculation can be made of the amount of fluorine present in the substance. (2.) Conversion of the Fluorine into Fluoride of Silicon, Wohler's process (Poggendorff's 'Annalen,' vol. xlviii.). The weighed substance is intimately mixed with pure silica, unless it already contains some ; the mixture is placed in a small flask which can be weighed on the balance ; very concentrated sul- phuric acid, which has been boiled, is added, and the flask quickly closed with a cork, through which a small tube, filled with fused chloride of calcium, and drawn out to a fine point, passes. The whole apparatus is now weighed, and then ex- posed to a proper heat, as long as gaseous fluoride of silicon is evolved. The last portions are removed by exhaustion under the air-pump. The loss of weight which it experiences is fluoride of silicon, from which the amount of fluorine is cal- culated. 1*395 parts of fluoride of silicon are formed for each part of fluorine. To test this method, the author estimated the amount of fluorine in fluorspar, and the results were ac- curate to the first decimal place. To avoid the use of the air- pump, Eresenius modifies the apparatus by adding a second tube, closed with a wax stopper, the lower end reaching nearly to the bottom of the flask ; the last traces of gas may, with this arrangement, be removed from the flask by simple suction : for which purpose a suction tube is applied, filled with dry cot- ton at the lower end, and with moist cotton in the centre. The above process is well adapted for the separation of fluorine from the silicates. Another method, proposed by Berzelius and modified by Eegnault, is the following : the finely levigated powder is kept for some time in a state of fusion, with four times its weight of carbonate of soda ; it is then extracted with boiling water, and the filtered solution is mixed with a solution FLUORINE. 509 of carbonate of zinc in ammonia, and evaporated to dryness. By digesting the dry residue with boiling water, a solution is obtained, which contains the whole of the fluorine in the form of fluoride of sodium, in conjunction with carbonate of soda. It is carefully neutralized with hydrochloric acid, and the car- bonic acid completely removed by allowing it to stand for some hours under a bell jar, by the side of a vessel containing solution of caustic potassa. The hydrofluoric acid is finally precipitated and estimated in the form of fluoride of calcium. Where the fluoride to be analysed contains water, the amount of that substance cannot be estimated by simply heating the fluoride, since the simultaneous action of the air and the water sometimes effects a partial decomposition of the fluoride, and a portion of hydrofluoric gas escapes along with the vapour of water. In such cases the analysis is performed as follows : The joint amount of water and fluoride is first determined by heating a known weight of the substance with concentrated sulphuric acid in the manner above described ; a fresh portion of the compound is then mixed in a little glass retort with about six times its weight of finely pulverized and recently ignited protoxide of lead ; the mixture is covered with a layer of oxide of lead, the retort is weighed, and heat is then applied, gentle at first, but gradually increasing to redness ; aqueous vapour, unaccompanied with the slightest trace of hydrofluoric acid gas, escapes. The retort is allowed to cool ; it is then weighed again, and the amount of water calculated from the loss which it has sustained. By the first operation the joint amount of water and fluorine has been learnt; and we have only now to deduct from that amount the quantity of water, indicated by the second experiment, to find the quantity of fluorine in the substance under examination. Separation of Fluorides from Phosphates. The process of Berzelius as modified by Kegnault, which we have described above, for separating fluorides from silicates, may also be em- ployed when phosphates also are present. In this case the solu- tion filtered from the silicic acid will contain phosphate of soda in addition to fluoride of sodium and carbonate of soda. It is analysed as follows : it is rendered aminoniacal, and mixed in a glass flask which can be closed air-tight with a solution of chloride of calcium ; the precipitate which is produced consists of fluoride of calcium and phosphate of lime. It is washed with the proper precautions, dried, ignited, and weighed ; it is 510 QUANTITATIVE ANALYSIS. then treated in a platinum crucible with concentrated sul- phuric acid, and heated until the whole of the hydrofluoric acid is expelled. The residue is treated with alcohol, in which sulphate of lime is insoluble, and the phosphoric acid dissolved in the alcohol is precipitated and estimated as pyrophosphate of magnesia. The amount of fluorine originally present in the mixed precipitate of fluoride of calcium and phosphate of soda, is learnt by the loss of weight sustained by the action of the sulphuric acid. 203. NITRIC ACID. (1.) Estimation of free Nitric Acid by Hydrate of Baryta. If nitric acid exist in a solution in a free state, baryta-water may be added until the liquid begins to acquire a slightly alkaline reaction : the whole is then evaporated to dryness. The dry mass is dissolved in water, arid the carbonate of baryta which i may have been formed during the evaporation, having been re- i moved by filtration, dilute sulphuric acid is added. The quan- tity of nitrate of baryta is -calculated from the weight of the i sulphate precipitated, and from this is deduced the weight of the nitric acid originally present in the solution. Care must i be taken not to add the baryta-water in too great excess. According to Schaffgotsch (PoggendorfTs ' Annalen,' .cviii. i p. 64), free nitric acid may be estimated in the form of nitrate of ammonia, a salt of perfectly constant composition, which can be easily weighed in a covered crucible, after being dried at 240 F. (2.) By the process of Acidimetry. This method, which is based on the determination of the amount of carbonic acid ex- pelled by a weighed quantity of the acid under examination, is best performed with the alkalimetrical apparatus of Fresenius and Will (see annexed figure). It being indispensable that the r bicarbonate employed for this experiment should be entirely free from neutral or sesquicarboiiate (though its perfect free- dom from any admixture of sulphate and chloride is nob of con- jj sequence) Fresenius and Will recommend to purify the bi- \ carbonate of soda of commerce in the following manner: ^Half a pound or a pound is reduced to a uniform powder, and a portion of it first tested with bichloride of mercury ; if the solution becomes merely turbid or milky at first, it is tole- \ rablv free from carbonates, the powder is then put into a glass jar and covered with the same amount of cold rain-water ; it is NITRIC ACID. 511 allowed to stand for twenty-four hours, with frequent stirring. The salt is then placed upon a funnel, the tube of which is stopped with loose cotton, so as to allow the ley to drop off. The salt is then washed several times with small quantities of cold rain-water ; what remains is generally pure, and adapted for acidi metrical purposes. It is dried between some sheets of blotting-paper, without the aid of heat, and kept for use in a well-closed glass jar. The ope- ration is conducted as follows : A portion of the fluid to be examined is carefully weighed into b ; if the acid be concen- trated, from four to eight times its amount of water should be Fig. 84. added. A little glass tube is provided, capable of holding about 75 grains of bicarbonate of soda ; it is filled nearly to the brim with the alkali, and, a silk thread being tied round its upper end, it is lowered down into b, so as to remain suspended perpendicularly when the flask is closed, the silk thread becoming fixed between the cork and the glass. The apparatus is then arranged as described (page 269), and accurately weighed. When the operation is about to be com- menced, the cork of the flask b is slightly loosened, so as to allow the little tube, together with its silk thread, to fall into the acid, and immediately fixed again air-tight. The evolution of carbonic acid commences instantaneously, and continues in a uniform and uninterrupted manner till the acid is completely neutralized. The operation may be accelerated by repeatedly shaking the apparatus. When no more bubbles of gas appear upon agitation, the flask b is placed up to its neck in hot water, and there left until the renewed evolution of gas has com- pletely ceased ; this is also promoted by agitating the appa- ratus. The wax stopper is then removed from d ; the flask removed from the hot water, wiped dry, and suction applied to e until the carbonic acid is completely removed. The appa- ratus is allowed to cool, and again weighed ; the amount of anhydrous acid in the sample examined is determined from 512 QUANTITATIVE ANALYSIS. the amount of carbonic acid evolved ; every two equivalents of the latter indicate one equivalent of nitric acid. It is evident that this method applies to all acids capable of decomposing bicarbonate of soda, and, consequently, that it cannot be em- ployed to determine nitric acid in the presence of any other free acid. Determination of Nitric Acid in combination with bases. Those nitrates whose bases form with sulphuric acid com- pounds which are not decomposed by mere ignition, may be analysed by heating them in a platinum crucible with sulphuric acid, and, having expelled the excess of acid by feeble ignition, determining the quantity of nitric acid expelled, by calculation from the weight of the sulphate obtained. Nitrates which con- tain no water of crystallization, may also be analysed by simple ignition, nothing but pure metallic oxide being in most cases left; the amount of nitric acid is inferred from the loss of weight. If the salt to be examined be mixed with two or three times its weight of perfectly anhydrous borax, the decompo- sition is effected at a lower temperature, and without danger of deflagration. A still better method is to fuse the alkaline ni- trate to a dull red-heat with 2J times its weight of bichromate of potassa : neither chlorides nor sulphates are under these conditions decomposed. Hydrated nitrates are analysed by determining the base in one portion, and the water and nitric acid in another, by the method of organic elementary analysis. Nitrates, the bases of which are precipitable by sulphuretted hydrogen, may be analysed by removing the base by that re- agent, excess of which should be avoided, and determining the amount of nitric acid in the nitrate by hydrate of baryta, in the manner described at the beginning of this section. The analysis of those nitrates which are insoluble in water may be effected also by sulphuretted hydrogen. They should be reduced to a state of minute division, and kept suspended in water while a stream of sulphuretted hydrogen is passed through the liquid. From baryta, strontia, and lime, nitric acid may be separated by mixing the solution wdth slight excess of sulphuric acid ; and, in the case of lime and strontia, adding also alcohol, to render the precipitation more complete. Baryta- water is added to the fluid filtered off from the precipitate, until it is slightly alkaline ; it is then evaporated to dryness on the water-bath, redissolved in water, filtered, and the filtrate treated as above NITRIC ACID. 513 directed. Another method, which may be adopted for deter- mining the amount of nitric acid in the nitrates, is to distil the nitrate with sulphuric acid diluted with twice its volume of water at a temperature not above 160 or 175. The opera- tion is conducted in a retort with the neck drawn out, and bent so that by means of an india-rubber tube it may be connected with a little receiver with three bulbs (similar to that used for collecting ammonia in the estimation of ammonia, by the method of Will and Varrentrapp), containing a known volume of a standard solution of potassa or soda. The distillation must be continued for three or four hours. It may be effected in a water-bath in a vacuum, either by means of an air-pump, or by expelling the air from the apparatus by boiling, and then closing hermetically. If the nitrate is mixed with chloride, a solution of sulphate of silver, or of moist oxide of silver, is added previous to distillation (Rose). For estimating nitric acid in the indirect way, Eeisch has proposed heating the nitrate at a barely visible red-heat, with four or six times their weight of pulverized quartz, which he prefers to bichromate of potassa, because the mass calcines without melting, and there is therefore no fear of portions being projected from the crucible, and because the presence of chlorides and sulphates is no inconvenience, the nitrates alone being decomposed. Cruin (Journ. fur Prakt. Chein., 41, 201) estimates nitric acid by transforming it into binoxide of nitrogen. The salt dissolved in water is introduced into a graduated jar over mer- cury, and then sulphuric acid added to the extent of three times the volume of the water used, after a few hours the binoxide is disengaged ; the volume of the gas in the tube is measured, due regard being had to the temperature, and the height of the barometer ; the binoxide is then absorbed by throwing up into the tube a solution of protosulphate of iron, and the amount of absorption being carefully noted, the quantity of nitric acid is calculated therefrom. Schlosing (Journ. fin? Prakt. Chem.,62, 142) also converts the nitric acid into binoxide, but instead of measuring the volume of the gas he reconverts it into nitric acid. The solution of the nitrate is boiled in a flask, till all the air is expelled ; an acid solution of protochloride of iron is then drawn into the flask, and the liquid is again boiled, binoxide of nitrogen is evolved, which is collected over mer- cury in a balloon filled with mercury and milk of lime : the gas 514 QUANTITATIVE ANALYSIS. is then brought, without loss, into contact with oxygen and water, s"o as to convert it again into nitric acid, which is esti- mated with normal alkali in the usual manner. This method, which may be used in the presence of organic matter, " has passed successfully," observes Fresenius (Quant. Anal., p. 309), " through an ordeal of numerous and searching experiments." It was devised by the author for estimating the amount of nitric acid in tobacco. Valuation of Nitre. i. Method of Pelouze (' Comptes Rendus,' January, 1847). This is founded on the peroxidation of a protosalt of iron by the oxygen of the nitric acid, a solution of permanganate of potassa being employed, as in Marguerite's process (see IRON), to indicate the completion of the oxidation. The operation is as follows : It is first accurately determined how much pure nitrate of potash is requisite to peroxidize a known weight of iron dissolved in excess of hydrochloric acid. Pelouze found that 2 grammes of pure iron (pianoforte wire) dissolved in a considerable excess of hydrochloric acid, required from 1/212 to 1*220 gramme ; on an average therefore 1'21G gramme of pure nitrate of potash. He examined the nature of the gases given off in this reaction, and found them to consist of hydro- chloric acid and binoxide of nitrogen. Converting these num- bers into equivalents, they correspond to 6 equivalents of iron j and 1 equivalent of nitrate of potash : the acid of this last salt is decomposed, therefore, into binoxide of nitrogen, which is | disengaged, and into 3 equivalents of oxygen, which deprive the f hydrochloric acid of 3 equivalents of hydrogen to form 3 equi- valents of water and liberate 3 equivalents of chlorine, which produce with the 6 equivalents of protochloride of iron 3 equi- valents of perchloride. These reactions will be better under- >, stood by putting them into an equation, thus : 4 HO + KC1 + N0 2 + 3 (Fe 2 Cl 3 ). It will be observed, by the way, that this decomposition of ni-!( trates by protosalts of iron in the presence of an excess of hy- drochloric acid furnishes a method of preparing binoxide of nitrogen. This decomposition being established, it occurred to Pelouze that it might be made the basis of a ready and simple method of analysing the nitrates. He found that the presence of sulphates and chlorides in no way interfered with the decom- NITRIC ACID. 515 position in whatever proportion they were present, and there only therefore remained to be found, a safe and ready method of determining, in the assay of an impure nitrate, the quantity of iron not peroxidized. Such a method has been described by Marguerite. Suppose, for instance, that having operated upon 3O88 grains of iron, and 18*77 grains of impure nitre, the per- manganate of potassa indicates that 3'08"> grains of iron have not been peroxidized, it may be concluded that 3O88 grains minus 3-088 grains, or 27*79 grains, have been peroxidized : now, if the nitre had been pure, the whole 30'88 grains would have been peroxidized ; we, therefore, have the proportion As 30-88 : 18-77 : : 27'79 : 16'89 ; the 18'77 grains of impure nitre consequently contain 16'89 grains of nitrate of potash, or 90 per cent. The following example will illustrate the practical details of the method. The specimen may be supposed to be taken from a sample of the crude nitre of commerce as it is sent to the refiner. 30'88 grains of pianoforte wire are placed in a flask of the capacity of about 10 cubic inches, and on them are poured about 1500 grains of strong hydrochloric acid, the flask is closed with a cork furnished with a drawn-out tube, and the iron dis- solved at a gentle heat. When the whole is dissolved, 18'52 grains of the nitre under examination are introduced, the flask is immediately closed, and the liquid boiled ; it becomes of a brown colour; dense vapours of hydrochloric acid, mixed with binoxide of nitrogen, issue from the orifice of the drawn-out tube, and prevent the access of air. As soon as the liquor loses the brown colour, it becomes yellow, and gradually brightens : after boiling for 5 or 6 minutes, and when the liquid has become perfectly transparent, the flask is removed from the fire, the liquid which it contains, and the wash water are poured into a flask capable of holding about 60 cubic inches, which is then entirely filled with ordinary water : upon this a solution of per- manganate of potash of known strength is gradually added from a graduated burette. The operation is then finished, and there remains only to calculate the result. Let us suppose that the solution of permanganate was of such a strength that 25 divisions of the burette were required to peroxidize 7'72 grains of iron, or 50 divisions for 15'44 grains of that metal ; and let us further suppose that to complete the preceding experiment 10 divisions of the same solution were 516 QUANTITATIVE ANALYSIS. required ; now, if 50 divisions of this permanganate suffice peroxidize 20 of iron, how much ought 10 divisions to per oxidize ? As 50 : 15-44 : : 10 : 3'088. 3'088 therefore are subtracted from 30*88 grains of iron, andii may be concluded that the remaining 27'888 have been per- oxidized by 18'52 grains of crude nitre : but it is known tha 30*88 grains of iron repreeent 1877 grains of pure nitrate o potash, the quantity of that salt therefore corresponding to 27'888 grains of iron is found by the following proportion :- As 30-88 : 1877 : : 27'888 : 16-95. In the 18*52 grains of saltpetre submitted to analysis there were therefore 16*95 grains of pure nitre = 91*4 per cent. In performing the experiment the operator must be careful in guarding against access of air : otherwise this would be apt to act upon the binoxide of nitrogen, and render it capable of per oxidizing a further quantity of iron, so that the amount of nitre would be exaggerated. However, the peroxidation of iron, when the metal is dissolved in a strongly acid liquid, is verys slow, as Marguerite has shown, and in the manner of conduct- ing the process, as described above, the inconvenience is avoided, for as soon as the iron has disappeared in the acid the flask is filled with hydrogen and hydrochloric acid gases : the nitrate which is introduced carries with it but a mere trace of air, and the liquid kept boiling disengages, by the drawn-out tube, acid and aqueous vapours, the issuing forth of which being always visible and readily maintained, does not allow the admission of any air. Pelouze states that he has confirmed the accuracy of his me- thod by the analysis of several other nitrates, such as those of soda, ammonia, lead, etc., and he observes further that it will enable us to determine the amount of water in certain nitrates, the composition of which is at present doubtful. This method has been subjected to an elaborate critical exa- mination by Abel and Bloxam (Quart. Jour. Chem. Soc., vol. ix. 97). They have arrived at the conclusion, that although the reaction upon which it is founded is perfectly correct, and that therefore it frequently happens that the most accurate re- sults are furnished by it, nevertheless, these results are liable to contain disturbing causes which are so far beyond the con- trol of the operator as to deprive the method of that certainty so essential to any process in use for commercial analysis. These NITRIC ACID. 517 disturbing causes are : 1. The action of air upon the nitric oxide gas by which nitric acid is re-generated. 2. The incom- plete expulsion of the nitric oxide from the fluid, in conse- quence of which it reduces a larger quantity of permanganate of potassa than corresponds with the amount of protoxide of iron contained in it ; this however is only to be feared in dilute solutions. 3. The evolution of nitric acid before it has acted upon the protochloride of iron, especially when the fluid boils very rapidly after the addition of the nitrate, and when the ex- cess of protochloride of iron is proportionally small. All these sources of error may, according to Fresenius(Liebig's 'Annalen,' May, 1858, p. 217)', be avoided, and perfectly satisfactory results obtained, by the following method of operating. Into the belly of a tubulated long-necked retort, capable of containing about 200 cubic centimetres, so fixed that the neck is directed a little obliquely upwards, is put about 1'5 gramme of fine pianoforte wire, accurately weighed, and from 20 to 40 cubic centimetres of pure fuming hydrochloric acid are added. Hydrogen gas washed in a solution of potash, is now passed in through the tubulure by means of a glass tube reaching about 2 centi- metres into the retort, and the neck of the retort is united to a U-shaped tube containg a little water. The belly of the re- tort is placed on the water-bath, and gently heated, till the iron is completely dissolved. It is then allowed to cool in the cur- rent of hydrogen, which is afterwards strengthened, and the ni- trate weighed in a small tube is thrown in with the tube, through the neck of the retort. The quantity of the nitrate should be calculated, so that it may only contain about 0'200 gramme of nitric acid. After the union of the neck of the retort with the U tube, the contents of the retort are heated on the water-bath for about a quarter of an hour, the water-bath is then removed, and heat is applied by a lamp so as to produce strong ebul- lition, until the solution, which has a dark colour from the ab- sorbed nitric oxide gas, has acquired the colour of perchloride of iron, after reaching this point, the boiling is continued for a few minutes. Every time the fluid is agitated, care must be taken that dry salt is never deposited on the walls of the retort. Before the boiling is stopped, the current of hydro- gen gas is strengthened in order to prevent the entrance of air through the U tube when the lamp is removed. The retort is allowed to cool in the current of hydrogen, the contents are much diluted with water, and finally the iron still existing PART II. 2 M 518 QUANTITATIVE ANALYSIS. as protoxide, is determined with solution of permanganate of potassa. This method may, according to the author, be employed with! the greatest confidence when no organic matters are present. Instead of estimating the amount of protoxide of iron on which the nitric acid has not acted, by permanganate of potassa, Braun (Jourii. fur Prakt. Chem., Ixxxi. 421) determines the" amount of peroxide formed, by means of iodide of potassium and hyposulphite of soda. The perchloride of iron decomposes the iodide of potassium, and sets the iodine at liberty, even in; the cold, but more completely when the reaction is assisted by a gentle heat. The free iodine is estimated by a normal solution; of the hyposulphite, after having added a small quantity ofl starch paste. The normal solution of hyposulphite of soda is I prepared either by dissolving a known weight of the salt inj a similarly known weight of water, or by standardizing with a! solution of hyposulphite of unknown strength, the iodine set! free by perchloride of iron prepared from a known weight of j metallic iron. ii. Gay-Lussac's method, modified by Abel and Bloxam. The nitrate is converted into carbonate by fusion with charcoal, and the amount of alkaline carbonate is determined by a standards solution of sulphuric acid. In order to moderate the evolution; of nitrogen gas, a certain quantity of pure common salt is addedj to the mixture. The proportions usually employed, are 1 partj of lamp-black, 4 of nitre, and 24 of salt, to 20 of nitre. The| nitre and charcoal are first mixed in a platinum crucible with a glass rod, the chloride of sodium is then added, and an intimate! mixture of the whole is made. The crucible is covered and! heated, cautiously at first, to prevent too violent deflagration,' and afterwards for a few minutes at a red temperature ; when cold, the contents of the crucible are washed into a flask,! treated with hot water, the residual carbon filtered off, and the! alkalinity of the solution then determined by a normal solution' of sulphuric acid. Abel and Bloxam, who have thoroughly investigated this method, likewise (Quart. Journ. Chem. Soc., vol. ix. p. 110,' and vol. x. p. 107) employ an acid so far diluted that a little less than 1000 grain-measures are required to neutralize the amount of carbonate of potassa, corresponding to 20 grains of nitre. An acid of this description contains from 8*0 to 8'5 grains of anhydrous acid in 1000 grain-measures. The strength NITRIC ACID. 519 of the acid having been approximately ascertained by the neu- tralization of a known weight of carbonate of soda, is finally determined with the greatest possible accuracy by precipitation as sulphate of baryta. In performing the alkalimetrical assay a sufficient quantity of aqueous solution of litmus is added to impart a deep blue colour to the liquid ; the standard acid is then added until it acquires a wine-red tint from the dissolved carbonic acid ; the liquid is then boiled, and the acid very care- fully added, drop by drop, until the peculiar onion-skin red colour imparted by the stronger acids to litmus, is obtained. The accurate appreciation of the colour is much facilitated by using as a standard of comparison, a similar quantity of co- loured water to which a few drops of sulphuric acid have been added. A great number of experiments made by Abel and Bloxam disclosed a considerable uncertainty in the above method ; this they traced to the following facts : 1. That a considerable quantity of nitre may escape decomposition at a high tempe- rature, even when intimately mixed with an excess of finely divided carbon, and although vivid deflagration of another por- tion of the nitre had taken place in its immediate neighbour- hood. 2. That in the deflagration of a mixture of nitre, salt, and charcoal, ~binoxide of nitrogen is evolved in large quantity. 3. That in the deflagration, cyanide of potassium is formed, and that, even where much care had been taken to remove moisture from the materials, a notable quantity of ammonia is evolved. 4. That the oxidation of cyanide of potassium by ex- posure to the air at a high temperature is not prevented by the presence of carbon in a finely divided state, or even by covering it with a layer of carbon. By substituting pure, ignited, and finely divided graphite for charcoal, in the conversion of the nitre into carbonate of potassa, and by the subsequent treat- ment of the fused mass with chlorate of potassa in order to oxidize the cyanate of potassa formed by the oxidation of the cyanide of potassium, and to obviate the errors which may arise from the reduction of any sulphates which may be present, Abel and Bloxam succeeded in obtaining very accurate results. The proportions they recommend are 20 grains of nitre, 5 grains of ignited graphite, and 80 grains of pure common salt. The crucible, covered with a jacket, should be loosely covered, and moderately heated for two or three minutes over the gauze gas-burner ; it should then be allowed to cool to such an extent 2M2 520 QUANTITATIVE ANALYSIS. that chlorate of potassa shall not fuse when sprinkled upon the mass. About 25 grains of the chlorate are added, so as tol form a layer upon the surface. A very gentle heat is first applied, until most of the chlorate has been decomposed, when; the temperature is raised to bright redness, and maintained in| that state for two or three minutes ; when cool the mass is to be carefully shaken out of the crucible into a funnel, and the! crucible and cover washed with boiling water. The mass is; dissolved by a stream of hot water from a washing-bottle, and; the solution allowed to run into the flask in which the determi- nation is to be made ; the liquid is coloured with litmus, and neutralized with the standard acid. Estimation of Nitric Acid by conversion into Ammonia (Schulz, Chem. Centralblatt, No. 58 ; and Yernon Harcourt,; Journ. Chern. Soc., vol. xv. p. 381). When a moderately con- centrated solution of potassa is poured upon a mixture of iron and zinc, hydrogen gas is freely disengaged, even without ther application of heat ; the action is electrolytic ; the zinc is oxi-i dized, and the hydrogen formed upon the surface of the iron4 If a nitrate be added to the mixture evolving hydrogen, ammonia is immediately produced, This reaction furnishes a good quali-, tative test for nitric acid. The fluid to be examined is reduced to a small bulk, and poured into a test-tube containing a mix- ture of granulated zinc and clean iron filings ; a small quantity of strong solution of potassa is then added; the mixture igj heated, and the gases evolved are led into a small quantity ot dilute hydrochloric acid. The acid solution is supersaturated; with potassa, and tested with a drop of potassio-iodide of mer-; cury (p. 64). To apply this reaction to the quantitative de-| termination of nitric acid in a nitrate, Harcourt proceeds as follows : A flask, having a capacity of about 12 cubic inches is connected by a bent tube drawn out and recurved at its ex< tremity, with a smaller flask or bulb, and the two are so arranged! that they both rest at a considerable angle of inclination, upor , a sand-bath. The smaller flask, into which a little water i& poured, is connected with a condenser which leads into a tubu- lated receiver containing normal sulphuric acid coloured witt litmus; the end of the distilling- tube reaches to about the middle of the receiver, through the tubulure of which is in- serted a tube with two or more bulbs, containing also coloured normal acid. In the upper part of the condensin^-tube is * small tubulure ; during the process of distillation, this is closec NITRIC ACID. 521 by an india-rubber plug, which may be removed after the opera- tion, is over for the purpose of sending a stream of water down the tube to remove all traces of ammonia from its sides. Into the larger, or distilling flask, are introduced 700 or 750 grains of granulated zinc, with about half that quantitv of clean iron filings, which have been recently ignited in a covered crucible, then the weighed nitrate with water sufficient to dissolve it, and lastly, a considerable excess of strong solution of caustic potassa ; the flask is immediately connected with the con- densing apparatus and heat applied ; when the distillation has actually commenced the water in the second flask is made to boil gently : by this arrangement the fluid is twice distilled, and any traces of potassa which may have passed over from the distilling-flask are sure to be retained. In his experiments with nitre, Harcourt has generally taken 0'5 grain of that sub- stance, 20 cub. cent, of water, and the same volume of solution of potassa, sp. gr. 1'3 ; with these quantities the distillation occupies about an hour and a half; it is completed when the hydrogen is pretty freely liberated, as the potassa becomes concentrated ; the lamp is then removed, and the whole allowed to cool ; the distilling-tube is rinsed into the receiver, and the determination is completed by adding standard alkali from a burette to the fluid in the receiver till a change of colour appears ; with alkaline nitrates this method gives very exact results ; with the nitrates of the heavy metals the results do not appear to be quite so satisfactory. PugJis method (Quart. Journ. Chem. Soc., vol.xii. p. 35). This process is based on the fact that when a nitrate is digested under pressure at a temperature of about 320 F., with an ex- cess of protochloride of tin, and hydrochloric acid, the following reaction occurs : 1 equivalent of nitric acid therefore, under the above condi- tions, converts 8 equivalents of tin from proto- to perchloride ; ^consequently if an unknown quantity of nitric acid be digested with sufficient excess of solution of protochloride of tin of known strength, and the quantity changed into perchloride be afterwards found by a suitable method of titration, the propor- tion of nitric acid will be found, supposing in all cases that no other substance is present capable of effecting the same change in the tin solution. Tor the purpose of estimating the amount of tin oxidized, Pugh adopts the method proposed by 522 QUANTITATIVE ANALYSIS. Streng (Pogg. Ann. xcii. 57), which consists in ascertaining how much of a solution of bichromate of potassa of known strength, is necessary to convert a given amount of protochlo- ride of tin in hydrochloric acid into perchloride, the point of complete chloridation being known by the deep blue colour produced by the liberation of iodine from iodide of potassium, in the presence of starch, by the first drop of bichromate of potassa solution above that necessary to raise the protochloride to perchloride. According to Sutton, however (' Volumetric Analysis,' p. 60), this method is not satisfactory, owing to the variable amount of oxidation which the tin solution undergoes when different quantities of water or acid are present during the titration ; he therefore adopts the following method pro- posed by Lessen : Metallic tin or its protosalt is dissolved in hydrochloric acid, if not already in solution, a tolerable quantity of the double tartrate of soda and potassa added, together with bicarbonate of soda in excess. If enough tartrate is present the solution will be clear; starch liquor is then added, and the mixture titrated with a decinormal solution of iodine, till the blue colour is permanent. For details respecting this elegant volumetric method we must refer to Pugh's memoir, or to Button's very useful work, referred to above. 204. CHLOBIC Acn>. When this acid exists in a free state in aqueous solution, it is most simply estimated, according to Fresenius, by converting it into hydrochloric acid by means of sulphurous acid, in the following manner: The solution is mixed with a considerable excess of an aqueous solution of sulphurous acid, and well agi- tated ; the flask is then stopped, and its contents digested for some time. The excess of sulphurous acid is then removed by bichromate of potassa, and the liquid, being acidified with nitric acid, is precipitated with nitrate of silver, and the amount of chloric acid originally present calculated from the quantity of chloride of silver obtained. Chlorates are analysed either by measuring the volume of oxygen gas which they evolve by their decomposition, or by exposing a known weight to ignition and calculating the amount of chloric acid from the loss of weight thereby ex- perienced ; should the salt under examination contain water of crystallization, it must be ignited in a retort of hard glass, to CHLORIC ACID. 523 which there is attached a weighed tube containing chloride of calcium. When a chloride and a chlorate are contained in the same so- lution, and are each to be estimated, the amount of chlorine in the chloride is first determined in a weighed quantity of the solution, by precipitating it as chloride of silver, another weighed portion of the solution is evaporated to dryness and ignited, by which the chlorate present is converted into chloride, it is then redissolved in water, and precipitated by nitrate of silver ; from the amount obtained we have merely to deduct that yielded by the first experiment to obtain the requisite data for calculating the quantity of chloric acid. 524 CHAPTEE IX. ON THE ELEMENTAKY ANALYSIS OF OK&ANIC BODIES. 205. Numerous and diversified as are the forms and properties of vegetable and animal substances, and complex as is, in general, their composition, they are nevertheless made up of a very few of the so-called elements of the material world. A vast number of organic substances consist but of three of these elements, viz. carbon, hydrogen, and oxygen,, and very many contain only two, carbon and hydrogen. There is, again, a very large class of organic bodies into the composition of which nitrogen enters in conjunction with carbon, hydrogen, and oxygen, and there are others to which, together with these four principles, sulphur and phosphorus, iron, manganese, and a few other elements appear to be essential. Thus, of the sixty-three substances at present re- cognized by chemists as elementary, not more than ten or a dozen are found in the organic world. It is true that organic compounds may contain other bodies ; but such are of artificial production, and do not therefore militate against the. fact we have just stated, viz. that by far the greater number of that almost bewildering variety of substances which we meet with as the products of animal and vegetable life, and of their reactions on each other, are constituted of three or four only of the elementary bodies, and that about a dozen of them are the most that are ever found to enter into their composition. The older chemists, being unacquainted with any methods of resolving organic substances into their ultimate constituents, or of re-arranging their elements into such forms of combination as might enable them to pronounce upon their constitution, were obliged to satisfy themselves with the examination of the results of their destructive distillation; they thus resolved them into a series of products which, being liable to infinite variation dependent upon the manner in which the distillation ELEMENTARY ANALYSIS OF ORGANIC BODIES. 525 was effected, and the degree of heat employed, could, it is ob- vious, throw but little light on the composition or properties of the substance under examination. A vast number of analyses were, however, made in this way by the earlier pneumatic chemists, of whom Hales and the celebrated Priestley studied attentively the gaseous products of the operation. The first chemists who attempted to combat the difficulties which surrounded the original attempts to gain an insight into the composition of organic substances, by the examination of the results of their combustion in oxygen, were Gay-Lussac and The- iiard. Their method was to mix a certain weight of the sub- stance to be analysed into one or more pellets with chlorate of potassa, and to drop them, by a suitable contrivance, into an ignited glass tube, receiving the product of the combustion into a jar filled w r ith mercury and inverted in a mercuric-pneumatic apparatus. It was soon found, however, that a perfect com- bustion of the substance could not be effected with chlorate of potassa, in consequence of the sudden manner in which this salt liberates its oxygen when exposed to heat. Gay-Lussac there- fore suggested the employment of oxide of copper, a substance which does not decompose when ignited alone, but which readily yields its oxygen when heated strongly with an organic sub- stance. The idea of using this oxide was a very happy one, and there can be no doubt that we owe to its application, in a great measure, that vast addition which has of late years been made to our knowledge in this department of chemistry. Dr. Prout employed an entirely different method ; it was complicated, though exceedingly accurate, as is evident from the fact that the analyses performed by its distinguished inventor, are still regarded as models of scientific research, and from the circumstance that Dumas was induced to recur to it notwith- standing its complexity. Dr. Prout's object was to burn the substance effectually in oxygen gas ; in which case, supposing it to contain three elements, hydrogen, carbon, and oxygen, one of three things must happen : 1st, the original bulk of the oxygen must remain the same, in which case the hydrogen and oxygen in the substance exist in it in the same proportion in which they exist in water ; or 2nd, the original bulk of the oxy- j gen is increased, in which case that element must exist in the compound in a greater proportion than it exists in water ; or 3rd, the original bulk of the oxygen is diminished, in which case the hydrogen must predominate. In the first case, the compo- 526 QUANTITATIVE ANALYSIS. sition of a substance may be known by determining the quantity of carbonic acid gas yielded by a known quantity of it ; while in the other two, the same can be readily ascertained from the same data, by noting the excess or diminution of the original bulk of oxygen. For full details respecting the manner of conduct- ing the whole operation, the student is referred to Dr. Prout's original paper in the Philosophical Transactions for 1827, or to an abstract of the same in Brande's 'Manual of Chemistry,' where he will also find woodcut illustrations of the different parts of the apparatus. We must content ourselves here with a general statement of the advantages of this method of analysing organic compounds, as summed up by the inventor. In the first and chief place, there is nothing to be apprehended from moisture. Whether the substance to be analysed be naturally a hydrate, or in whatever state it may be with respect to water, the results will not be affected ; and the great problem whether the hydrogen and oxygen exist in the substance in the proportions in which they form water, or whether the hydrogen or the oxy- gen predominates, will be equally satisfactorily solved, and that (within certain limits) independently of the weight of the sub- stance operated on. When, however, it is the object to ascer- tain the quantity of carbonic acid gas and water yielded by a substance, it is of course necessary to operate on a known weight ; but this being once determined, there is no fear, as in the then common methods, of exposing the substance to the atmosphere as long as may be necessary. The hygrometric properties of the oxide of copper, as well as its property of con- densing air, are also completely neutralized, for the whole, at the end of the experiment, being left precisely in the same state as it was at the commencement, the same condensation of course must take place, and any little differences that may exist,, are rendered quite unimportant from the bulk of oxygen gas operated on, which of course, in all cases, should be consider- ably greater than that of the carbonic acid gas formed. An- other advantage of this method is, that perfect combustion is ensured by it. There is, also, no trouble of collecting and esti- mating the quantity of water, a part of the common process attended with much trouble, and liable to innumerable acciden- tal errors ; here, on the contrary, the results are obtained in an obvious and permanent form, and, from the ease with which they are thus verified, comparatively very little subject to error. We shall now proceed to describe, somewhat in detail, the more ELEMENTARY ANALYSIS OF ORGANIC BODIES. 527 modern methods of analysing organic substances, for the perfec- tion of which the scientific world is mainly indebted to Baron Liebig. This method combines with great exactness such sim- plicity of execution, that it is frequently resorted to as the easiest method of testing the purity of a substance; and it is generally admitted that to its invention must be ascribed the rapid progress of organic chemistry within the last few years. The process consists in burning the organic substance together with oxide of copper or chromate of lead, by which its carbon is converted into carbonic acid and its hydrogen into water, both of which are collected and weighed ; if the body contain nitrogen, it is either collected in a gaseous state, or is deter- mined by other methods, and the oxygen is calculated by de- ducting from the original weight of the substance analysed, the combined weights of the carbon, hydrogen, and nitrogen formed. 206. Preliminary testing of the Organic Substance. Before commencing the ultimate analysis of an organic compound, the operator must assure himself of its purity. The presence or absence of inorganic non-volatile matter is easily ascertained by heating the substance on platinum foil over a spirit lamp, and, if necessary, urging the flame with the blowpipe ; if it does not wholly disappear, the residue must be minutely ex- amined ; and if it has been determined to proceed with the ulti- mate analysis of the compound, notwithstanding the presence of inorganic matter, the amount of the latter must be accurately ascertained. We will suppose, however, that the substance has been proved to be free from inorganic non-volatile matter ; it has next to be tested for nitrogen, sulphur, and phosphorus. a. Testing for Nitrogen. If the quantity of this element be large it is easily detected by simply burning tire body, which emits during ignition a smell of singed hair. In smaller quan- tities, it may be detected by mixing the substance with soda-lime, and heating it in a test-tube ; the water of the hydrate is thereby decomposed, its oxygen passing to the carbon and hydrogen of ^ the organic substance, and its hydrogen to the nitrogen (if it jj contain any), giving rise to the formation of ammonia, which I may be detected at the orifice of the tube by holding near it la feather, moistened with strong acetic acid. The following j beautiful and delicate test for minute quantities of nitrogen in organic substances, was recommended by Lassaigne (Biblio- theque Univ., April, 1843). It is based on the facility with I which cyanide of potassium is formed when an organic sub- 528 QUANTITATIVE ANALYSIS. stance, containing even very little nitrogen, is calcined at a red- heat and protected from the atmosphere, together with an ex- cess of potassium. A small piece of this metal is placed at the bottom of a tube, and the matter to be tested is then added. The end of the tube is heated to a dark red in the flame of a lamp, then left to cool, and the potassa salt, formed by the cal- cination, dissolved with four or five drops of distilled water. The decanted liquor is tested with a drop of solution of ferroso- ferric oxide (magnetic oxide of iron) ; a dirty greenish precipi- tate is immediately formed, which being brought into contact with a drop of hydrochloric acid, becomes a dark blue, if the matter under examination contain a trace of nitrogen. In the contrary case, the precipitate of the hydrated oxide of iron is entirely redissolved, without any blue colouring. Lassaigne found that neither potassa nor its carbonate could be substituted for potassium in this process, neither is it applicable when the organic substance is accidentally mixed with a nitrate, or with an ammoniacal salt. Of these two methods of testing for ni- trogen, the latter is in its indications the least liable to miscon- struction. If the organic compound contain nitrogen in chemi- cal combination with oxygen, the presence of that element is readily detected by heating the substance in a tube and holding over the mouth a piece of paper moistened with solution of starch, to which a drop of iodide of potassium has been added, a blue colour is imparted to the paper if nitrous or nitric acids be present. b. Testing for and estimation of Sulphur. A variety of me- thods have been proposed for detecting and estimating the amount of this element in organic substances. The following are the most reliable : i. Ruling's process (Chem. Gaz.,vol. iv. p. 362). The finely pulverized substance is fused in a spacious silver dish, with about twelve times its weight of caustic potassa, perfectly free from sulphuric acid ; it is then mixed with pure nitre, in weight equal to about half the potassa employed, and heated till the fused mass appears perfectly white. The whole is then super- saturated with hydrochloric acid, and a salt of baryta added : the formation of a white precipitate indicates sulphur. It con- sists of sulphate of baryta, and is collected, washed, dried, and ignited. It weight furnishes the datum for calculating the amount of sulphur originally present in the substance. ii. Kemp's process (Chem. Gaz., vol. iv. p. 371). Fine ELEMENTARY ANALYSIS OF ORGANIC BODIES. 529 quartz sand, which has been carefully tested for sulphuric acid, is mixed with chlorate of potassa, in the proportion of two-thirds of the former to one-third of the latter. The substance to be examined is intimately mixed with the above, introduced into a large platinum crucible, and exposed to a high temperature over a spirit-lamp. The process is completed in a few seconds. The contents of the crucible are treated with boiling distilled water, the insoluble portion separated by filtration, and washed till no traces of sulphuric acid remain: the sulphates and phosphates, if any exist, may by this plan be estimated simultaneously by means of a salt of baryta. The object of the sand is to prevent the too rapid oxidation of the organic body, and the risk of a portion being thrown out of the crucible. in. Weidenbusctis process (Liebig's 'Annalen,' March, 1847). The substance is mixed with an excess of nitrate of baryta, and stirred in a beaker glass, with the most concentrated fuming nitric acid, to a thin paste ; the beaker is covered and exposed to the temperature of the sand-bath, replacing the evaporated nitric acid until the organic substance is entirely destroyed, which is perceptible from the circumstance that on evaporating the nitric acid still present, the mass begins to thicken, but no longer ascends in large bubbles in the glass, and dries quietly without frothing. The decomposed mass is now carefully washed into a platinum dish, dried at 212, and then heated gradually, until the saline mass is perfectly liquid. If the organic substance had not been completely destroyed by the nitric acid, it frequently happens that the mass takes fire, and a portion is projected from the slight deflagration. This must be avoided, as it occasions a loss of substance ; but, even though this were not the case, a smaller quantity of sulphate of baryta would probably be obtained, since it appears that the sulphur still contained in organic combination is not oxidized so readily by the deflagration of the mass, as by the slow action of the nitric acid. The fused mass is carefully removed from the platinum dish, treated with dilute acetic acid, and heat applied as long as any evolution of gas is perceptible. The carbonate of baryta is decomposed, and the residue (if any) consists of pure sulphate, which is collected on a filter, washed, dried, heated to redness, again digested with acetic acid after ignition, and again heated to redness. After the second igni- tion it is perfectly pure, and the amount of sulphur may be calculated from it. 530 QUANTITATIVE ANALYSIS. iv. Heintz's process (Poggendorff's 'Annalen, 3 Ixxii. p. 145), The sulphur is converted into sulphuric acid by burning the organic substance with oxygen and oxide of copper. For this purpose an ordinary combustion-tube, open at both ends, is half filled with copper turnings, and a little vessel containing the dry and pulverized substance placed in the tube, which is then connected with a bulb apparatus, filled with solution of potassa, perfectly free from sulphuric acid, and the combustion made in a current of oxygen. The contents of the bulb appa- ratus are now decanted into a spacious flask, containing a warm solution of chlorate of potassa in dilute nitric acid, and closed with a glass stopper. The sulphurous acid absorbed by the potassa, and which had formed from the decomposition of a portion of the sulphate of copper, is by this means perfectly oxidized. The mixture of oxide, sulphate, and metallic copper is treated in a beaker with the above mixture, and entirely dis- solved at a gentle heat ; the sulphuric acid is precipitated by chloride of barium, and estimated as usual. As, along with the sulphurous acid, sulphuric acid always passes over into the bulb apparatus, it is advisable, in order to avoid all loss of sul- phuric acid, which might arise from condensation of the latter on the cork of the combustion tube, to employ a tube which is drawn out at one extremity at a very obtuse angle into a fine tube, which fits into the stronger arm of a bulb apparatus made for the purpose, and then connected with caoutchouc. The author examined taurine by this process, and found in three experiments, in the first, 25*68 ; in the second, 25'66 ; and in the third, 25*49 per cent, of sulphur. Eedtenbacher's formula, C 4 H 7 N0 6 S 2 , requires 25*60 per cent. The combustion must, not be carried on too quickly, otherwise a slight loss might readily occur from vapour passing through the solution of. potassa without being absorbed thereby. In illustration of the importance of carefully examining organic substances for sul- phur, and of the great mistakes which may be committed from neglecting to do so, it may be mentioned that the substance above alluded to, viz. taurine, which contains nearly 26 per cent, of sulphur, had originally assigned to it the formula C 4 H 7 N0 16 and was supposed to be altogether free from that element. v. By Acetate of Lead. A very convenient method of de- tecting minute quantities of sulphur in organic substances, which does not, however, enable us quantitatively to estimate ELEMENTAEY ANALYSIS OF ORGANIC BODIES. 531 it, is to boil the compound nearly to dryness with a concentra- ted solution of caustic potassa. The residue is dissolved in water; and introduced into a flask provided with a tube funnel ; a strip of paper which has been thoroughly moistened with acetate of lead, and subsequently touched with a few drops of carbonate of ammonia, is suspended inside the flask between the cork and the neck; on pouring a little dilute hydrochloric acid slowly through the tube-funnel, sulphuretted hydrogen gas will be evolved, should sulphur be present, and the lead- paper will turn brown. c. Testing for, and estimation of Phosphorus. The substance may be oxidized by the process of either Ruling or Kemp, or with fuming nitric acid, or with a mixture of nitric acid and chlorate of potassa, and the solutions obtained may be examined for phosphoric acid by one of the methods described in the sec- tion On " PHOSPHORIC ACID." (1.) Analysis of non-volatile Solids, not containing Nitrogen. The first important point is to dry the substance thoroughly. This is done either in the water or air-bath, or in vacuo over sulphuric acid, or by any other of the methods described in the first chapter of this treatise, the means employed depending on the physical and chemical characters of the substance under ex- amination. With the view of illustrating the details of the whole operation, we will take, as the simplest case that can be brought before the student, the analysis of the crystallizable sugar of manna, namely, mannite, giving the results as they oc- curred in an actual analysis. The crystalline sugar, having been finely pul- verized, is dried in a watch-glass in the water- bath until two weighings, made at the interval of half an hour from each other, furnish a con- stant result. The fine powder is now introduced into a small stoppered, bottle, the exact size and shape of Fig. 85. The weight of the bottle (which is kept in the water-bath along with the substance, to ensure its being thoroughly dry) is, when empty 76-66 grains, and, when filled with the mannite, 84-76 grains. A Hessian crucible, of the capacity of about four ounces, is nearly filled with oxide of copper, carefully prepared in the following manner: Some shreds of pure copper (clippings from elec- A Fig. 85. 532 QUANTITATIVE ANALYSIS. trotypes) are dissolved in pure nitric acid; the solution is evaporated to dryness in a porcelain basin, and the residual basic salt heated to moderate redness in a Hessian crucible, until fumes of nitrous vapour cease to be evolved. It is con- stantly stirred during the ignition with a copper rod, to pro- mote the escape of the gas. The last traces of nitric acid ad- here somewhat obstinately to the oxide of copper ; the process should not, therefore, be hurried, but it is better to keep up a moderate heat for a considerable time than to attempt to quicken the operation by urging the fire, as in this case the lower portions of the oxide are apt to fuse, and, on cool- ing, the mass becomes so hard that it is difficult to remove it from the crucible, and subsequently, to reduce it to powder. Before considering the process to be finished, a portion of the oxide should be tested ; with which view a small quantity is introduced into a tube of combustion glass and strongly heated, the aperture of the tube being closed with the finger. On removing the latter the smell will generally indicate the existence of nitrous vapours, and the tube itself should be carefully examined for the appearance of ruddy fumes. The decomposition being complete, the oxide is allowed to cool, then, transferred to a mortar, in which it is reduced to a tolerably fine powder, after which it is fit for use, and may be preserved in a wide-mouth stoppered bottle. The same oxide may be employed for several experiments : by merely pouring a little acid on it, the metal is oxidized afresh, and the oxide is restored to its former purity by calcination. The Hessian crucible above alluded to is nearly filled with oxide of copper thus prepared : its cover is put on, and it is placed among the coals of a coal on charcoal fire, taking the greatest care that no organic matter from the fire comes accidentally into contact with it. It is heated to low redness : meanwhile, the other materials required for the analysis are being prepared. (a.) The Combustion Tube. This is drawn out from a tube of hard German glass, with the aid of a blowpipe lamp (i'ig. 13, p. 20). A tube free from flaws, and which has been previously tested as to its capability of bearing the requisite heat without blowing out, is selected, from thirty inches to three feet long ; it is heated, first gradually, and then with the full power of the lamp, in the middle, until it becomes sufficiently soft to rend asunder ; the two ends are then drawn from each other, and by a dexterous turn of the wrist two tubes are prepared, shaped as ELEMENTARY ANALYSIS OF ORGANIC BODIES. 533 shown in Fig. 86, each from fifteen to eighteen inches long. The fine points of the tubes are then sealed, and the open ends are fused at the edges, so as to enable them to bear the pres- sure of the cork without splitting. The tube, which is selected Fig. 86. for the combustion, is thoroughly cleaned and dried ; for this purpose a piece of clean filtering-paper is twisted round the end of a long stout copper wire, and with this the tube is well wiped out ; it may perhaps be necessary to wash it out with distilled water. In this case, the greatest care must be taken to dry it thoroughly afterwards, which may be done either by folding loosely over the mouth a piece of filtering-paper an'd burying it in the sand of the sand-bath, or by heating it from end to end over a spirit-lamp, drawing a current of air through it the whole time by putting a long tube, open at both ends, to the bottom, and sucking with the mouth. If the tube has been washed out with water, a small drop of the liquid will often obstinately remain at the very point of the sealed end, which is expelled with a sort of explosion on putting the point into the flame of a spirit-lamp. The tube being thus cleaned and dried, is kept corked till about to be used, in order to prevent the en- trance of any impurity. (b.) The Chloride of Calcium TufaThis is shown in Fig. 87. The chloride of calcium is prepared by dissolving chalk in hy- drochloric acid, filtering the solution, and evaporating it to dry- ness ; the dry mass is then fused in a Hessian crucible : it must Fig. 87. be kept in a well-stopped bottle. In filling the tube, a little cotton wool is first introduced at the large end, which is then closed with the finger ; vigorous suction is now applied at the small end (seen in the figure with its cork attached), the finger is then suddenly removed, and the cotton is forced up to the end of the bulb, where it acts as a plug in preventing any of PART II. 2 K 534 QUANTITATIVE ANALYSIS. the small fragments of the chloride from getting into, and chok- ing up, the narrow neck of the tube. The bulb is next filled with coarse pieces of chloride of calcium, and the remainder of the tube, to within half an inch of the end, is then filled up with finer fragments ; a second plug of cotton-wool is introduced, and finally the tube is closed with a good sound cork, well se- cured with sealing-wax, nicely smoothed round the edges ; through the centre of this cork there proceeds a piece of stout quill tube about three inches long, through which connection is established between the chloride of calcium tube and the po- tassa apparatus. A small hook is seen in the figure attached to the tube. This is made of platinum wire, and is intended to offer a facility for hanging the apparatus to the scale of the balance, in which it is accurately weighed. (c.) The Potassa-apparatus. The construction of this ap- paratus (the invention of which by Liebig may be considered as one of the leading causes of the rapid progress of organic chemistry within the last few years, from the remarkable man- ner in which it has facilitated the analysis of organic bodies) will be understood by reference to Fig. 88. It consists of a tube on which five bulbs are blown, three at the bottom, the middle of which is the largest, and one on each of the sides ; the three bottom bulbs communicate with each other by pretty wide openings, but each outer bulb is separated from the other by about two inches of tube ; one of the side bulbs, &, must be larger than either of the others. The object of this arrangement is to expose a large surface of caustic potassa ley to the gases evolved during the combustion of the organic sub- stance, so as to ensure the complete absorp- tion of the carbonic acid. The method of filling the bulbs with the alkaline ley is shown in Pig. 89. A porcelain basin is filled with a solution of caustic potassa, specific gravity about 1'27. This is prepared in the usual manner, or more conveniently by the follow- Fig. 88. i n g receipt, given by Dr. Gregory (' Instruc- ] tions for the Chemical Analysis of Organic Bodies,' by Justus Liebig, translated by Gregory) : Two parts of the carbonate of potassa are dissolved in twenty or twenty -four of boiling water. One part of quicklime is slaked by being covered with hot water in any convenient vessel. In this way the whole of the lime ELEMENTARY ANALYSIS OF ORGANIC BODIES. 535 is converted into a uniform cream, without the formation of any hard sandy particles which occur in the ordinary method of slaking. The cream of lime is added in small portions to the boiling solution of car- bonate of potassa, which is boiled a few minutes after each addition, water being occasionally added to sup- ply the loss by evaporation. When the whole of the lime has been added, the mixture is boiled for a short time longer, and is then allowed to cool, the cover of the ves- sel being carefully closed. After twelve hours the whole of the ley may be decanted perfectly clear and quite caustic, if the vessel has been nearly full. The carbonate of lime is sandy, and occupies a very small bulk. The clear liquid is now to be boiled down rapidly in a clean iron vessel, till small crystals begin to separate. It is then allowed to cool in a stoppered bottle of green glass, when it deposits the whole of the sul- phate of potassa originally present in the carbonate, that salt being entirely insoluble in a strong solution of caustic po- tassa. The specific gravity of this ley is from 1*25 to 1'27 ; it is nearly pure, containing no foreign matter except a little chloride of potassium, and is perfectly adapted for organic analysis. The basin being filled with the ley, the end of the tube com- municating with the largest of the side bulbs, is dipped into the liquor, and to the extremity of the tube, communicating with the smaller bulb, a suction tube is attached by means of a perforated cork : suction is now applied with the lips, and the ley is thus drawn into the apparatus to the extent shown in Fig. 88. The bulbs are then moved from their inclined position, the suction tube taken off, and the interior of the tube which had been immersed in the alkaline liquor is wiped dry with a twisted slip of filtering-paper ; and finally, the whole appara- tus is wiped thoroughly dry with a clean cloth. It is now ac- curately weighed, being suspended from the pan of the balance in the manner shown in Fig. 42, p. 218. 536 QUANTITATIVE ANALYSIS. (d.) The Combustion Furnace. This is shown in Fig. 90. It is made of sheet iron, from 20 to 24 inches long, and 4 inches high. The bottom is 3 inches wide, and is furnished with a series of small slips or apertures, which serve as a grate. The side walls are slightly inclined outwards, and are, at the top, about 5 inches apart. To support the combustion-tube, a series of straight and upright pieces of strong sheet-iron are riveted to the bottom ot the furnace, between the apertures ; they are of exactly equal heights, and their upper edges are slightly curved, so as to coincide exactly with the aperture at the end Fig. 90. of the furnace. To confine the fire to particular portions of the tube, two screens are employed, seen in their places in the figure. The openings in these screens must be sufficiently large to receive the analysing-tube without difficulty ; a screen is placed at the anterior part of the furnace during the com- bustion, to protect the cork of the tube from the heat of the coals. The furnace is placed, during the combustion, on a large tile, resting on a block of wood, a wedge being thrust be- tween the tile and the wood, to give the air free access to all the holes of the grating, or it may be placed on an iron tripod, as shown in Fig. 93. (e.) Charging the Combustion Tube. A sheet of clean glazed paper being laid on the table, the cork is removed from the combustion-tube, and warm oxide of copper introduced to about a (Fig. 86), this serves as a measure for the quantity required for the analysis ; it is shaken from the tube into a clean warm porcelain mortar, which should have a rough or unglazed bottom, as rendering the mixture less liable to adhere to it ; about an inch of oxide is left at the end of the tube. The sub- stance to be analysed is now shaken from the little stoppered bottle, in which it has been weighed, into the oxide in the mortar, and the whole is gently, but well mixed together with the pestle. The incorporation being complete, the mixture is transferred from the mortar to the tube, which is best done by depressing ELEMENTARY ANALYSIS OF ORGANIC BODIES. 537 the open extremity into the mixture, and turning the tube round ; the portions which cannot in this way be got into the tube, are poured on a smooth glazed card, and thus introduced. The mortar is now rinsed out with a fresh quantity of warm oxide, which is also transferred to the tube, and which will oc- cupy the space from a up to about b ; lastly, the tube is filled to within an inch of the end with pure oxide from the crucible. The distribution of the mixture in the tube is therefore as \ Fig. 91. follows : From the extremity to c, pure oxide ; from c to , the mixture of oxide and the' organic substance ; from a to I, rinsings of the mortar ; and from b to d, pure oxide. The pro- cess of filling the tube should be completed before the oxide of copper has got cold. The cork is now put in its place, and the tube being grasped between the forefinger and thumb of each hand, is tapped smartly a few times on the table, so as to shake together the contents, to free the pointed extremity entirely from oxide, and to have a channel above the mixture from end to end, to allow of the escape of the gaseous products of de- composition which would otherwise force the mixture from the tube. Fig. 91 shows the tube as it should appear with the mixture shaken into its place. Now, as during the process of Fig. 92. mixing, the oxide of copper may have absorbed aqueous vapour from the air, and as this moisture would vitiate the results of the analysis, by giving too high a number for the hydrogen, 538 QUANTITATIVE ANALYSIS. the next step of the operation is to get rid of the moisture, which is done by placing the tube in a wooden trough and co- vering it with hot sand, the temperature of which should be above 212, but not sufficiently high to char paper. To facili- tate the removal of the aqueous vapour, the air is exhausted from the tube, while it is surrounded with the hot sand, by means of a small exhausting syringe, as shown in Fig. 92 ; after a few strokes of the piston the stopcock is opened, by which fresh air is admitted, which becomes completelv dried by pass- ing through a chloride of calcium tube. This process of alter- nate exhaustion and readmissiou of air, is repeated ten or a dozen times, and the absence of all hygroscopic moisture is thereby ensured. (f.) Arrange/merit of the Apparatus for tlie Combustion. This will be perfectly understood from an inspection of Fig. 93. Fig. 93. The chloride of calcium tube is connected with the combustion tube by means of a perforated cork, which has been thoroughly dried in the water bath ; the operator must be careful in the selection of this cork, as, if a deficiency of tightness is detected when the apparatus comes to be examined previous to com- mencing the analysis, it is here that the fault will generally be found to lie. The cork should be thoroughly sound, and should be cut smooth and tapering; the hole through which the chlo- ride of calcium tube enters should be round and smooth. The chloride of calcium tube is connected with the potassa apparatus by a double caoutchouc tube secured by strong silk cord ; the caoutchouc should be soft so as to yield to the pressure of the cord, which should be tied in a bow, and not in a knot, in order that it may be readily removed at the end of the analysis. It will be noticed that it is the extremity of the potassa apparatus, in connection with the large bulb which is attached to the chlo- ELEMENTARY ANALYSIS OF OE.GANIC BODIES. 539 ride of calcium tube. The combustion tube should project a full inch from the end of the furnace, the screen being placed over it to protect the cork, as shown in the figure. The bulbs of the potassa tube rest on a folded towel, or silk handkerchief placed on blocks of wood raised to the requisite height, and the apparatus is inclined slightly by placing a cork underneath one of the side bottom bulbs, as seen in the sketch. The whole arrangement should have a slight inclination forwards, from the sealed end of the combustion tube to the junction of the chloride of calcium tube with the potassa bulbs. The ap- paratus being thus far disposed, the next step is to see that the junctions are all perfectly air-tight ; this is done in the fol- lowing manner: A red-hot coal is held near the large side- bulb of the potassa tube, the enclosed air becomes hereby ex- panded, the ley sinks, and, as the heat increases, a portion of the expanded air escapes in bubbles through the several bulbs, Fig. 94. out at the open end of the tube ; when about a dozen bubbles have in this way been expelled, the hot coal is removed, and the bulbs are set level on the cloth ; as the bulb cools, the air re- maining in it contracts, and the potassa ley rushes back and fills it more or less. It is now left for a few minutes, during which time the operator may be employed in re-weighing the little bottle which contained the substance about to be ana- lysed ; if at the end of five minutes or so the potassa ley should show no disposition to recede from the large side-bulb, and once more take up a horizontal position, it may be concluded that the apparatus is perfectly tight : on the other hand, should 540 QUANTITATIVE ANALYSIS. the liquid not preserve the level which it had acquired after the cooling of the large bulb, then the junctions are not tight, and the fault will most likely be found in the cork. Supposing, however, the apparatus to have been proved to be air-tight^ the potassa apparatus is once more inclined, and the combus- tion commenced. (g.) Management of the Combustion. A good supply of red- hot charcoal being provided close at band, that part of the tube included between the anterior part of the furnace and the first screen is quickly surrounded with small pieces of hot coal, so as to bring it speedily to a red-heat; the potassa in the large bulb will almost immediately be seen to recede, owing to the expan- sion of the heated air ; the screen is then gradually shifted backwards about an inch at a time, exposing a fresh portion of the tube, which, as before, is surrounded with red-hot charcoal ; as soon as the heat reaches the mixture of the organic sub- stance and the metallic oxide, the decomposition of the former commences, carbonic acid and aqueous vapour begin to be evolved, the latter is seen lining the interior of the front part of the chloride of calcium tube in the form of dew, while the former passes into the potassa apparatus, driving forward all the air present in the bulbs. At first, the bubbles appear to pass through the ley without being absorbed ; but as the com- bustion proceeds the absorption is more and more complete, and, after a time, if the substance under analysis be free from nitrogen, only a solitary bubble occasionally breaks through the liquid, nothing put pure carbonic acid passing through the chloride of calcium. The combustion should be so managed that the bubbles of gas shall follow each other in a uniform, uninterrupted, and tolerably rapid stream ; and the screen should be gradually moved backwards until the whole length of the tube is surrounded with red-hot charcoal. As soon as the evolution of gas in some degree slackens, the heat is raised by fanning the embers with a piece of pasteboard ; if the ope- ration has been managed according to the above directions, and if the mixture of the organic substance with the oxide of copper has been thoroughly made, the gas generally ceases suddenly. On this taking place, the cork is removed from under- neath the small side-bulb of the potassa apparatus, and the latter laid level on the cloth. The alkaline ley soon begins to recede, slowly at first, but, when it reaches the large bulb which is left at the end of the combustion full of carbonic acid, it rushes ELEMENTARY ANALYSIS OF ORGANIC BODIES. 541 rapidly into it ; there is no danger, however, of any of the liquid being carried into the chloride of calcium tube if the proper proportion ^between the bulbs has been observed in the construction of the, potassa apparatus. As soon as the operator sees the potassa rushing back into the large side-bulb, he clips oif the extremity of the tube with a pair of nippers ; this allows air to enter, and the alkaline liquor immediately recedes. The hot coals are now removed from the posterior part of the tube, a long tube open at both ends is held over the broken extre- mity, either by an assistant or supported by some convenient rest, and the operator, having inclined the potassa bulbs, places a suction tube on the end, and draws a current of air through the whole apparatus ; by this means all the carbonic acid and aqueous vapour which remained in the apparatus are swept into the chloride of calcium tube and potassa bulbs. The dis- position of the apparatus at this period is shown in Fig. 94. The analysis is now finished, and it only remains to take the apparatus apart, and, after the lapse of about half an hour, to weigh the potassa apparatus, and chloride of calcium tube, and calculate the results. (h.) Calculation of Results, The particulars of the analysis of mannite are as follows : Grains. Weight of the bottle and substance 83'76 Ditto of the bottle and residue 79*64 "Weight of substance analysed 4*12 Weight of the potassa apparatus after combustion 586*64 .Ditto before combustion 58O65 Weight of carbonic acid produced 5 -99 Weight of the chloride of calcium tube after combustion 276-06 Ditto before combustion 273 '20 Weight of water produced 2'86 These numbers furnish the data for calculating the per- centage composition of the substance. We have proved by the preliminary experiments that there is no nitrogen, sulphur, iphosphorus, or inorganic matter present; it is evident there- ifore that, after deducting from the weight of the mannite submitted to analysis the amount of carbon and hydrogen ob- itaiued in the forms of carbonic acid and water, the remainder 542 QUANTITATIVE ANALYSIS. must be calculated as oxygen. Let us see what quantities of these several elements we obtain as the result of our experi- ment. In the chloride of calcium tube we obtain 2*86 grains of water; how much hydrogen, then, in 4'12 grains of man- nite. Now, as every 9 grains of water contain 1 grain of hydrogen, we find the quantity contained in 1 grain by the proportion : 91 1 1 As 9 : 1 :: 1 : af t or y = , or9= , or #= ; that is, the amount of water obtained, divided by 9, gives the o.or> amount of hydrogen ; consequently, - !_=-3l777=the hydro- y gen in 4*12 grains of mannite ; but it is far more convenient to express the amount in percentage parts, thus : * = 7'71= percentage amount of hydrogen. y X 4<' \.2t Again, in the potassa apparatus we have condensed 5'99 grains of carbonic acid ; what is the percentage amount of carbon ? 22 grains of carbonic acid contain 6 grains of carbon. As 22 : 6 : : 11 : 3, As 11 : 3 : : 1 : a. #= ; consequently, the quantity of carbonic acid found, mul- tiplied by 3 and divided by 11, gives the amount of carbon ; and! we get it in percentage parts by multiplying by 300 and di- viding by 11 times the quantity of substance analysed, thus: cr.QQ y QHA _ ^. = 39-61 = percentage amount of carbon. The composition of the mannite, according to the above ana- lysis, is therefore Carbon 39'61 Hydrogen ...... 771 Oxygen. ...... 52*68 100-00 Construction of a Formula from the percentage composition a substance. Although we may not, as in the case of mannite, be able always to determine the atomic weight of a compound, it is an easy matter to ascertain the relative proportion of the equivalents of the elements of which it is composed, provided we have before us a correct statement of its percentage com- position ; in other words, we can, from such data, deduce its ELEMENTARY ANALYSIS OF ORGANIC BODIES. 543 empirical formula ; for this purpose we have only to divide the percentage number of each element by its atomic weight, and to reduce the figures obtained to the simplest expression pos- sible ; thus, in our experiment with mannite, we have = 6-601==equivalent of carbon. 7'71 = 7'710 = equivalent of hydrogen. 52 ' 68 = 6-585 = equivalent of oxygen. 8 Now here the numbers 6*601 and 6'585 are so nearly alike that it is at once evident that in mannite the number of atoms of carbon and of oxygen are the same ; and, on considering the above figures, it is seen that the simple numbers with which they best correspond are 6, 7, 6. Let us see, then, how nearly the numbers calculated in percentage parts from the formula C 6 H 7 6 agree with the experimental ones : Theory. Experiment. C 6 . . 36 . . 39-57 . . 39-61 H 7 . . 7 . . 7-69 , . 7-71 O 6 . . _48 . . 52-74 . . 52-68 91 . . 100-00 . . 100-00 The coincidence is so close that there can be no doubt that the above formula is the correct one, though we have no way of confirming it by determining the atomic weight of this peculiar kind of sugar. We shall give presently a case in which the absolute as well as the relative proportion of the atoms of an I organic substance can be determined, and a rational formula j deduced for it. Combustion with CJiromate of Lead. Some organic sub- i stances cannot be thoroughly oxidized by burning them with ' oxide of copper, partly, according to Dumas, in consequence ; of the formation of a small quantity of carbide of copper, i together with the reduced metal, and partly owing to the depo- ! sition here and there of small particles of carbon. "With such S substances this source of error is sometimes obviated by intro- ducing into the posterior end of the combustion tube some chlorate of potassa, mixed with oxide of copper; this salt, being heated at the end of the combustion, gives off pure . oxygen, in which all the combustible matter must be thoroughly oxidized ; care must be taken to heat the part of the tube con- 5M QUANTITATIVE ANALYSIS. taining the chlorate gradually, so as to avoid a too rapid evolu- tion of oxygen. The same object is attained by employing chromate of lead instead of oxide of copper. This substance is admirably adapted for oxidizing purposes, from its yielding up a portion of its oxygen by heat alone, a property not pos- sessed by oxide of copper. Chromate of lead is prepared by precipitating a clear solution of acetate of lead, containing slight excess of acetic acid, by bichromate of potassa, washing the precipitate thoroughly on a linen strainer, drying, and igniting to fusion in a Hessian crucible. The fused mass is pulverized, and kept in a stoppered bottle. By heat, either alone or in contact with an organic substance, it is decomposed into green oxide of chromium and basic chromate of lead. The combustion of an organic body, with chromate of lead, is conducted precisely in the same manner as with oxide of copper. The chromate is finely pulverized, and mixed while warm with the substance ; but, as it is not in the least degree hygroscopic, the process of exhausting the an aly sing-tube while buried in hot sand may be omitted ; a narrower tube than that employed with oxide of copper may be used, and the chromate which has served for one experiment may be used again after refusion. In the case of substances which burn with difficulty, anthra- cite coals for example, the chromate of lead may be advanta- geously mixed with aboiit one-tenth of its weight of fused and pulverized bichromate of potassa. Such substances are how- ever best analysed by burning them with oxide of copper in a current of dry oxygen gas. Mene (' Comptes Eendus,' Ivi. 1033) suggests the substitu- tion of melted chlorate of potassa for oxide of copper in the analysis of organic substances, by which he thinks that the pro- cess is rendered much easier, and the drawbacks arising from the frequent incomplete combustion of the substance arising from insufficient oxygenation and other drawbacks are entirely obviated. His process is as follows : Into the end of a closed tube of white glass, about 20 inches long, 0*39 inch wide, and 0'039 inch thick, a quantity of chlorate of potassa (melted and pounded) is introduced equal to about three-quarters of an inch ; the mixture of the organic substance (about 15 grains) with sufficient chlorate nearly to fill the tube is then poured in in the usual manner ; when full, the tube is closed with an ami- anthus plug, then with a small cork traversed by a small tube for the disengagement of the gases. The tube thus prepared ELEMENTARY ANALYSIS OF ORGANIC BODIES. 545 is suspended by two iron wires, so that it may be free, and rwithin the reach of the operator ; then by means of an india- ; rubber tube it is put in communication with the potassa appara- itus and chloride of calcium tube in the ordinary manner. An ordinary spirit-lamp is placed under the tube, beginning to beat near the amianthus plug. The chlorate of potassa melts almost immediately, burning the matter and forming carbonic acid and water, which are disengaged quietly. Though in many cases (when the substance contains much carbon), igni- tion, and sometimes even deflagration accompanies combustion, yet there is no danger. When the substance is burnt in the place first heated, the lamp is shifted, so as successively to heat the whole length of the tube. The chlorate at the closed end of the tube is heated last, so as to disengage oxygen, and thus bake with it all the analysable gases which may remain in the tube. When the operation is presumed to be at an end, the bube is detached from the estimating-apparatus, and all the chloride of potassium is eliminated by washing to make the tube ready for a fresh analysis. An analysis of this kind oc- cupies twenty minutes, and the operation may be carried on at the will of the operator. (2.) Analysis of volatile compounds (solid and liquid) not containing Nitrogen. The analysis of waxy substances, and of volatile organic solids, is performed in tubes which are some- what longer and wider than those generally employed. The analysing-tube is filled also in a different manner, as it would Obviously be inadmissible to triturate volatile substances with hot oxide of copper, and the nature of oily and waxy ?bodies renders their admixture in the usual man- !ner impracticable. Solid fusible fats, oils, etc., are (weighed in a little tube, the size and shape of Fig. 95. iThe chromate of lead or oxide of copper is introduced, Iwhile hot, into a wide thin glass tube, about twenty inches long, or into a copper tube which can be Closed air-tight with a brass cap and lead washer, Fig. 96 ; in this it is allowed to cool ; if chromate pf lead be the oxidizing agent employed, a layer of jibout two inches is first introduced into the combustion tube ; ibut, if oxide of copper is to be used, the same quantity of a mixture of chlorate of potassa and oxide is shaken in ; the little tube, with its charge of substance, is then dropped into ;the analysing-tube, and heat applied so as to melt it, and cause 546 QUANTITATIVE ANALYSIS. it to spread over the lower side ; the tube is then filled with the oxidizing agent by introducing its open end within the copper or wide glass tube, containing the oxide of copper or chromate of lead, and turning it round. The tube being filled, the substance to be analysed is diffused through the oxide or chromate by placing it for a few minutes in hot sand, keeping it well corked during the time ; the combustion is then proceeded with in the usual manner. If the substance is not of a waxy or fatty nature, it is mixed with the oxide or chromate by means of a copper wire, the end of which is formed like a corkscrew ; but the operation requires to be conducted with skill and rapidity. Volatile liquids, such as alcohol, ether, etc., are weighed in two or three small glass bulbs, the size and shape shown in Fig. 97. These bulbs should be drawn out of a long piece of quill tubing, in order to prevent any moisture from getting into them while they are.; being expanded by the breath. They are filled by heating them slightly over a spirit- lamp, and immersing their pointed extremi- ties in the fluid to be analysed ; on cooling, the liquid enters the bulb from the contrac- tion of the heated air, the bulb is again gently heated, on which more air is expelled, and then, on once more dipping the open end into the liquid, an additional quantity Fig 96 eil ^ ers - I n ^ s manner the bulb is about ' three parts filled with the liquid ; upon which its extremity is sealed, and it is weighed, arid, the weight of the empty bulb having been pre- viously taken, that of the included liquid is found. The analy sing-tube employed should be longer and wider in proportion as the liquid is more vola- tile. The oxidizing agent is allowed to cool in the closed copper or glass tube, 1'ig. 96, and, a layer of two inches having been introduced into the com- bustion tube, the centre of the neck of one of the ^. little bulbs is slightly scratched with a file, the point is quickly broken off, and both bulb and point are dropped into the tube and covered immediately with three or four inches of oxide of copper ; another bulb is then introduced in a ELEMENTAEY ANALYSIS OF ORGANIC BODIES. 547 similar manner, covered with a layer of oxide, and so on with ^a third, should the substance to be analysed be included in more than two bulbs. The tube is then tapped to open a channel for the escape of the gases. The analysis of volatile liquids requires a great deal of care |on the part of the operator. Thus, if a high heat were sud- denly applied to one of the bulbs, a sudden gush of vapour [would be produced, a portion of which might either be expelled from the tube without undergoing combustion, or, if it were resolved into carbonic acid and water, these products would be evolved with such impetuosity as to eject the potassa from the bulbs. The greatest care is therefore necessary in conducting the operation. About six inches of the anterior portion of the tube are first heated to redness, and a screen placed behind to prevent the communication of heat to the locality of the bulbs ; [the fluid in the first bulb is then gently distilled out by hold- Hng a red-hot charcoal near it, and the vapour, thus gradually [coming into contact with the oxidizing agent, is quietly burned. fin this way, the contents of all the bulbs are brought into seontact with the heated oxide, and it is advisable to keep a few i pieces of hot charcoal against the closed end of the tube, to iprevent any of the liquid from distilling back and condensing jbhere. (3.) Analysis of organic substances containing Nitrogen. The determination of the nitrogen of an organic compound is Always an operation distinct from that of the determination of ^he carbon and hydrogen. It is necessary, however, to modify |.n some degree the arrangements for the latter process, in the Jinalysis of substances containing nitrogen, or the number for : ;he carbon would always come out too high. The reason of :j;his is that when an organic compound containing nitrogen ;|S burnt with oxide of copper or chromate of lead, a portion of i;he nitrogen becomes converted into nitric oxide gas, which, on ;oming into contact with the potassa in the bulbs, is further oxidized into nitrous acid, and is retained by the alkali. It is lecessary, therefore, to decompose these oxygen compounds of litrogen before they can leave the combustion tube ; which >bject is perfectly attained by introducing a layer of four or ;ive inches of fine and pure copper-turnings into the anterior part |>f the tube and keeping them strongly ignited during the whole analysis. The nitric oxide and nitrous acid are completely de- mposed by copper at a high temperature, nitrogen being 548 QUANTITATIVE ANALYSIS. evolved in a gaseous form, which passes unabsorbed through the chloride of calcium tube and potassa apparatus. The copper turnings employed for this purpose must be free from impurities, to ensure which it is advisable to tarnish them by heating them in a crucible, with free access of air, and then to reduce the film of oxide which has formed on their surface by heating them to redness in a glass tube, through which a current of hydrogen is passing. There are three different ways of estimating the nitrogen of an organic substance, namely, 1st, by collecting the gas in a free state and measuring its volume ; 2nd, by determining the relative proportion between the evolved carbonic acid and nitrogen gases ; and 3rd, by converting the nitrogen into am- monia, and estimating it as such. (a.) Estimation of Nitrogen in a gaseous state. This is the method of Dumas, and is applicable to the analysis of all nitrogenous substances, in whatever form that element may exist in them. The substance is burned in the usual way with oxide of copper or chromate of lead ; a layer of reduced copper- turnings being placed in the anterior portion of the tube, which should be at least two feet long and not drawn out at the closed extremity, but merely sealed ; and from five to eight inches of dry bicarbonate of soda at the posterior end, the object of which is to enable the operator to expel all the at- mospheric air from the analy sing-tube at the commencement of the operation, and all the nitrogen from it at the end by a current of carbonic acid. t The mixture of the substance with the oxidizing agent is made in the manner already de- scribed, and the tube filled as shown in Pig. 98 ; thus, from a to b, bicarbonate of soda ; from b to c, pure oxide or chromate ; Fig. 98. from c to d, the intimate mixture of the substance with the oxidizing agent ; from d to e, rinsings of the mortar ; from e ELEMENTARY ANALYSIS OF ORGANIC BODIES. 549 to j^ pure oxide or chromate ; and fro mf to g, copper turnings. The tube is arranged in the furnace, as shown in Fig. 99, and, the first object being to expel the whole of the atmospheric air, a screen is placed about three inches from the end, and hot coals are applied round the tube before it, so as to heat intensely the bicarbonate of soda, and fill the whole tube with carbonic acid. The conducting tube c d, fitted into the combustion tube a b, with a good sound cork, passes under the surface of mercury contained in a small pneumatic trough T. The bubbles of gas, Fig. 99. they break through, are received from time to time into a ; test-tube filled with caustic potassa ; and, when they are com- pletely absorbed, it may be concluded that the atmospheric air I has been thoroughly expelled from the apparatus. As the j success of the experiment depends in a great measure on thus I perfectly driving off the air, the operator must not hurry this !i preliminary process, but continue to apply heat to the bicarbo- nate of soda till a dozen or so bubbles following are completelv absorbed by the caustic ley. A stout glass cylinder c. ver'v accurately graduated into cubic centimetres, or hundredths of a cubic inch, and of the capacity of about 200 c. c. or 13 c. i., is now filled two-thirds with mercury and one-third with potassa ley and inverted in the pneumatic trough immediately over the conducting-tube a;* the combustion is then proceeded * As caustic potassa acts on the cuticle, making the skin slippery, the best way of filling the cylinder is first to pour in the mercury, then the requisite quantity of potassa ley, and lastly pure water; the jar is then inverted by laying on its edge (which should be ground perfectly smooth) a ground glass plate. If the fingers are allowed to become soiled with the alkali, there will be a danger of the tube slipping through them when the operator is in the act of inverting it in the pneumatic trough. PART II. 2 O 550 QUANTITATIVE ANALYSIS. with in the ordinary manner, the carbonic acid yielded by the decomposition of the substance is absorbed by the ley, while the nitrogen passes into the cylinder in its gaseous form. At the end of the operation, the remainder of the bicarbonate of soda is strongly ignited, and the carbonic acid evolved sweeps the whole of the nitrogen still contained in the tube into the graduated cylinder, while it is itself absorbed by the potassa. The analysis being concluded, the cylinder containing the nitrogen is allowed to remain at rest for some time, to ensure the complete absorption of the carbonic acid ; it is then care- fully transferred to a large deep glass vessel containing water, by placing underneath its edge the ground glass plate, on removing which under the water the mercury sinks to the bottom of the vessel, its place being supplied by the water. The water within and without the cylinder is then brought to the same level, and the volume of gas read off, the state of the barometer and thermometer being at the same time noted. The analysis is now concluded, and it only remains to calculate the results. The method of doing this will be best understood by quoting the results of an actual experiment. Weight of substance submitted to analysis, '600 grm.= 9-264 grains. Volume of nitrogen gas collected, 41 '5 cubic centimetres = 2-531 cubic inches ; thermometer, 1S'5 Cent. = 65'3 F.; barom. 7573 millimetres = 29'9 inches. Correction of the Gas for temperature. It is assumed that gases when heated from the freezing to the boiling-point of water, that is, from to 100 of Celsius's thermometer, ex- fVQ Af*^C pand 0'3665, or, in other words, that they expand - _LUu = 0-00366 for every degree of the Centigrade scale ; to find, therefore, what volume the 41'5 cub. cent, of nitrogen observed at 18'5 would occupy at 0, we must multiply the number of degrees above by the expansion coefficient for one degree, add the product to 1, and divide the observed volume by this sum, thus : lume the 0-00366) gas would occupy at Cent. This volume has now to be ELEMENTARY ANALYSIS OF ORGANIC BODIES. 551 corrected for pressure ; but before doing so, it is necessary to take into consideration the tension of the aqueous vapour at the observed temperature : for since the gas was measured in its moist state it must have contained a quantity of vapour of water diffused through it proportional to the temperature, which would counterbalance part of the column of air pressing on the gas ; this portion must, therefore, be subtracted from the observed height of the barometer, in order to arrive at the true pressure. In the Appendix will be found a table (ex- tracted from Fresenius's 'Quantitative Analysis'), in which the force of the aqueous vapour at every degree of Celsius from to 40 is expressed in millimetres. Referring to this table, it will be found that at the temperature at which the experiment we are now considering was made, viz. 18*5 Cent., the tension of the aqueous vapour amounts to 15 '85 milli- metres ; this, therefore, being deducted from 757*3 millimetres, the observed height of the barometer, gives 741-45 millimetres as the true pressure from which to correct the volume of gas. Correction of the Gas for pressure. The normal pressure is : assumed to be 760 millimetres (29-9 inches), therefore ( 760 : 741-45 : : 38'87 : 37'92 cubic centimetres = the volume j of nitrogen corrected for temperature, pressure, and aqueous vapour. The next question is What is the weight of 37 '92 cubic centimetres of dry nitro- gen gas at Cent, and 760' millimetres barom. ? The specific gravity of nitrogen is 0'9706, atmospheric air | being unity ; aud it has been ascertained by careful experi- f ments that 1 litre (1000 cubic centimetres", or 61 -03 cubic inches) weighs at Cent, and 0*76 metre barom. 1*2609 gramme (19-46 grains), therefore: As 1000 : 1-2609 : : 37'92 : x. x -0478 grm. (0'738 grains) = the weight of the nitrogen yielded by 0'600 grm. of the substance at the standard tem- perature and pressure ; and As -600 : -0478 : : 100 : 7'96 = the percentage amount of nitrogen in the substance analysed. Simpson (Annal. d. Chem. u. Pharm. 95, 74) modifies this process by burning the substance with a mixture of oxide of copper and oxide of mercury, and sweeping the air out of the tube by a current of carbonic acid, obtained by heating carbo- nate of protoxide of manganese placed in the posterior end of 2o2 552 QUANTITATIVE ANALYSIS. the combustion tube, instead of bicarbonate of soda. The mixed gases are collected in a pear-shaped receiver of peculiar construction, filled with and inverted over mercury ; a little strong solution of caustic potassa being thrown in to absorb the carbonic acid ; when the combustion is finished the re- ceiver is detached from the aualysing-tube, and the nitrogen is forced into a graduated receiver by the pressure of a column of mercury. Having thus detailed the general principles of the method of deducing the percentage amount of nitrogen from the volume of gas obtained, we give a formula founded on the above considerations, by which the calculation may be made in a few minutes, and a good deal of trouble saved. Let v be the observed volume of gas. B the height of the barometer. p the tension of aqueous vapour. t the observed temperature. m the weight of the substance analysed. Then, 0-12609 v (B-p) m 0-76 (1 + 0-003665 t) 0-1659p(B-p) m (1 + 0-0036650" the percentage amount of nitrogen. If the analysis quoted above as an illustration be calculated according to this formula, the result will be found to be the same as that at which we arrived by the more circuitous method. (b.) Estimation of Nitrogen by determining the Relative Pro- portion between the Carbonic Acid and Nitrogen yielded by the Decomposition of the Substance (Liebig's method). This me- thod, which is applicable only to the analysis of substances rich in nitrogen, is performed by burning the substance in the usual way with oxide of copper in a long tube not drawn out at the closed end, and containing at least 5 inches of reduced copper- turnings at its anterior end. After a considerable portion of the substance has been decomposed, and when it is concluded that all the atmospheric air has been expelled from the tube, the evolved gases are collected over mercury in a series of gra- duated tubes, about a foot long, and half an inch wide. These tubes are then transferred, one at a time, to a tall cylinder (Fig. 100) containing mercury, where the united volumes of carbonic ELEMENTARY ANALYSIS OF ORGANIC BODIES. 553 acid and nitrogen are accurately measured, the mercury within and without the tubes being brought to the same level : a small quantity of potassa ley is then thrown up into the tubes from the siphon (Fig. 101) by gently blowing, the bent extremity be- ing introduced at the open end of the tubes ; and after the absorption of the carbonic acid is complete, the residual gas is again measured, the mean result of all the tubes is taken, and the volume of the carbonic acid relative to that of thenitrogengas,expressesimmediately, without any further calculation, thepro- portion of the equivalents of carbon to those of nitrogen, since one equivalent of carbon is consumed by two equiva- lents of oxygen, without the volume of the latter undergoing any alteration, yielding therefore two volumes of car- bonic acid ; and one equivalent of ni- trogen yields also two volumes of ni- trogen gas. The amount of carbon must have been ascertained by a previous experiment ; an( j j ^ at b e i n g known, the weight of the ni- trogen may be calculated from the atomic weights of carbonic acid and nitrogen: thus, suppose the proportion of carbonic acid to nitrogen to be as 10 to 4, then the compound analysed contains 10 x 6 = 60 carbon to 4 x 14 = 56 nitrogen. This is the proportion in which these elements exist in uric acid, for the analysis of which this method answers very well ; but in cases where the proportion of nitrogen is very small, or where it is evolved in a variable proportion at different stages of the analysis, it is inapplicable, and the method before de-- scribed, or the one we shall next consider must be adopted. (c.) Bunsen's method. A stout tube of difficultly fusible glass about 14 inches long, and from half to three-fourths of an inch internal diameter, is drawn out at one end, and then thickened, but not closed in the blowpipe flame. From half to three- fourths of a grain of the substance is well mixed with about 100 grains of oxide of copper, and a small quantity of clean 101> Fig. 100. 554 QUANTITATIVE ANALYSIS. copper turnings, and the mixture is introduced into the tube, the other end of which is then drawn out so as to allow a space of at least seven inches from ~b to c. The tube, thus prepared and filled, has the form shown in Eig. 102. The end a is con- Fig. 102. nected with an apparatus for generating hydrogen gas, and the end d with an exhausting syringe. The gas, previously well dried by passing through strong sulphuric acid, is allowed to flow through the tube until all atmospheric air has been ex- pelled, and replaced by hydrogen gas, which is then rarified by raising the piston of the exhausting syringe, and the tube is im- mediately closed at b by the blowpipe flame. It is then exhausted as completely as possible, and the end d closed by the blowpipe flame at c. To decompose the substance, the tube is now buried in freshly -prepared gypsum paste laid in a strong iron mould made in two parts, which can be well secured together with bolts, and in which several holes are bored, to allow free issue to the aqueous vapour. As soon as the gypsum has thoroughly set, the mould is slowly heated to bright rednesss in a suita- ble furnace, and kept at that temperature for about half an hour, it is then allowed to cool, after which the mould is taken apart, the tube is carefully removed from its plaster bed, and the point broken off under a graduated tube filled with and in- verted over mercury ; a drop of water is previously introduced into the tube, to saturate the moist gas with aqueous vapour. The volume of the gas being taken, due regard being had to the state of the barometer and thermometer, a ball of moist- ened hydrate of potassa attached to a platinum wire is intro- duced into the tube, by which the carbonic acid is absorbed ; the residual nitrogen gas being dried, by the introduction of a second unmoistened ball of hydrate of potassa, its volume is measured, the volume of both gases being reduced to the same tempera- ture, pressure, and dryness, their relative volumes are ascer- tained, and also the relation of the equivalents of carbon to nitrogen in the substance analysed. This method requires practice and dexterity in the manipulation, but it is, notwith- standing the smallness of the quantity operated upon, suscep- tible of a high degree of accuracy. ELEMENTARY ANALYSIS OF ORGANIC BODIES. 555 (d.) Estimation of Nitrogen by converting it into Ammonia. This method, which was invented by Drs. Will and Varren- trap, and which is founded on the action of the hydrates of the alkalies on nitrogenous compounds at high tempera- tures, has already been described (page 273). This pro- cess, though accurate, is tedious, on account of the necessity for evaporating the platinum salt on the water-bath, and it is expensive, on account of the alcohol and ether which are required. An important modification of the method was, in consequence of these objections, devised by Peligot ('Comptes Eendus,' March 29, 1847). The combustion is executed in the same manner as in Will and Varrentrap's method, and the ammo- nia is condensed in a similar bulb apparatus (see annexed cut, Fig. 103), but which, instead of containing hydrochloric acid, is Fig. 103. irtly filled with a fixed weight or volume of sulphuric acid of known strength. Now, as the ammonia which combines with the acid lowers its strength, by determining after the combus- tion the composition of this liquid, and comparing it with that it previously possessed, it is easy to ascertain the quantity of am- monia that has been condensed, and consequently the amount of nitrogen yielded by the substance submitted to analysis. This operation is executed with as much rapidity as precision, by means of an alkaline solution, likewise of known strength. Peligot recommends a solution of lime in syrup. By triturating slaked lime with a solution of sugar, it dissolves in far greater propor- tion than in pure water, and the compound formed, possesses the same alkaline reaction as if the base which it contains were free. The operation is conducted as follows : A certain quantity of sulphuric acid of known strength is accurately measured into the condensing apparatus. The acid employed by Peligot contains 61*250 grammes of S0 3 HO to 1 litre of water; consequently, 100 cubic centimetres corre- 556 QUANTITATIVE ANALYSIS. spond to 2'12 grammes of ammonia, or 1/75 gramme of nitro- gen. The exact strength of the acid is determined by precipi- tating an accurately-measured volume with chloride of barium, and determining with the greatest care the quantity of sulphate of baryta produced. "When the combustion is finished, the acid which has served to condense the ammonia is poured into a cylindrical glass, the apparatus carefully washed, and to this liquid, diluted with much water, a few drops of litmus are added to give it a red colour. "By means of the solution of lime in sugar, which is contained in a burette graduated in cubic centimetres and in tenths of a cubic centimetre, the acid liquid is accurately saturated, being guided by the blue colour which the liquor suddenly acquires when the point of saturation is attained. The quantity of alka- line liquor required to produce this effect is read off on the burette ; and as the amount of the lime compound which satu- rates 10 cubic centimetres of the normal solution of sulphuric acid has been determined by a previous experiment, by sub- tracting from this quantity that found for the acid which has absorbed the ammonia from the nitrogenous substance, we find the volume of the acid solution which has been saturated by the ammonia, and consequently the weight of the nitrogen which the body contained. Shortly after the publication of this method, some experi- ments to test its accuracy were undertaken by the author (Chem. Gaz., vol. v. p. 411). The substance experimented upon was vitelline, a nitrogenous principle extracted from the yolk of the egg. The standard sulphuric acid employed was of the specific gravity 1'0504 ; the quantity of dry acid which a small bottle of about one cubic inch capacity contained, was very ac- curately determined by precipitation with chloride of barium, to be 15-8 grains, which agrees exactly with the theoretical amount calculated from the specific gravity by Tire's table. By a simple calculation it was ascertained that this quantity of dry acid is equivalent to 6'715 grains of ammonia, or to 5'53 grains of nitro- gen. The solution of lime in sugar was of such a strength that it required 54'25 divisions of the burette to neutralize exactly the bottle full of standard acid. The condensing apparatus employed was larger than it is usually made : it had three bulbs instead of two, and held about two-thirds of the contents of the bottle without being sufficiently full to excite any apprehension of the rushing back of the acid into the combustion tube at the close of the operation. ELEMENTARY ANALYSIS OF ORGANIC BODIES. 557 "Example. 7*53 grains of vitelline were burned with soda- lime, and the ammonia condensed in the standard acid, which, after the combustion was finished, was transferred to a beaker, and the apparatus rinsed out repeatedly with distilled water ; the remainder of the acid in the bottle was then added, the whole diluted to about half a pint, and a few drops of litmus added ; 45 burette measures of the sugar-lime were required to saturate the acid, 9*25 measures (54*2545) therefore repre- sent the quantity of sugar-lime equivalent to the ammonia given off by the substance ; a simple calculation shows this to be equal to O941 grains of nitrogen, or 12*5 per cent. In a second experiment, in which 10*17 grains were burned, 42 measures of the sugar-lime were required to saturate the acid, which gives 12*25 as the representative of the sulphuric acid condensed by the ammonia, and 1'246 grains as the amount of nitrogen = 12*35 per cent. In a third experiment a different standard acid was employed, and a fresh preparation of vitelline analysed. The acid was equivalent to 4'004 grains of ammonia, or 3*297 grains of nitro- gen ; and it required for neutralization 32 divisions of sugar- lime, which was also a fresh preparation. 7*24 grains were burned; after the combustion 23*25 mea- sures of sugar-lime were required to saturate the acid; 32 23*25 = 8*75 measures represent, therefore, the sulphuric acid neutralized by the ammonia; this gives 0*9016 for the nitrogen in the substance, or 12*45 per cent. These three experiments agree, it will be observed, very closely with each other. Two experiments, in which the ni- trogen was estimated as double chloride of platinum and ammo- nium, gave 13*02 and 12*6 per cent, of nitrogen, numbers which are rather higher than those obtained by the new method ; in- deed, from the very principles of Peligot's process, the results must always be a trifle below the truth, it being always neces- sary to add a slight excess of the alkaline solution, in order to see, by the change of the liquid from red to blue, when neutra- lization has taken place. It is true that when we are experi- menting with a perfectly colourless acid, the smallest addition of the lime solution above that required for the perfect neutra- lization of the acid immediately determines the blue colour : but, after the acid has served to condense the ammonia pro- duced by the combustion of an animal substance with soda-lime, it is always more or less coloured ; nor can a considerable tinge 558 QUANTITATIVE ANALYSIS. of yellow be prevented, although it may be much diminished, by using a long combustion-tube and a great excess of soda-lime : this colour renders it somewhat difficult to hit the precise mo- j ment when neutralization takes place; if the sugar-lime be added i till the liquor becomes distinctly blue, a deficiency of nitrogen amounting nearly to 1 per cent, may easily occur, and it is only practice that will enable the operator to stop exactly at the proper time. The best practical rule is, to use a long combus- tion-tube, to keep its anterior portion very hot, and to dilute the acid considerably with water previous to adding the alkaline solution. One great advantage which the method of Peligot possesses over all others, is the saving of time which it effects, as by it, a nitrogen determination may be executed in less than half an hour with as much accuracy as by the other methods which require at least three hours. This process is equally applicable to the analysis of substances which are not very rich in nitrogen, as wheat, vegetable soil, etc., provided a sufficient quantity of the substance be employed. It has been objected that the sugar solution does not keep, but if air be perfectly excluded it remains clear for a considerable length of time, it may however be replaced by a standard solution of soda, which should be perfectly free from carbonic acid. For the estimation of nitrogen in manures, and for commercial purposes, where great accuracy is not required, the ammonia produced by burning the substance with soda-lime may be con- densed in a solution of tartaric acid in strong alcohol. The tube conveying the ammonia must not be allowed to dip into the liquid, the receiver containing which, should be provided with a safety tube and connected with a second. The ammonia is collected, and precipitated in the form of bitartrate of ammonia, a salt which is quite insoluble in strong alcohol ; it is washed with alcohol on to a tared filter, and dried at about 180 F. ; 100 parts of the dried salt contain 10-2 parts of am- monia = 8*4 of nitrogen. Should the preliminary examination of the organic compound have indicated the presence of chlorine, it must be burned with chromate of lead, since, if oxide of copper were employed, a portion of subchloride of that metal would form, and condense in the chloride of calcium tube ; and should sulphur be present, a tube of perfectly dry peroxide of lead must be included be- tween the chloride of calcium tube and the potassa bulbs, in ELEMENTARY ANALYSIS OF ORGANIC BODIES. 559 order to convert into sulphuric acid, and retain as sulphate of lead all the sulphurous acid which may have formed from the oxidation of the sulphur. The amount of chlorine in chlorinated compounds is deter- mined by ignitiog them with soda-lime, or by transmitting their vapour through soda-lime, should they be of a volatile nature. The ignited mass is dissolved in dilute nitric acid, and the chlo- rine determined as chloride of silver. When the organic com- pound is of an acid nature, Lowig dissolves it in dilute potassa ley, evaporates the solution to dryness, and ignites the residue, by which the chlorine present is converted into a metallic chloride. 207. Determination of the Rational Formula and Atomic Weight of an Organic Substance : (1.) Of an Organic Acid. An excellent method is to con- vert it into a silver salt, and to determine accurately the amount of metal which it contains. Example. Erom O320 grm. of a certain crystalline sub- stance having acid properties, there are obtained O825 grm. of carbonic acid, and 0'1715 grm. of water. Another portion of the acid is exactly neutralized with ammonia, and the solu- tion mixed with nitrate of silver. A white curdy precipitate falls, which is washed on a filter with cold water, and purified by two or three crystallizations out of boiling water ; O384 grm. of this new substance, being ignited, yields 0'1485 grm. of pure silver. These numbers calculated centesimally give the following as the composition of the acid, and salt : Acid. Carbon 70-31 Hydrogen . 5'95 Oxygen 23-74 Hydrated acid lOO'OO Salt. Oxide of silver 47'73 Anhydrous acid 52'27 Silver salt lOO'OO ISTow, the atomic weight of oxide of silver is 116, therefore 560 QUANTITATIVE ANALYSIS. As 4773 : 52-27 : : 116 : a?, x = 127-03, the atomic weight of the anhydrous acid, and 127-03 4- 9 = 136-03, the atomic weight of the hydrated acid ; and, to deduce 'from this number the formula of the hydrated acid, we have, first, the proportions As 100 : 70-31 : : 136'03 : a?; x = 95'64 As 100 : 5-95 : : 136'03 : a?; x 8'09 As 100 : 23-74 : : 136'03 : x\ x = 32'29 Then by dividing each of these numbers respectively by the atomic weights of carbon, hydrogen, and oxygen, we have = 15-94 for the carbon, b = 8-09 for the hydrogen, qq.oq - = 4-036 for the oxygen, 8 numbers which evidently correspond closely with the formula C 16 H 8 4 , which is, therefore, the rational formula of the acid in question ; and which, when calculated in percentage parts, agrees almost exactly with that determined bv experiment. Thus Calculation. Experiment. C . . 70-58 ........ 70-31 H. . 5-88 ........ 5-95 O . . 23-54 ........ 23-74 100-00 100-00 ' (2.) Of an Organic Base. The atomic weights and formula? of these interesting bodies are best determined by analysing with great accuracy the insoluble double salts which they form with bichloride of platinum. Example. Hofmann and Muspratt obtained by acting on nitrotoluide with sulphuretted hydrogen a crystalline body having basic properties, 0'1955 grm. of which, by combustion with oxide of copper, gave 0*5630 grm. of carbonic acid and 0-1515 grm. of water. A double salt of the hydrochlorate of this base and platinum was formed, 0'4337 grm. of which gave, by cautious ignition, 0'1372 grm. of platinum. Now there is every reason for supposing that this double salt is analogous ELEMENTARY ANALYSIS OF ORGANIC BODIES. 561 in composition to the double chloride of platinum and am- monium if so, its atomic weight is found from the proportion As 0-1372 : 04337 : : 98'68 (the atomic weight of platinum) : x, as = 31 1 1 93, and by deducting from this number the sum. of the atomic weights of one equivalent of bichloride of platinum (169'68) and one equivalent of hydrochloric acid (36'5), we get the atomic weight of the new base, 311-93 206-18 = 105-75. The percentage composition of the base, as found experi- mentally, is Carbon .......... 78'53 Hydrogen ......... 8'61 Nitrogen ......... 12-86 100-00 from which numbers its empirical formula may be deduced thus : As 6 : 78-53 : : 1 : #; x = 13'09 As 1 : 861 : : 1 : x; x 8'61 As 14 : 12-86 : : 1 : x; x = 0'918 or C 14 H 9 N, which agrees best with the experimental numbers. Thus : C 14 ...... 84 .... 78-50 H 9 ...... 9 .... 8-41 N ...... JL4 . . . . 13-09 107 100-00 The theoretical number 107 agrees so well with the experi- mental number 105*75, that there can be no doubt that C 14 H 9 N is the correct formula for the new substance. Gas as Fuel in Organic Analysis. For the admirable com- bustion-furnace in which the source of heat is a mixture of coal gas and atmospheric air, shown in Fig. 104, we are indebted toHofmann, by whom it is thus described (Journ. Chem. Soc., vol. xi. p. 30) : " Some years ago, a peculiar clay burner was introduced under the name of atmopyre. This burner, shown in Fig. 105, is a hollow cylinder of burnt clay, closed at the top, open at the bottom, and with numerous perforations through the sides. Those which I use are 3 inches high, and of |- inch exterior, and \ inch interior diameter. The perforations, of about the QUANTITATIVE ANALYSIS. ELEMENTARY ANALYSIS OF ORGANIC BODIES. 563 thickness of a pin, are made in rows ; their number varies, those which I employ have ten rows, each of fifteen holes. Prom such a clay-cylinder, fixed upon an ordinary bat's-wing burner, the stopcock of which has been appropriately adjusted, the gas burns with a perfectly blue flame, which envelopes the cylinder and renders it in a short time incandescent. These clay-burners are readily obtainable and very indestructible, and by appropri- ately grouping together a number of them, a com- plete combustion-furnace for organic analysis may be arranged. " The disposition of the apparatus being obvious from the Fig. 104, a few explanatory remarks will be suf-Fig- ficient. Into a brass tube from 3 feet to 3 feet 8 inches in length (shown in section in Fig. 106) which communicates at both ends with the gas main of the laboratory, there are screwed from twenty to thirty tubes. These tubes, I inch thick, and 7 inches high, are provided with stopcocks, and carry brackets of 4 inches length and -f inch diameter for the reception of five ordinary bat's-wing burners (each con- suming from 3 to 4 cubic feet of gas per hour for a full luminous effect), upon which a corresponding number of clay- burners is fixed. Three of these burn- ers are 3 inches high, the other two are only 1-|- inches high, and have only ri g- 106 - seventy or eighty perforations ; supports for two combustion- tubes, side by side, are thus provided, which are bedded in channels of heated fire-clay. The system of brackets lying side by side acquires sufficient stability by a strong iron frame which rests upon two firm supports of cast-iron. The iron frame has, moreover, a groove for the reception of move- able side plates of fire-clay ; they are of the same height as the H high burners, over which they project by about |. inch in con- k sequence of their resting upon the iron frame ; these side plates (which in Fig. 106 are not drawn quite high enough) are in- tended as supports to covering plates of fire-clay so that during a combustion the whole or part of the burners may be enclosed. " It deserves to be noticed that the efficiency of the furnace 564 QUANTITATIVE ANALYSIS. essentially depends upon the correct disposition of the gas-jets. The most appropriate space between the several burners, ac- cording to numerous experiments made for the purpose, is about i inch. It is very important for the attainment of a perfectly uniform temperature, that the several brackets bearing the burners should be equidistant ; their position is therefore spe- cially secured by every bracket being fixed in an aperture formed in the iron frame. " The use of the furnace scarcely requires any special remark. According to the length of the combustion tube, from eight to ten stopcocks (under all circumstances the largest possible number) are opened at once at the commencement of a com- bustion. If care has been taken to regulate the amount of gas, either by the stopcocks in the horizontal gas-pipe, or by those in the separate supply tubes, the lighted portion of the furnace in ten or twelve minutes will be in a state of perfect incandes- cence, comparable only to the ignited mass of charcoal in an ordinary combustion-furnace. After this, it is only necessary to open the remainder of the stopcocks in appropriate suc- cession to ensure a slow and regularly progressing combustion. The time required for the completion of an analysis varies from forty minutes to an hour. " The heat obtained by this furnace is extremely uniform, and since it is conveyed to the combustion-tube chiefly by radiation from the incandescent mass of surrounding clay, every part of the tube is equally heated. The temperature which it is ca- pable of yielding is entirely at the command of the operator; when strained to its full power, it gives a heat equal to that of the strongest charcoal combustion-furnace. By appropriately adjusting the stopcocks, however, it is possible to maintain the furnace at any desired temperature." "With regard to the quantity of gas consumed by this furnace, Hofmann found that a combustion, lasting one hour, and re- quiring the whole length of the furnace (34 rows of burners) consumes from 80 to 90 cubic feet. For a carbon determina- tion with 24 rows of burners, which generally lasts about 40 minutes, from 50 to GO cubic feet are required, and for a nitro- gen deter minati on from 25 to 30 cubic feet. 565 PART III. SPECIAL METHODS CHAPTEE X. ON THE ANALYSIS OF MINERAL WATEES. 208. Qualitative Analysis. The principal constituents are detected qualitatively in a brief and simple manner ; the rarer substances are most conveniently found in the quantitative analysis. i. Detection of Carbonic Acid. Add to a portion of the water which has been recently collected and kept in a well- stopped bottle, a small quantity of lime water ; if a precipitate be formed, carbonic acid is present. ii. Detection of Sulphuretted Hydrogen. The odour of the water is, in most cases, a sufficient evidence of the presence of this gas, which is confirmed by adding to a small flask, nearly filled with the water, a few drops of solution of acetate of lead, and well agitating ; a black precipitate, or a more or less brown coloration of the fluid, indicates sulphuretted hydrogen, or an alkaline sulphide. in. Detection of Sulphuric Acid. A portion of the water is slightly acidified with hydrochloric acid and gently warmed, chloride of barium is then added ; the formation of a white pre- cipitate indicates sulphuric acid. iv. Detection of Chlorine. A portion of the water is acidified with nitric acid, warmed, and nitrate of silver added ; a curdy precipitate, or a turbidity, indicates chlorine. v. Detection of Nitric Acid. A considerable quantity of the water is evaporated to dryness, the residue is dissolved in a small quantity of water, and tested as directed in section 203. vr. Detection of Iodine. A large quantity (four or five gal- lons) of the water is reduced to a small bulk by evaporation, PAET II. 2 P 566 QUANTITATIVE ANALYSIS. filtered, and a portion of the filtrate is tested with starch paste and nitric acid as directed in section 199. vu. Detection of Bromine. Another portion of the residual fluid is treated with chlorine water and ether, as described in section 200. Detection of Bases. I. Detection of Lime. Add to a portion of the water, concen- trated if necessary by evaporation, oxalate of ammonia, and allow it to stand for a considerable time in a warm place ; a white precipitate indicates lime. IT. Detection of Magnesia. Concentrate the filtrate from the oxalate of lime, if necessary, by evaporation, and add to the clear residue ammonia and phosphate of soda - y the formation, either immediately, or after agitation and rest, of a crystalline white precipitate, indicates magnesia. in. Detection of Lime in tlie form of Bicarbonate. To a por- tion of the water, solution of chloride of ammonium is added, and then ammonia ; the formation of a white precipitate indicates bicarbonate of lime. Should a notable quantity of iron be present, the precipitate has a more or less yellow tinge ;. and this is also the case if the water contain organic matter. iv. Detection of the Alkalies. Should the water contain iron, that metal must be brought to the state of sesquioxide by means of chlorine, it is then precipitated, together with all the other bases but magnesia, by carbonate of ammonia ; the further pro- cess for the detection of the alkalies is conducted as described in section 159. v. Detection of Ammonia. A. few drops of hydrochloric acid are added to a considerable quantity of the water, which is then evaporated very nearly to dryness at a gentle heat ; the resi- due is tested for ammonia by caustic potassa or by hydrate of lime. Besides the above acids and bases, there are found occa- sionally in mineral waters the following substances :lithia, alumina, strontia, oxides of manganese, zinc, and copper, amongst bases; and sulphurous, lor acic, phosphoric, hydrofluoric, and sili- cic acids among electro-negative bodies. These bodies are never however found all together in the same water, and as they generally exist in small quantities only, they are, as before ob- served, best detected in the quantitative examination. 209. Quantitative Analysis. The substances to be estimated may be, amongst bases, potassa, soda, lithia, ammonia, lime, mag- ANALYSIS OF MINERAL WATERS. 567 nesia, strontia, alumina, oxides of manganese and iron; and amongst electro-negative bodies, sulphuric, phosphoric, silicic, nitric, and carbonic acids ; chlorine, bromine, iodine, fluorine, and sulphuretted hydrogen, crenic, and apocrenic acids. If the operator should have the opportunity of collecting the water himself at the source, he should pay attention to the fol- lowing particulars : 1st, its temperature as compared with that of the air ; 2nd, the geological formation of the locality whence the spring originates ; 3rd, the yield of water ; 4th, its physical appearance ; 5th, whether or not it deposits any solid matter in the pipes or channels through which it makes its exit from the spring or well ; 6th, its smell and taste : in short, he should make himself acquainted as accurately as possible with every particular he can obtain with respect to the general history of the water he is about to analyse. The quantity of water required for a thorough examination is about 12 gallons ; if, however, the analyst should have an opportunity of evaporating down a considerable quantity (say 10 gallons to 2 or 3 pints) on the spot, for the detection and estimation of those ingredients which only exist in minute quan- tities, such as iodine, bromine, phosphoric acid, fluorine, oxide of manganese, alumina, strontia, lithia, and the organic acids; or, if he should choose to dispense with a search for these bodies, then a few pints will suffice for a complete analysis of the water for the other ingredients. I. ^Estimation of Carbonic Acid. The water for this purpose must be collected at the spring or well by immersing in the water a large pipette of" exactly known capacity ; four or five stoppered bottles, each of about a pint capacity, should be charged each with one ounce of a strong solution of chloride of calcium and an ounce and a half of ammonia, free from carbonic acid ; they should then be filled with the water poured quietly in, to prevent any loss of carbonic acid from splashing ; all the car- bonic acid contained in the water will thus be precipitated in the form of carbonate of lime. Thus CaCl + .NH 4 + C0 2 = CaO,C0 2 + NH 4 C1. The precipitates in the different bottles, which should be nearly alike in quantity, are collected on a filter, dried at 212, and weighed. They are then mixed together, and analysed by Ere- senius and Will's method (p. 268) ; the result shows the total amount of carbonic acid in the united volumes of the several bottles of water. 2p 2 568 QUANTITATIVE ANALYSIS. Should the water contain sulphuretted hydrogen, the bottles may be charged with a small quantity of a solution of arsenious acid in hydrochloric acid, and the hydrosulphuric acid estimated as sulphide of arsenic (p. 443) ; or the estimation may be made directly in the water by a standard solution of iodine in iodide of potassium (p. 442). ii. Determination of the Specific Gravity of the Water. This is effected in a very simple manner by the method described in page 228. As the specific-gravity bottle has been adjusted with distilled water at 60 F., it is of course necessary that the mi- neral water should be brought to the same temperature previ- ous to weighing it. The bottle contains exactly 1000 grains of distilled water ; the specific gravity, therefore, of the mineral water is exactly as its weight, water being 1000. in. Determination of the total amount of Solid Ingredients. A quantity of the water (varying from 1000 to 10,000 grains) is carefully evaporated to dryness on the water-bath, the residue is dried at various temperatures, and the weight taken when no fur- ther diminution is perceptible. If the water contain any con- siderable amount of earthy chlorides, and especially if it con- tain ammoniacal salts, a very exact result cannot in this man- ner be obtained regarding the amount of the solid contents, for the earthy chlorides retain a small proportion of water at 212, while, at a higher temperature, some hydrochloric acid may be set free ; chloride of ammonium is also partly volatilized at temperatures higher than 212. If the water do not contain ammonia, oxide of iron, or alumina, Schweitzer (Mem. Chem. Soc., vol. ii. p. 201), recommends, as a control, to change the carbonates, chlorides, etc., into sulphates, the total amount of which must closely correspond with the amount of the various ingredients when computed as sulphates ; this method is indeed applicable when ammonia is present, as that substance being dissipated by heat must be deducted from the total amount of ingredients obtained by analysis. Another method, in cases where great accuracy is desirable, is to evaporate the water with a proportionate quantity of chemically pure anhydrous carbonate of soda sufficient to decompose the earthy chlorides and sulphates. This method of evaporating with carbonate of soda is advantageous, moreover, in preventing the formation of sulphate of lime, which is a troublesome ingredient, inasmuch as it adheres closely to the evaporating vessel, is not readily soluble in dilute hydrochloric acid, and very sparingly soluble in water. ANALYSIS OF MINERAL WATERS. 569 As the solid matters obtained by the evaporation of a mineral water are frequently very deliquescent, the crucible should be cooled underneath a receiver in close proximity to a vessel con- taining sulphuric acid (p. 247), and weighed as quickly as possible. The general character of the water having been ascertained by the qualitative examination, the quantitative analysis is pro- ceeded with in the following manner : IT. Estimation of the Sulphuric Acid. Ten thousand grains (more or less, according as the preliminary experiment has indi- cated a greater or less amount of sulphuric acid) are acidified with hydrochloric acid, and precipitated at a boiling temperature by chloride of barium ; the precipitate is allowed to settle, then collected on a filter, washed, dried, ignited, weighed, and the amount of sulphuric acid calculated as directed, page 438.* v. Estimation of the Chlorine, Iodine, and Bromine (should the water contain the two latter), conjointly. Two thousand or three thousand grains of the water, according to the quantity of chlorine, etc., supposed to be present, are acidified with ni- tric acid, the chlorine precipitated with nitrate of silver, and estimated as directed, page 485. Should the water contain organic matter, it should be freed from it before estimating the chlorine, or it would be partly precipitated, together with the silver-salt, thereby increasing its weight ; with this view it should be evaporated with carbonate of soda (perfectly free from chloride of sodium), the liquid filtered from the earthy carbonate precipitated, evaporated to dryness, the residue fused, redissolved in water, nitric acid added, and then precipitated with nitrate of silver. As an illustration of the interference of organic matter with a correct determina- tion of the chlorine, Schweitzer obtained from the Bonnington water, as a mean of several experiments, 10'053 grains of chloride of silver from 4 ounces of the fresh water, and 9'694 grains after evaporating it with carbonate of soda, as above described. vi. Estimation of the total amount of Lime and Magnesia, Si- licic Acid, and Oxide of Iron. From 1000 to 10,000 grains of the water are acidified with nitric acid, and evaporated to dryness * It is better to determine the various ingredients of a mineral water from weighed than from measured quantities of the water ; if the operator have not the advantage of the large balance, Fig. 43, he can employ the spe- cific-gravity bottle, the contents of which, when quite full and the stopper in its place, are known accurately. 570 QUANTITATIVE ANALYSIS. at a heat below ebullition. The perfectly dry residue is digested with dilute hydrochloric acid ; the silicic acid remains undis- solved, having, by the evaporation, been reduced to its insoluble condition. It is separated by filtration, and estimated as di- rected in section 195. The filtrate which contains the iron, lime, and magnesia in the form of chlorides, is supersatu- rated with ammonia, by which oxide of iron is precipitated, and is collected and determined as directed in section 169. The lime is thrown down as oxalate, from the filtrate from the oxide of iron, by oxalate of ammonia, and estimated as carbonate or sul- phate in the manner described in section 158, and the magnesia is precipitated from the filtrate from the oxalate of lime (con- centrated, if necessary, by evaporation) with phosphate of soda, and estimated as pyrophosphate, as directed in section 159. Schweitzer proceeds in a different manner. He commences with evaporating a large quantity of the water down to a few ounces, together with a proportionate quantity of anhydrous carbonate of soda ; he throws the residue on a filter, and lixi- viates the solid matter with boiling distilled water till the fil- tered liquor gives no indications of sulphates and chlorides; he then makes analyses of the soluble and insoluble ingredients separately. The soluble portion contains the alkalies and mag- nesia in the form of sulphates, chlorides, nitrates, bromides, or iodides ; and the insoluble portion contains the earths in the form of carbonates ; silica, oxide of iron, oxide of manganese, etc. VIT. Estimation of the Lime, Magnesia, and Oxide of Iron dissolved in the water ~by free Carbonic Acid. Ten or twelve thousand grains of the water are boiled for some time in a flask, the free carbonic acid is hereby expelled, and the substances which it held in solution are precipitated. The operator must be careful not to allow the volume of the water to be much diminished by the ebullition, or a portion of sulphate of lime may be precipitated, together with the earthy carbonates ; boil- ing distilled water should be added from time to time to the flask. The precipitate is collected on a filter, dissolved in hydrochloric acid, a few drops of nitric acid added, and the oxide of iron, lime, and magnesia separated from each other in the usual manner. Schweitzer dissolves the insoluble matter obtained by his process in nitric acid, evaporates the solution to dryness, and leaves the residue for several hours in contact with nitric acid ; he then boils with hydrochloric acid, and, having sepa- ANALYSIS OF MINERAL WATERS. 571 rated the silicic acid, be precipitates the iron by succinate of ammonia. To the filtrate from the persuccinate of iron he adds sulphide of ammonia, which throws down the manganese ; he boils the sulphide of manganese with hydrochloric acid, preci- pitates it with carbonate of soda, washes and ignites the pre- cipitate, which he estimates as red oxide of manganese. Having removed the excess of sulphide of ammonium from the fluid filtered from the sulphide of manganese, by boiling it with hydrochloric acid, Schweitzer mixes it in a vessel that can be closed, with ammonia: the precipitate which hereupon forms, he filters quickly, protecting it as much as possible from the air, and having washed and ignited it, he redissolves it in hydro- chloric acid. He mixes the solution with bicarbonate of po- tassa, and, having dissolved the precipitate (should any form) in hydrochloric acid, he boils the solution with caustic soda ; if a precipitate should hereby be produced, it is oxide of iron, and must be redissolved in hydrochloric acid and precipitated by ammonia: the caustic ley may contain alumina ; to obtain which, it is supersaturated with hydrochloric acid, and precipitated by ammonia. All the magnesia present is retained in solution by the bicarbonate of potassa ; to obtain it, Schweitzer neutra- lizes the fluid with hydrochloric acid, evaporates it quickly with carbonate of potassa, and, when nearly dry, mixes it with boiling water, and filters ; he then washes, ignites, and esti- mates the residue as magnesia. The ammoniacal solution from which the magnesia had been precipitated contains the lime, which is precipitated with oxalate of ammonia. The detection of small quantities of manganese in the iron precipitated is best effected by Crum's process, which consists in treating the precipitate with nitric acid, and adding a small quantity of linoocide of lead. If manganese be present, the liquid exhibits, as soon as the precipitate has subsided, the magnificent colour of permanganic acid. vin. Estimation of the Lime and Magnesia in the filtrate from the precipitate produced by ebullition. They are determined in the usual manner, the lime being first precipitated by oxalate of ammonia, and the magnesia by phosphate of soda. The joint amount of the oxide of iron, lime, and magnesia obtained in the precipitate by boiling, and, in the filtrate, ought to cor- respond with the total quantity of those ingredients found by the previous experiment. ix. Estimation of the Alkalies. About 10,000 grains of the 572 QUANTITATIVE ANALYSIS. water are evaporated to about one-half, then mixed with excess of baryta water, filtered, carbonate of ammonia added, again fil- tered, the filtrate evaporated to dryness, ignited, and the alka- lies separated from each other, and determined as directed, page 259. If however the water contain a fixed alkaline car- bonate, a different plan is adopted. The water is concentrated by evaporation, filtered, and divided into two parts ; in one, the amount of chlorine is at once determined by nitrate of eilver ; the other is evaporated to dryness with slight excess of hydrochloric acid, the residue gently ignited, redissolved in water, and precipitated in its turn by nitrate of silver. The amount of chlorine obtained from this second portion above that obtained from the first, is the equivalent of the quantity of carbonated alkali originally present in the water. Having thus ascertained the amount of carbonated alkali, the total quantity of the alkalies present is found by boiling the water with chloride of barium for some time, and then adding baryta water and filtering ; the filtrate is precipitated with carbonate of ammonia, again filtered, and evaporated to dryness ; the re- sidue consists of the alkalies in the form of chlorides. x. Estimation of the Iodine. The filtrate from the precipi- tate produced by evaporating 10 or 12 gallons of the water down to 2 or 3 pints, is evaporated to dryness, together with, carbo- nate of soda, the residue is digested repeatedly with alcohol, the alcoholic extract evaporated to dryness, and the residue gently ignited in a platinum crucible to destroy the organic matter ; it is then dissolved in water and filtered, the clear liquor is carefully neutralized with dilute hydrochloric acid, and then precipitated with protochloride of palladium ; the prot- iodide of palladium, after standing at rest twelve hours, is collected on a filter, washed with warm water, dried at a tem- perature below 212, and weighed. For its composition, see page 496. xi. Estimation of the Bromine. The fluid filtered from the protiodide of palladium is saturated with sulphuretted hydrogen gas, in order to remove the excess of the palladium salt ; it is then filtered, boiled with nitric acid, filtered again if necessary, and then precipitated with nitrate of silver; the precipitate which contains all the bromine, together with the chlorine in the form of bromide and chloride of silver, is analysed according to the directions given in section 200. Schweitzer(Phil. Mag., vol. xv. p. 51), having separated the excess of the palladium salt, ANALYSIS OF MINERAL WATERS. 573 evaporates to dryness,redissolves in water, concentrates, and then adds to the solution a few drops of an ammoniacal solution of chloride of silver, prepared by mixing one part of a saturated solution of chloride of silver in ammonia, with one part of ammonia and one of water ; this solution, though it produces no turbidity in a saturated solution of chloride of sodium, he finds to indicate a very minute quantity of bromine, which it precipitates as bromide of silver. xn. Estimation of the Nitric Acid. About a gallon of the water, more or less, according to the amount of solid ingredients, is evaporated with an excess of pure carbonate of soda : the pre- cipitate which is formed is filtered off and washed, the filtrate is evaporated to dryness, and the residue well mixed and weighed ; it is divided into two or more parts, each of which may be analysed for nitric acid by Pelouze's method, as modified by Presenius (page 517). For methods of detecting minute quanti- ties of ammonia and nitric acid in rain-water, the student is referred to a valuable memoir read at the meeting of the British Association in 1854, by Messrs. Lawes and Gilbert, and to a paper by Dr. E. Pugh (Quart. Journ. Chem. Soc., vol. xii. p. 35). xiii. Estimation of the Ammonia. A considerable quantity of the water is concentrated by evaporation, having been previ- ously acidified slightly with hydrochloric acid. The concentrated i fluid is mixed in a tubulated glass retort with considerable j excess of caustic soda, and boiled for a long time, the evolved j vapours being condensed in a receiver containing dilute hydro- { chloric acid, and kept cool by being surrounded with ice or I cold water. The boiling having been continued till only two or three ounces of fluid remain in the retort, the distilled fluid is mixed with bichloride of platinum, and evaporated in a por- celain vessel. The dry residue is lixiviated with a mixture of two volumes of alcohol and one of ether, in which the am- monio-chloride of platinum is perfectly insoluble ; it is then dried at 212 and weighed : for its composition, see page 272. To control the correctness of the weight, the double salt may be ignited, and the weight of the platinum obtained compared with theory. Schweitzer states that the addition of bichloride of platinum to the distillate previous to evaporating it even at a temperature below 212, is quite essential to obtaining correct results, inasmuch as chloride of ammonium rises with the aque- ous vapours, a fact easily ascertained by covering the vessel during evaporation with filtering-paper, which when lixiviated 574 QUANTITATIVE ANALYSIS. clearly evince the presence of chloride of ammonium, as had been noticed by Berzelius. xiv. Estimation of Litliia. Should there be reasons for sus- pecting the presence of this alkali, it must be looked for in the filtrate, produced by evaporating 10 or 12 gallons of the water, about a pound of which should be retained for the purpose. It is boiled to dryness with excess of carbonate of soda ; the re- sidue redissolved in water, filtered, and again evaporated to dryness with phosphate of soda ; the residue is digested at a gentle heat with water, then ammonia added, and the precipi- tate (if any) filtered off after standing for some hours : it is washed on the filter with ammonia-water, and dried at a mode- rate heat. Basic phosphate of lithia (3LiO,P0 5 ) contains 37 per cent, of lithia. xv. Phosphoric Acid. This is looked for in the precipitate formed by boiling down several gallons of the water. A por- tion is dissolved in hydrochloric acid, and the phosphoric acid tested for, and estimated as directed in section 193. Another portion of the precipitate may be examined for iron, alumina, and manganese, with which view it is redissolved in hydrochloric acid, and nitric acid having been added to peroxide of iron, the solution is supersaturated with caustic potassa, by which the iron and manganese are precipitated, and the alumina retained in solution ; from the alkaline solution the alumina is preci- pitated by chloride of ammonium : the mixed oxides of iron and manganese are redissolved in hydrochloric acid, and the oxides separated as directed, page 313. xvi. Fluorine. This element is to be sought for in the inso- luble portion of the saline residue from a considerable quantity of the water. About 500 grains are mixed with a small quantity of pure, finely divided quartz, and boiled for about an hour \ with concentrated sulphuric acid. If fluorine be present, it is | disengaged in the form of fluoride of silicon, which is passed ' into water, the resulting hydrofluosilicic acid is decomposed by i ammonia, and the fluoride of ammonium, after separating the ; silica by filtration, is evaporated, and gently heated in a platinum crucible with concentrated sulphuric acid ; hydrofluoric acid is disengaged, which is recognized by its action on glass. Should fluorine be present in some quantity, it will become perceptible at once on passing the fluoride of silicon into water, a floccu- lent precipitation of silica taking place, which becomes aug- mented by the addition of ammonia. ANALYSIS OP MINERAL WATERS. 575 XYII. Estimation of Organic acids {Crenic and Apocrenic). According to the observations of Mulder, which will be referred to again in the section on the analysis of soils, there exist in the soil at least seven different organic substances, viz. crenic acid, apocrenic acid, ge'ic acid, Jiumic acid, ulmic acid, liumin, and ulmin ; he divides these substances into two groups the Jiumic, including ge'ic acid, Jiumic acid, ulmic acid, humin, and ulmin, and the crenic, including crenic and apocrenic acids ; of these bodies humin and ulmin are insoluble in water and in alkalies, the others are more or less soluble in both ; the members of the humic group are precipitated from their alkaline solution by an acid, while apocrenic and crenic acids are retained in solution ; none of these acids contain nitrogen as a constituent element, though they all stand in the closest relation to ammonia, with which they form salts in the soil, and the same is probably the case when the two latter exist in mi- neral waters. The crenic and apocrenic acids are not precipitated from their alkaline solutions by an acid, as above stated. To de- Ftect and estimate them in water, therefore, a weighed por- i tion of the precipitate, produced by boiling down some gallons, i is boiled for some time with potassa ley, then passed through a i filter, acidified with acetic acid, and acetate of copper added ; the formation of a brownish precipitate indicates apocrenate I of copper, which contains, according to Mulder, about 50 per [cent, of apocrenic acid. The fluid filtered from this precipi- tate is mixed with excess of carbonate of ammonia, and then boiled ; if crenic acid be present, a bluish-green precipitate is formed, which, when dried at 281 F., contains, according to eri- Unt. Sul- phuric Acid. Chlo- rine. Iodine with traces of Bro- Silica. Carbonic Acid free and com- Carbonic Acid as Carbonate of Lime and Carbonate of Car- bonic Acid, Phos- phoric Acid. Apo- crenic Acid. mine. bined. Magnesia. 24-00 2^82 0-27 j) 2-38 5J 23-80 2-80 5? 0-29 5 2-36 5} JJ JJ 3an 23-90 2-81 0-0094 0-28 5-403 2-37 3-133 0-031 0-112 PART II. 2Q 582 QUANTITATIVE ANALYSIS. Verification for Lime. Experiment. Lime precipitated by boiling. Lime left in solu- tion after boiling. Total by Calculation. Total by Experiment. I .. 2-282 3-280 6-108 6-422 II 2-857 3-310 6-077 6-257 Mean 2-842 3-296 6-092 6-340 Verification for Magnesia. Experiment. Magnesia precipitated by boiling. Magnesia left in solution. Total by Calculation. Total by Experiment. L... 0-11 4-735 4-845 4-748 II. . 0-15 4-785 4-945 4-783 Mean 0-13 4-755 4-890 4-765 Residue left on Evaporation. Experiment. Mineral Residue. Organic Residue. Total Residue. I. . 51-02 0-114 51-114 II. . 51-21 0-110 51-320 Mean 51-11 0-112 51-217 Solid contents of an Imperial Pint as determined ly Analysis. Carbonate of lime 5'0760 Carbonate of magnesia ...... 0'2630 Lime (not as carbonate) 3'2200 Magnesia (not as carbonate) .... 4*7580 Potassa 0-6930 Soda 9-5460 Chlorine 2'8100 Sulphuric acid 23-9000 Silica . 0-2800 ANALYSIS OF MINERAL WATERS, 588 Phosphoric acid 0-0310 Bromine traces Iodine 0'0094 Crenic acid traces Apocrenic acid . 0*1120 50-6984 Residue left on evaporation . 51*2170 Saline Constituents of one Imperial Gallon. Grains, Carbonate of lime . . . 40-608 Carbonate of magnesia ...... 2*104 Sulphate of potassa ........ 10-264 Sulphate of soda . , 174*904 Sulphate of lime 62-560 Sulphate of magnesia 76*592 Chloride of magnesium ...... SO'OOO Iodide of sodium 0'088 Silica 2*080 Phosphoric acid . 0*248 Crenic acid traces Apocrenic acid 0*896 Bromine. .......... traces 400*344 xx. Determination of the Volume of the Gases. All water, including even that which has been recently dis- tilled, contains air; if it be desired to collect and estimate the volume of these gases, the process invented by Professor Bun- sen may be followed. This consists in expelling the gases by ebullition in vacuo, and is conducted in the following manner (Hofmann, Journ. Chem. Soc., vol. vii. p. 166) : " A globular glass flask, a, over the neck of which is fitted a strong vulcanized caoutchouc tube, is immersed in the water until it is filled, and the tube is then closed by means of a brass screw-clamp, which is screwed immediately above the neck of the flask. In this manner a quantity of water is secured which has not been in contact with the air, and the exact volume may be readily determined by experiment. The caoutchouc tube is then connected with a small glass globe, >, 2Q2 584 QUANTITATIVE ANALYSIS. partially filled with water, and provided with two necks oppo- site each other ; the one corresponding in diameter to the neck . Fig. 10.7. of the flask, the other the size of an ordinary quill pen. By the latter, the globular vessel, which may be called the ' boiler globe ' may be connected with the glass tube, c, about ten inches long and three-quarters of an inch wide, and terminating on each side in a narrow open connecting tube, corresponding in width to the narrow tube of the boiler-globe. The connection of the latter, with the tube which may be termed the ' gas-receiver,' is effected by means of a caoutchouc tube just sufficiently long to admit between the two glass tubes, which are fixed to its ends, the application of a small brass screw-clamp, by means of which arrangement communication between the two vessels may be established or interrupted. In a similar manner the other extremity of the gas-receiver is connected with an ordi- nary delivery tube, which discharges into water. " The water flask, a, being closed with its screw-clamp, the water in the boiler-globe is heated to ebullition by means of a spirit-lamp, and kept boiling for about a quarter of an hour, by ANALYSIS OF MINERAL WATERS. 585 which time all the air in the apparatus is perfectly replaced by steam, which may be ascertained, moreover, by not a trace of air collecting if the delivery tube be made to discharge for a few minutes under an inverted bell-jar filled with water. The caoutchouc tube at the upper end of the gas-receiver is now closed by a brass clamp, and the spirit-lamp simultaneously re- moved, when the ebullition in vacua will continue for a consi- derable time. The boiling out of the water in the flask may now commence. For this purpose the brass clamp is removed, and the flask submitted to a gradually increasing heat. Air bubbles soon make their appearance, but a quarter of an hour generally elapses before the caoutchouc tube, which is collapsed and almost flattened by the atmospheric pressure, is sufficiently expanded to admit of a free communication between the flask and the upper part of the apparatus. The water gradually enters into violent ebullition, in which state it is kept for about an hour, so as to ensure the complete expulsion of every trace of gas. When this point has been reached, the heat is increased for a moment, a good deal of steam is generated in the flask, which causes the water gradually to rise from the boiler-globe into the narrow end of the gas-receiver. At this moment the flask is closed with the clamp, and the source of heat removed ; the water in the boiler-globe begins to cool, gradually contracts, and descends in the narrow tube of the gas-receiver ; the very moment its level disappears under the caoutchouc connector, the clamp between the boiler-globe and gas-receiver is fastened, and the whole apparatus disconnected. The gas-receiver, closed at both its extremities by caoutchouc, contains the whole of the gas, together with a few drops of water ; the volume of the gas may be readily measured by opening one of the clamps under mercury and allowing the metal to enter the gas-receiver until it stands at the same level within and without the tube ; the clamp is then secured again, and the mercury transferred into a graduated cylinder. The difference between the volume of mercury and the known capacity of the gas-receiver represents the volume of the gas at the existing barometric pressure and temperature. The only conditions necessary for the success of this ap- parently complicated process are a sufficiently long ebullition, and a proper proportion in the capacities of the water-flask and gas-receiver, which can only be attained by a few preliminary operations. The capacity of the gas-receiver must be greater 586 QUANTITATIVE ANALYSIS. than the volume of the gas in the water-flask, when measured under the ordinary pressure of the atmosphere. If the volume of the gas were greater, the gas-receiver or the caoutchouc con- nector might burst, and if they actually withstood the outward pressure, a certain quantity of gas would be lost on opening the clamp under mercury. 587 CHAPTEE XI. ON THE ANALYSIS OF SOILS. 210. In the examination of soils, the analytical process is divided into two parts : 1. The mechanical analysis ; 2. The chemical analysis. 1. Mechanical Analysis. (a.) Selection of the sample. Too much care cannot be taken to obtain a fair average specimen. For this purpose one or two pounds should be taken from each of four or five different parts of the field where the soil appears to be nearly the same. These should be well mixed together, and a pound or so selected for analysis ; all samples should be kept in well-corked bottles. It is not unfrequent to see in a field otherwise fertile, a few patches almost barren, where plants, especially when the field is in white crop, spring up, and for a time look quite healthy, but soon become diseased, assume a yellow colour, and die. Spe- cimens from such parts should on no account be mixed with the rest ; they should be examined by themselves, and the results compared with those given by the fertile parts : by following this course, the cause of sterility and the means of curing it are most likely to be discovered. (b.) Determination of Water. Spread a weighed quantity (say half a pound) of the soil upon a sheet of white paper, and expose it to the air in a dry room for several hours, weighing it at intervals of two or three hours till the weight remains constant : the loss indicates the amount of water which has eva- porated, but by no means the whole of the water which the soil contains. To determine which, heat about 500 grains of the air-dried soil in a small glass beaker plunged into an oil bath, the temperature of which is kept between 300 and 350 F., till it ceases to lose weight, the result gives a close approxima- 588 QUANTITATIVE ANALYSIS, tion to the amount of water. Absolute desiccation cannot, how- ever, be accomplished except at a heat close upon redness, which is, of course, inadmissible, as the organic matters the soil con- tains would thereby become altered or destroyed. (c.) Absorbing power. Allow the 500 grains of soil dried as above to cool in a covered vessel ; then spread it out on a sheet of paper, and expose it to the air for twenty-four hours : note the increase of weight which is due to absorption of water, and if it amounts to ten grains, it is, so far, an indication of great agricultural capability. (d.) Power of holding Water. Put 1000 grains of air-dried soil into a filter enclosed in another placed in a funnel ; pour cold water drop by drop on the soil until it begins to trickle down the neck of the funnel ; cover with a piece of glass, and allow it to stand for an hour or two, adding a few drops of water from time to time until it is certain that the whole soil is per- fectly soaked ; remove the filters from the funnel, and open them upon a linen cloth to remove the drops of water adhering to the paper. The outside filter is now placed in one pan of the balance, and the inner one containing the soil on the other ; and the whole being carefully balanced, the true weight of the wet soil is obtained: suppose this to be 1400, then the soil is capable of holding 40 per cent, of water. (e.) Rapidity of drying. Expose the saturated soil with its filter on the plate to the air, for four, twelve, or twenty -four hours, weighing from time to time. The loss of weight, indi- cating the tendency of the soil to dry, may convey useful infor- mation as to the necessity or otherwise of drainage. (f.) Relative proportions of gravel, sand, and clay. Rub a quantity of air-dried soil between the hands, and remove and weigh any stones which may be present. Weigh off 4000 grains, and pass them, through a sieve (No. 1) of copper- wire gauze, the meshes of which are about T ^th of an inch in diameter.* Kemove the sieve from its bottom, and place it over a deep eva- porating basin ; throw a gentle stream of water upon the con- tents, and stir with a spatula or the hand until the water passes through clear. Transfer the residue to another basin, and place it in the water oven to dry ; then weigh, after which ignite in the air, and when cold, weigh again. The first weighing gives the amount of coarse gravel, and the second indicates the pro- * Sieves of various degrees of fineness may be obtained at most of tho philosophical instrument makers, or they may easily be made. ANALYSIS OF SOILS. 589 portion of organic matter which this gravel contains. Transfer the soil which has passed through sieve No. 1 to sieve No. 2, the meshes of which are about ^th of an inch in diameter ; treat the residue on the sieve precisely as before, dry at 212, weigh, ignite, and weigh again ; the results give the amount of gravelly sand, and of organic matter mixed with it. Dry a por- tion of the soil which has passed through sieve No. 2 in the water oven, and weigh off 500 grains ; transfer to a deep basin or flask, and boil for twenty minutes or so, with water. The boiling must be continued until the particles are thoroughly separated from each other. ^ The' coarse sand, fine sand, and finely divided particles are then separated from each other by the following simple process, recommended by Schultz : The boiled soil is allowed to cool, and is then washed into an elutri- ating glass, which is merely a tall champagne glass 7 or 8 inches deep and about 2f inches wide at 'the mouth, round which is fastened a brass ring about half an inch broad, with a tube slightly inclined downwards proceeding from its side. A gentle stream of water is caused to pass continuously into the elutri- ating glass in such a manner as to cause a constant agitation of the particles, whereby the finest are washed away through the tube at the top of the glass, and received in a beaker or any other convenient vessel. This stream of water is best kept up and regulated by causing it to flow from a reservoir provided with a stopcock, to which is attached a tube funnel from 12 to 18 inches long, drawn out to a point, with a fine aperture. The end of this tube is placed nearly at the bottom of the elutriating glass, and the supply of water so adjusted that the funnel tube always remains half full of water. When the water runs off from the discharge tube nearly clear, the stopcock of the reser- voir is closed, and the elutriating glass being removed, the water is decanted from it, and the residue is washed into a small dish, where it is dried and weighed, after which it is ignited and weighed PI gain ; the two weights give the proportion of coarse sand and its organic matter. The elutriated turbid fluid is al- lowed to stand for several hours, and the water is then poured off into another beaker. The deposited matter, consisting of fine sand and fine soil, is then subjected to a second elutriating process, conducted as before, except that the force and volume of the washing water is considerably lessened. The operation is con- tinued until the wash water passes off quite clear ; this some- times takes three or four hours, but it is, with the arrangement 590 QUANTITATIVE ANALYSIS. we have described, a self-acting process, requiring no personal superintendence. The residue in the elutriating glass is fine sand, which, with its organic matter, is estimated as before, by drying, weighing, igniting, and re- weighing. We have only now to deduct from the original 500 grains the quantities of coarse and fine sand, to obtain the proportion of finely-divided matter. The results of this mechanical analysis may be tabulated thus (Schultz) : Fixed Combustible Substances. Substances. 6>go / Gravel (coarse) 6'90 0-00 \ Organic matter Gravel (fine) 6'43 - Organic matter 0'67 qn^ft / Sand ( coarse ) 34-37 \Organicmatter 1-13 / Sand (fine) 38'50 - \ Organic matter 1'50 Fine soil 9'50 Organic and other volatile substances . . . I'OO 100-00 95-70 4-3 Stones, 2'10 per cent. This mechanical treatment of soils we deem to be of high im- portance, and it is to be regretted that so few of our English soils have hitherto been so examined. The same remark applies to the analysis of clays, to which the above operations equally apply, except that in the case of clays we have not to look for gravel. The following are the results of the mechanical analysis of two clays, one a " fat" and the other a " poor " variety. Fat Clay. Poor Clay. Coarse sand 6'66 . . . 24r68 Fine sand 9'66 . . . 11'29 Clay 74-82 . . . 57'82 Water 8-86 . . . 6-21 100-00 100-00 These analyses will illustrate the usefulness of the mechanical analysis. To render the matter complete, however, the gravel and sand should be moistened and examined under the micro- scope with the view of ascertaining if they are wholly siliceous, or if they contain fragments of different kinds of rock, sand- stones, slates, granites, traps, limestones, or ironstones.. A few drops of strong hydrochloric aoid should also be added, when the presence of limestone is shown distinctly by an efferves- ANALYSIS OF SOILS. 591 cence ; of peroxide of iron by the brown colour which the acid speedily assumes ; and of black oxide of manganese by the smell of chlorine, which is easily recognized. (g.) Determination of the density of a Soil. Dry a sample of the soil (from which the large stones have been picked out) at 212, in the water-oven, till it ceases to lose weight. Fill a perfectly clean and dry common phial with distilled water up to a mark made with a file on the neck, and weigh it carefully ; or the specific-gravity bottle with a long neck (Fig. 44, p. 228) may be used, but it should be of stout glass. Pour out part of the water, and introduce into the bottle in its stead 1000 grains of the dried soil ; shake the bottle well to allow the air to escape from the pores of the soil ; fill up again with water to the mark on the neck, and again weigh. The weight of the soil divided by the difference between the weight of the bottle with soil and water, and the sum of the weights of the soil and the bottle of water together, gives the density or specific gravity. Example. Grains. The bottle with water alone weighs . . 2000 The dry soil 1000 Sum (being the weight which the bottle with the soil and water would have had could the soil have been introduced without displacing any of the water) . 3000 Actual weight of soil and water . . . 2600 Difference (being the weight of water taken out to admit 1000 grains of soil) 400 Therefore 1000 grains of soil have the same bulk as 400 grains of water : that is, the soil is 2^ times heavier than water, since 1000 - = 2'5 = its specific gravity. (h.) Determination of the absolute iveight of the Soil. Weigh an exact imperial half-pint of the soil in any state of dryness. "When this weight is multiplied by 150, it will give very nearly the weight of a cubic foot of the soil in that state. 2. Chemical Analysis. Determination of the total quantity of organic substances, volatile Salts, and chemically combined Water. Having mixed well together the various samples taken from different parts of 592 -QUANTITATIVE ANALYSIS. the field, allow the mixture to remain for a whole day in the water-oven, then weigh off 1000 grains, and ignite gently in a platinum or porcelain crucible with constant stirring until all blackness has disappeared from the mass, allow it to cool, then drench it with water saturated with carbonic acid, dry, ignite, and weigh. The object of drenching with carbonic acid water is to restore any carbonic acid which may have been expelled by ignition from the chalk which the soil may have contained. In- stead of carbonic acid water, a concentrated solution of carbo- nate of ammonia may be used for this purpose. Pass the ignited soil through sieve No. 2, and submit to analysis the finer par- ticles consisting of clay and sand, weigh off" 200 grains, and di- gest for several hours on the water-bath with excess of a mixture of equal parts of strong hydrochloric acid and water, then add water, and filter into a graduated jar or flask. "Wash the resi- due on the filter with hot distilled water till it passes through tasteless, then break a hole through the bottom of the filter with a stirring-rod, and wash its contents into a small beaker, put it aside for the present, covering it with a glass plate, and marking it (A), calling the hydrochloric solution (B). Analysis of the Hydrochloric Solution, (B.) Dilute the solu- tion and washings with distilled water till 100 measures are obtained, well mix by agitation, and divide into four equal parts, I., II., III., IV. i. Determination of Sulphuric Acid. Heat to ebullition, then add chloride of barium in excess, put aside for twenty-four hours, pour off the clear liquid, transfer the precipitate to a filter, wash, dry, ignite, and weigh 100 parts 34'31 sulphuric acid. Multiply by 2, the product gives the amount of sulphu- ric acid in 100 grains of the soil. ii. Determination of Phosphoric Acid. Boil with a little nitric acid to ensure the peroxidation of the iron, then precipi- tate with ammonia, filter, and having washed the precipitate on the filter, redissolve it in dilute nitric acid. To determine the phosphoric acid in the nitric solution, Eresenius directs to proceed as follows : dissolve 20 grains of molybdic acid in 180 grains of strong ammonia, mix this with 400 grains of nitrie acid, and add the mixture to the nitric solution, digest for twenty-four hours on the water-bath, and wash the precipitate (if any) with the same solution of molybdic acid with which it was produced ; let the filtrate stand for some time on the water- bath to see whether a further precipitate will form, and if so, ANALYSIS OF SOILS. 593 add this precipitate to the first. Dissolve the precipitate on the filter in ammonia, wash, and agitate the solution well with solution of sulphate of magnesia, the precipitate which is thereby- produced is ammonio-phosphate of magnesia. After standing for twenty-four hours it is collected on a filter, and washed with water containing one-sixth of ammonia; it is then dried, ignited, and weighed : 100 parts correspond to 63*96 parts of phosphoric acid. This method, though somewhat troublesome, is accurate. If, however, the operator fail in obtaining any proof of phos- phoric acid, he is recommended to repeat the experiment, em- ploying the entire hydrochloric solution of 100 grains of the soil; it is advisable, also, to prepare a considerable quantity of the molybdic solution, as it is employed for washing as well as for precipitation. An important advantage in this method of determining phosphoric acid is, that the presence of alumina and sesquioxide of iron does not interfere with it, it is how- ever, necessary to add the molybdic solution in very consider- able excess ; and it is inapplicable where the phosphoric acid is in considerable quantity, which is not however likely to be the case in the analysis of soils. Chancel's method, by acid nitrate of bismuth, is stated also to give accurate results. See " PHOS- PHORIC ACID." Schultz recommends the following process : A considerable quantity of the ignited soil (700 or 800 grains) is digested with hydrochloric acid; the filtrate from the insoluble matter is nearly neutralized by dilute ammonia, taking care not to preci- pitate permanently the alumina and oxide of iron. To the so- lution, which should amount to about 1^ pint, 35 to 45 drops of perchloride of antimony are added, and set aside for twenty- four hours. A yellowish- white flocculent precipitate is obtained which contains all the phosphoric acid, but it is composed principally of antimonic acid, carrying with it some oxide of iron and alumina, and containing besides a quantity of ammonia in proportion to that of phosphoric acid. The precipitate well washed with water, is digested with soda-ley containing a cer- tain amount of silicate. After boiling, the liquid is filtered, the oxide of antimony, now changed into antimoniate of soda, re- mains on the filter with the alumina and oxide of iron. The solution containing the phosphoric acid with a little alumina and silica is supersaturated with hydrochloric acid, then ammo- nia added ; it is concentrated by evaporation, again treated with ammonia, and then filtered. The precipitate of alumina and silica thus obtained carries with it a small quantity of phos- 594 QUANTITATIVE ANALYSIS. phoric acid ; to remove this, the precipitate is redissolved in a drop or two of hydrochloric acid, the solution is evaporated to dryness, and the residue is warmed with a little acidulated water. After having filtered again to separate the silica, a small quan- tity of tartaric acid is added to the filtrate, which is then mixed with the preceding ammoniacal solution containing the greater part of the phosphoric acid, which is then precipitated as usual by sulphate of magnesia. in. Determination of the Iron. All fertile soils contain this metal when in an efficient state of cultivation, as sesquioxide ; not unfrequently, however, especially where the draining has been neglected, part, and occasionally indeed the whole, of the iron exists as protoxide. JSTow, as the latter oxide is positively injurious, while the former is probably absolutely necessary, ife is important to ascertain whether any, and if any, how much of the iron exists in the soil in the form of the hurtful oxide. For this purpose a portion of the fine non-ignited particles is extracted with hydrochloric acid, and filtered; the filtrate is nearly neutralized with ammonia, diluted with water, and, if necessary, filtered again ; a drop or two of solution of yellow prussiate of potash is added to the clear filtrate. If the preci- pitate which falls is dark blue, the whole of the iron exists as sesquioxide ; if the precipitate be pale blue, then part at least of the iron is in a state of protoxide. Suppose, first, that the whole of the iron exists as sesquioxide, one fourth part of the hydrochloric solution of 200 grains of the soil is transferred to a flask, a little more acid added, arid then some small fragments of pure metallic zinc, the flask being placed in a slanting position to avoid any loss, evolution of hydrogen gas takes place, which gradually reduces the iron to the state of protoxide, and the solu- tion at last becomes colourless. The amount of iron is now de- termined by the volumetric method of Marguerite. It is calcu- lated as sesquioxide (56 iron = 80 sesquioxide), and the amount obtained, multiplied by 2, gives the percentage quantity of sesquioxide of iron in the finely divided matter of the soil. Suppose, however, that the qualitative experiment has shown that some of the iron is in the form of protoxide, in that case a i second determination of iron is made in the hydrochloric solu- tion of a weighed quantity of the non-ignited fine matter of the soil, not previously reducing with zinc, the quantity found is calculated upon 100 grains of the soil, and the result deducted from the total quantity of iron found ; the difference gives the proportion of the metal as sesquioxide. ANALYSIS OF SOILS. 595 iv. In this portion of the hydrochloric solution, silicic acid, alumina, lime, and oxide of manganese are determined. It is evaporated to perfect dryness, a little nitric acid being added during the evaporation to peroxidize the iron, the dry and cold residue is sprinkled with hydrochloric acid, and after standing for half an hour, boiling distilled water is poured over it, and it is thrown on a filter; the silicic acid, having by this treat- ment been rendered insoluble, remains on the filter, on which it is washed, then dried, ignited, and weighed. The filtrate and washings contain the oxide of manganese, alumina, lime, and magnesia, which are separated from each other and estimated as in the analysis of iron ores (see "IRON"). If no oxide of man- ganese be present, the process may be much simplified. The solution, after the separation of the silicic acid, is at once pre- cipitated by slight excess of ammonia, and filtered as quickly as possible. The precipitate on the filter consists of alumina and sesquioxide of iron, it is washed, dried, ignited, and weighed. From the weight obtained, that of the sesquioxide of iron al- ready determined is deducted, the remainder is alumina ; the lime and magnesia are contained in the filtrate which is treated in the usual manner. v. Determination of the Alkalies. From 50 to 100 grains of the soil are taken for this purpose, and the method followed is that of Dr. Smith (p. 473), the results are very exact ; the fused mass should be digested with pure water instead of hy- drochloric acid as first recommended by Dr. Smith. vi. Determination of Carbonic Acid. This is effected on a separate portion of the gently ignited, finer part of the soil by the process of "Will and Fresenius. A quantity not less than 100 grains of the soil are introduced into the larger bottle A, Fig. 108, together with a little water ; B contains strong sulphuric acid. The upper aperture of the pipette c, which is filled with nitric acid, is stopped by a wax plug. The appara- tus being accurately weighed, the wax plug is loosened, upon which the nitric acid descends into A, and de- composes any earthy carbonate that *nay be present, the carbonic acid v Fig. 108. 596 QUANTITATIVE ANALYSIS. evolved escaping through the sulphuric acid in B, and thereby becoming dried. "When the effervescence is over, A is warmed, and air is drawn through the apparatus by applying suction at d, the wax plug is then replaced, and when cold, the apparatus is re-weighed, the loss of weight indicates the amount of car- bonic acid in the form of carbonates in 100 grains of the soil. If the soil be eminently calcareous, the nitric acid must be allowed to flow into the bottle B very gradually. The Organic Constituents of the Soil. Mulder (' Chemistry of Vegetable and Animal Physiology,' English Translation) re- duces the number of organic substances as at present known to exist in the soil to seven. The varieties would indeed be in- calculable but for the existence of a general cause which reduces them to a small number. This cause is the transformation and decomposition which vegetables undergo at their death, by which they are converted chiefly into a substance called humus. This change is remarkably uniform, since from the innumerable organic combinations which exist in plants and animals, the same few constituents of the black layer of soil are derived ; and if we exclude those substances which are mixed accidentally with the black soil, as well as those substances which have not yet undergone sufficient decomposition, its constituents are limited to a small number of organic substances, substances existing in it everywhere, and on the decomposition of which the growth of plants depends. These substances have the following properties: Some of them are soluble in water, others in alkalies ; others again are insoluble in both, while some dissolve more or less readily in alcohol and ether. The latter are of a resinous nature, and do not appear to have any share whatever in vegetation. If a soil be exhausted by means of water, a great many salts may be extracted from it, among which may be mentioned the alkalies, lime, magnesia, and ammonia, in combination with formic, acetic, carbonic, crenic, apocrenic, and liumic acids. From the soils thus treated with water, alkaline solutions ex- tract substances whicli may be precipitated by acids. In dif- ferent kinds of soils the substances thus extracted are some- times different. They all consist, however, of one or more of the following : Geic acid ..-.... C 40 H 13 O 14 Humic acid C^H^O^ Ulmic acid CHO ANALYSIS OF SOILS. 597 The latter is that which, in neutral vegetable substances under- going decay, is formed first. From it, by absorption of oxygen from the air, hmnic acid is produced, and finally, by a further absorption, gelc acid. The substances which are insoluble in alkalies, ulmin and humin, can be rendered soluble, and so converted into ulmic and humic acids respectively, by the de- composition which is always going on in the constituents of the soil. The base by which Mulder supposes these acids to be ren- dered soluble in water is ammonia. He does not adopt the idea of Liebig, that this ammonia is carried down to the soil from the atmosphere by means of rain water, but he accounts for its presence in the following manner : nitrogen in the state of pure gas, and also atmospheric air, are possessed of one common property, namely, that when in contact within an en- closed space with putrefying substances, from which hydrogen is in consequence given off, the nitrogen combines with the hy* drogen, and forms ammonia. This property of nitrogen is well known. It is the principle on which saltpetre is formed, the production of that substance being always preceded by that of ammonia. Now air is contained in the soil, and is in continual contact with moist and decomposing substances. This air could produce saltpetre, if there were only a sufficient abun- dance of bases, and that even without the presence of putrefy- ing organic substances. There are in Ceylon numerous natural saltpetre grottos, where there are np organic substances from which nitrogen might be supplied. Nitrogen is derived from the air con- tained in these caverns, and in favourable circumstances even the water is decomposed, and ammonia at the same time pro- duced, which afterwards is oxidized into nitric acid by the oxygen of the air in places where it has more easy access ; this acid then combines again with the bases from the walls of the grottos, and forms nitrates. All this would happen in the soil if organic substances where not present to absorb the oxygen, and thus to prevent the oxidation of the nitrogen in the am- monia, and the formation of saltpetre. In a porous soil in which moist air is contained, nitrogen is combined with the hydrogen of the organic bodies only into ammonia, the oxygen of the water and from the air being consumed in the higher oxidation of the organic substances themselves. In this way the first product of the decomposition of organic substances, PART II. 2 R 598 QUANTITATIVE ANALYSIS. namely ulmic acid (C 40 H 14 12 ), is converted into Tiumic acid (C 40 H 12 O 12 ), and this again into yeic acid (C 40 H 12 O 14 ), which in its turn is further oxidized into apocrenic acid (C 48 H 24 S2 ), and finally into crenic acid (C 24 H 12 16 ). It is to this formation of ammonia from the constituents of atmospheric air and water that we must look, according to the Dutch chemist, for the cause of one of the most important peculiarities in the growth of plants. It is owing to this slow formation of ammonia, that the organic substances of the soil insoluble in water, are rendered soluble, and can be offered to plants as organic food, even without a supply of ammoniacal manure to the soil. In other words, it is owing to this cause that the five acids already mentioned, can be converted into so- luble ammoniacal salts. It is particularly worthy of remark, and a fact of great interest, that these humic substances, which can be extracted from the soil by alkalies, and precipitated by acids, have a great uniformity, from whatever kind of soil they may have been prepared ; and that they are remarkably similar to those substances which by the action of several chemical agents may be obtained from the materials which are generally diffused through the animal and vegetable kingdoms. The manner in which sugar is changed into ulmin, humin, and ge'ic acid, by a simple transposition of its elements, is thus suggested by Mulder: c H Q J of 7 equivalents of sugar -f- O 2 . 42 35 37 2 equivalents of carbonic acid ..204 19 ditto water . . . . ' Q 19 19 1 ditto ulmin .... 40 16 14 42 35 87 % of 7 equivalents of sugar -f 4 . 42 35 39 2 equivalents of carbonic acid . . 2 04 23 ditto water . . . .* 23 23 1 ditto humin .... 40 12 12 42 35 39 % of 7 equivalents of sugar + 6 . 42 35 41 2 equivalents of carbonic acid ..204 23 ditto water .... 23 23 1 ditto ge'ic acid . . . 40 12 14 42 35 41 ANALYSIS OF SOILS. 599 In the same manner the conversion of cellulose, starch, gum, pectin, and other analogous substances, which are so very much diffused through plants, into the humic substances, may be re- presented ; and the way in which his protein even may be con- verted into humic acid, by the influence of hydrochloric acid and oxygen, is shown in a very ingenious manner by Mulder. The crenic and apocrenic acids, like the humic acids, exist in the soil in the state of salts of ammonia, potassa, and soda, combined with lime, magnesia, and oxide of iron ; they never exist in the soil uncombined. They unite intimately with am- monia, which, however, can be completely separated from them by caustic potassa. Apocrenic acid, when artificially prepared, is & five-basic acid. It is produced in the form of apocrenate of ammonia when humic acid in whatever way prepared, or wood charcoal, is exposed to the action of nitric acid. Its value in the soil is very highly estimated by Mulder, as in con- sequence of its Jive-basic character it may supply plants either with several bases at once, or these bases may be alternately exchanged according as the proportion of a more powerful base in the soil is less, and that of a more feeble one tempo- rarily greater ; in consequence of these properties he assigns to it a much higher rank than the humic class of acids. The formation of apocrenate of ammonia, by the oxidation of humate of ammonia, is continually going on in the soil during the warmth of summer (except on the very surface which is directly exposed to the air). Each minute portion produced can be taken up by the roots of plants, in the form of double apocrenates of ammonia and various fixed bases, provided there be a sufficient supply of water at hand ; and, while in this way the soil loses its apocrenates, a new portion of apocrenate of ammonia is formed from the humic acid or Jiumin, which is present in large excess. We may thus call the production of apocrenic acid in one respect an organic nitrification. The series of organic acids present in the soil is, according to the views of Mulder, concluded by a fifth important sub- stance, a final product of the oxidation of organic substances before they are entirely changed into carbonic acid and water, namely, crenic acid, the composition of which is C 24 H 12 lvi . It also is combined with ammonia in the soil, and forms double salts which are soluble in water. These exist, along with the apocrenates, in all kinds of water, which have been in contact with organic substances in the soil. They were first found by 2 n2 600 QUANTITATIVE ANALYSIS. Berzeiius in spring water; they exist also in the waters of ditches, marshes, and bogs. Grenic acid is four-basic. By the continual tendency to nitrification in the soil, apocrenic acid must always be converted into crenic acid ; the series of ulmic, hiwniC) ge'ic, and apocrenic acids thus terminating with crenic acid. In the upper layer of the soil, however, where the air is not enclosed,and consequently no tendency to nitrification exists, crenic acid must conversely be changed into apocrenic acid. Johnston thinks (translation of Mulder's work, p. 184), that to the five acids, mentioned by Mulder, several others may be added, as occurring in certain circumstances in the soil. Thus, from a sample of compressed peat, he extracted by ammonia a substance, from which hydrochloric acid threw down a dark- brown acid, having the composition C 2 ^H 12 O 13 ; the same peat, when digested with solution of caustic potassa, or of carbo- nate of potassa, or carbonate of soda, and precipitated by an acid, gave a substance, having the formula C 24 H lt O 9 ; agree- ing therefore with the ulmic acid of Mulder, in containing an excess of hydrogen, but not reducible to his ulmic group, the excess of hydrogen being very much greater than in ulmic acid, and the equivalent containing only 24 instead of 40 equivalents of carbon. These acids agree with those of Mul- der in their tendency to unite with several bases at once, and to combine with oxygen and nitrogen from the air, pro- ducing ammonia, and acids containing more oxygen than the humic and ulmic acids. Again, from a substance called Pigo- tite, found on certain parts of the rocky coasts of Cornwall, where caves occur, Johnston extracted an acid, represented by the formula C 12 H 5 8 . This acid was obviously formed from the decaying vegetable matter of the soil, and was combined with alumina. It approaches very closely in composition to the crenic acid of Mulder, and is called by its discoverer M u- deseous acid ; on treating its compound with alumina, with nitric acid, Johnston obtained still another acid, the formula of which is Cjg H 5 10 . The above is a very brief and imperfect outline of the pre- sent state of our knowledge regarding the nature of the or- ganic constituents of the soil. It will be seen that the subject is quite in its infancy, and that no attempt can as yet be made to point out a method of analysing a soil with the view of iso- lating any particular acid. The following mode of examina- tion must therefore be adopted : ANALYSIS OF SOILS. 601 Til. Determination of the organic acids of the Sumic group, Digest 1000 grains of the air-dried soil at a temperature under boiling, with a strong solution of carbonate of soda for several hours ; filter, and supersaturate the filtrate with hydrochloric acid. A greater or less quantity of a brown flocculent matter separates ; this is collected on a weighed filter, and washed with distilled water till the wash-water begins to be coloured ; it is then dried at 212, and weighed, after which it is burnt, and the weight of the ash (after subtracting the filter ash) deducted from that of the dry mass ; the difference gives the acids of humus. If a preliminary experiment has shown a con* siderable proportion of these acids, a smaller quantity of soil may be operated upon. vin. Determination of Humus Coal. The soil is boiled for several hours, preferably in a silver dish, with moderately strong solution of caustic potassa, replacing the evaporated water from time to time with distilled water; it is then diluted consi^ derably with water, and filtered. Excess of hydrochloric acid is added to the filtrate, and the precipitated matter collected on a tared filter, washed, dried, and weighed ; now, by the ac- tion of the caustic alkali, the humus coal is converted into humic acid, the difference between the weights obtained re^ spectively in this experiment and in the last, expresses there-, fore the amount of humus coal. ix. Determination of the total amount of Carbon in a $0*7. < This is effected by burning a known weight of the soil (which has been previously treated with dilute hydrochloric acid, and thoroughly washed) with oxide of copper in an ordinary com" bustion tube, and receiving the carbonic acid produced, in caustic potassa, contained in an apparatus in which it can be weighed, The increase in the weight of the caustic potassa is due to car- bonic acid, derived from the carbon of the organic matter in the soil (22 CO 3 = 6 C). According to Schultz, 58 parts of carbon correspond on an average to 100 parts of organic matter in the soil. For almost every purpose, the estimation of the total amount of carbon in the manner here indicated is suificient, and the process is attended with very little trouble, Very little reliance can be placed on the loss of weight which the dried earth suffers by ignition, on account of the tenacity with which some clayey soils retain water. x. Determination of the Nitrogen in a Soil. Combined nu trogen may exist in a soil in three different forms, viz., 3 as 602 QUANTITATIVE ANALYSIS. ammonia, as nitric acid, and as organic matter. If the ana- lysis be intended to be of a rigorous character, it will, of course, be necessary to examine the soil specially both for ammonia and for nitric acid ; in almost every case, however, a determina- tion of the percentage amount of nitrogen conveys all the in- formation that is desired on this head. It is effected by burn- ing a known weight of the air-dried soil in a hard glass tube with soda lime (Will and Varrentrapp's process, p. 555). The ammonia condensed in the acid may be estimated either by a standard alkaline solution, or by converting it into the double chloride of platinum or ammonium. A large com- bustion-tube is required, it being ncessary to burn a consider- able quantity of the soil. xi. Analysis of the residue (A) insoluble in Hydrochloric Acid. Evaporate off the water, or draw it off with a pipette ; transfer to an evaporating dish, and dry the residue tho- roughly on the sand-bath, after which remove it to a pla- tinum or porcelain crucible, and ignite it gently. Digest with excess of sulphuric acid mixed with a little water, until nearly the whole of the acid is driven off (this must be done under a hood). Let the mixture cool, and then add water. Transfer to a filter, and well wash the residue (which will con- sist generally of pure siliceous sand), dry, ignite, and weigh. The solution containing alumina, with perhaps a little oxide of iron, is precipitated by ammonia, the precipitate is collected on a filter, washed, dried, ignited, and weighed. The weight is entered in the tabular statement of results of the analysis as " alumina not soluble in hot hydrochloric acid." The accuracy and care with which the whole of these pro- cesses have been conducted, is tested by adding together the weights of the several substances that have been separately obtained. If this sum does not differ more than two per cent, from the weight of the soil employed, the results may be con- sidered as deserving of confidence. One of the points in which a beginner is most likely to err, is in the washing of the several precipitates. As this is a tedious operation, he is very likely to wash them at first only imperfectly, and thus have an excess of weight when his quantities are added together, whereas a small loss is unavoidable. The precipitates should always be washed with distilled water, and the washing continued till a drop of what passes through leaves no stain when dried upon a piece of glass. ANALYSIS OF LIMESTONES. 603 211. ANALYSIS or LIMESTONES. The importance of these minerals in agriculture and in the arts, renders the determination of their general composition a matter of frequent occurrence to the practical chemist. The following is the method of procedure: Dissolve 100 grains of the specimen (which should be well selected and averaged) in dilute hydrochloric acid, and evaporate the so- lution to dryness on the sand-bath ; moisten the residue with hydrochloric acid, let it stand for half an hour, then add boiling distilled water, and filter through a filter which has been previ- ously dried at 212 in the water oven, and weighed ; wash on the filter till the wash water passes through tasteless, then return the filter with its contents to the water oven, and dry, until it ceases to lose weight. The increase in weight is the silicic acid, sand, insoluble clay, and (perhaps also) organic matter in .100 grains of the limestone. It is rarely necessary to do more than de- termine the total amount of silicic acid in this insoluble ref sidue, which is done by fusing it in a platinum crucible with four times its weight of mixed carbonates of potassa and soda, and evaporating the fused mass carefully to dryness with ex- cess of hydrochloric acid ; when perfectly dry it is allowed to cool, then moistened with strong hydrochloric, boiled up with water, and filtered ; the residue on the filter is pure silicic acid; it is well washed with boiling distilled water, dried, and ignited in a platinum crucible. The crucible should be weighed with its cover on, as silicic acid is apt to take water from the air. In the great majority of cases it is sufficient to determine the weight only of the ignited residue of a lime- stone, insoluble in hydrochloric acid, without submitting it to analysis. The filtrate from the insoluble residue, with the washings, is divided accurately into three equal parts, each part representing one-third of 100 grains of the limestone. One part is mixed in a flask with a little concentrated nitrio acid, or chlorine water, to peroxidize any iron that may be present, heated nearly to boiling for some time, and then allowed to cool ; it is now transferred to a beaker or other convenient vessel that admits of being covered with a glass plate, and ammonia added in slight excess, the precipitate which falls is alumina (with a little silicic acid) and perhaps sesquioxide of iron and oxide of manganese ; it is collected on a filter, washed, dried, ignited, and weighed. Except for 604 QUANTITATIVE ANALYSIS. scientific (geological) objects, it is rarely necessary to submit this precipitate to analysis. In the filtrate from the precipitate by ammonia, the lime and magnesia are determined in the usual manner. La order to determine the total amount of car- bonic acid, the excellent method proposed by SchafFgottsch may be followed. A known weight of the limestone is heated intensely in a platinum crucible, with four times its weight of recently fused borax; the loss of weight when cold, indicates the amount of carbonic acid, plus the water, which the limestone may contain. The amount of the latter is determined with sufficient accuracy for all practical purposes by keeping 100 grains of the mineral in the air-bath heated to 300, until its weight is perfectly constant ; or the carbonic acid may be determined in a weighed quantity of the limestone finely pulverized, by the process of Will and Fresenius. It will be understood that for minera- logical purposes, the residue insoluble in hydrochloric acid and the precipitate by ammonia must be submitted to minute ana- lysis, If the limestone be intended to be used as a flux (in the iron blast furnace, for example) it may be necessary to examine the residue insoluble in hydrochloric acid for pyrites ; a portion of it should be fused with a mixture of carbonate of soda and nitrate of potassa ; the fused mass digested with dilute hydrochloric acid, and evaporated to perfect dryness ; the dry residue moistened with a few drops of hydrochloric acid, then diluted with distilled water, filtered, and the sulphuric acid in the filtrate determined as sulphate of baryta by the addition of chloride of barium. . 212. ANALYSIS OP CLAYS. For technical purposes much information respecting the value of a sample of clay may be derived from a mechanical analysis of it (see ANALYSIS or SOILS) ; by this is learnt what are its compound parts, how much is sand, and how much clay proper. The chemical analysis is conducted as follows : i. Moisture, combined water, and organic matter. The first, is determined by drying 100 grains of the sample finely pounded, in the water oven, and the two latter, by heating it to redness for a considerable time in a platinum crucible. ii. Determination of the constituents. About 30 grains of the sample are digested for several hours in a platinum crucible with concentrated sulphuric acid ; the best method of doing this, is to invert a funnel over the mouth of the crucible, and ANALYSIS OF CLAYS. 605 place it on a sand-bath underneath a hood. After the diges- tion has gone on for two or three hours, the funnel is removed, and the greater part of the acid expelled by evaporation ; dis- tilled water is now added, and the insoluble residue filtered off, this consists of silicic acid (which was probably present in the clay in the form of hydrate) and sand. It is boiled two or three times with a strong solution of soda which dissolves the silicic acid and leaves the sand ; the alkaline solution is diluted with water, filtered off from the sand, the filtrate supersaturated with hydrochloric acid, and then evaporated to perfect dryness. When cold it is washed on to a filter, and there treated with repeated affusions of boiling water, after which it is dried, ig- nited, and weighed in a covered platinum crucible ; the weight of the sand is also determined after ignition. The filtrate from the sand and silicic acid, is warmed with a little nitric acid, and then precipitated by slight excess of pure ammonia ; it is filtered as rapidly as possible, and the precipitate on the filter consisting of alumina and peroxide of iron is well washed with hot water ; the filtrate may contain lime and magnesia which are determined in the usual manner, The pre- cipitate of alumina and oxide of iron is dissolved off the filter with hot hydrochloric acid, the filter being thoroughly washed ; the filtrate and washings being mixed, are accurately divided into two equal parts, either by weighing or measuring, the for- mer being the most accurate, one half is precipitated by slight excess of ammonia, filtered, and the precipitate washed, dried, ignited and weighed ; the weight being calculated for the whole quantity of solution, gives the total weight of alumina and oxide of iron ; the other half of the solution is nearly neutra- lized with pure carbonate of soda, and then boiled in a silver or porcelain dish with pure caustic potassa, this is repeated twice or even three times, the alumina is hereby dissolved and the sesquioxide of iron (with perhaps a trace of oxide of man- ganese) is left behind. The insoluble residue is washed with boiling water, redissolved in hydrochloric acid, precipitated by ammonia, washed, dried, ignited, and weighed ; its weight being calculated for the whole solution gives the amount of sesqui- oxide of iron with sufficient accuracy for most practical pur- poses ; but it is not exact, it being almost impossible to accom- plish the perfect separation of alumina from sesquioxide of iron by caustic potassa. In cases where great accuracy is required, the following method of Bivot may be adopted. The mixed 606 QUANTITATIVE ANALYSIS. precipitate of alumina and sesquioxide of iron is ignited and accurately weighed, it is placed in a small porcelain boat, and heated to redness in a porcelain tube, through which a stream of dry hydrogen gas is passing, the oxide of iron is hereby re- duced to the metallic state, the alumina remaining unaltered. "When no more water is formed, the reduction of the iron is known to be complete, and the tube is allowed to get cold, the current of hydrogen continuing to pass through it ; the little boat is then removed, and its contents digested in very dilute nitric acid, which dissolves only the iron, from which it is after- wards precipitated as sesquioxide, by means of ammonia. It is almost needless to observe that this troublesome process is only to be followed in cases where extreme accuracy is required, as in the analysis of minerals for mineralogical purposes. The process recommended (p. 315) may likewise be employed for the separation of these two oxides. 213. ANALYSIS or MANUBES. (1.) Guano. The extensive use of this complex manure, its high price, and the consequent inducement held out to un- principled vendors to sophisticate it, render it a matter of im- portance that the real value of a sample, as a fertilizing agent, should be capable of being tested, without submitting it to the costly and tedious process of an elaborate analysis. "Imposi- tions," says Professor Johnston ('Elements of Agricultural Chemistry'), ''have been practised on unwary farmers, by selling as genuine guano, artificial mixtures, made to look so like guano, that the practical man in remote districts is unable to detect it. A sample of such pretended guano, which had been sown in the neighbourhood of Wigtown, and had been found to pro- duce no effect upon the crops, was examined in my laboratory, and found to contain, in the state in which it was sold, more than half its weight of gypsum, the rest \)Q\i\gpeat or coal ashes, with a little common salt, sulphate of ammonia, and either dried urine, or the refuse of the glue manufactories, to give it a smell. I could not satisfy myself that it contained a particle of real guano." Again, he says, "Eour vessels recently sailed hence (Liverpool) for guano stations, ballasted with gypsum (plaster of Paris) ; this substance is intended for admixture with guano, and will enable the parties to deliver from the vessel a nice- looking and light-coloured article. Parties purchasing guano are very desirous of having it delivered from the vessel, as they ANALYSIS OF MANURES. 60? believe that they thereby obtain it pure. The favourite material for the adulteration of guano is umber, which is brought from Anglesea in large quantities. The rate of admixture, we are informed, is about 15 cwts. of umber to about 5 cwts. of Peru- vian guano, from which an excellent-looking article, called African guano, is manufactured." In selecting a guano the following points ought to be at- tended to by the farmer, (Anderson, ' Elements of Agricul- tural Chemistry ') : 1. The guano should be light-coloured and dry, cohering very slightly when squeezed together, and not gritty. 2. It should not have too powerful ah ammoniacal smell, and should contain lumps which, when broken, appear of a paler colour than the powder. 3. A bushel should not weigh more than from 56 to 60 pounds. These characters are, however, imitated with great skill, so that they cannot be implicitly relied upon, and they are appli- cable to Peruvian guano only. The following are the chief adulterations in guano, which chemical analysis alone can de- tect : A sort of yellow loam, very similar in appearance to guano, sand, brick-dust, chalk, umber, gypsum, common salt, and occasionally also, ground coprolites and inferior guano. These substances are rarely used singly, but are commonly mixed in such proportions as most closely to imitate the colour and general appearance of the genuine article. The adulteration always takes place in large towns, because it is only there that facilities exist for obtaining the necessary materials, and carry- ing it out without exciting suspicion. " The sophisticated ar- ticle," observes Anderson, " then passes into the hands of the small country dealers, to whom it is sold with the assurance that it is genuine, and analysis quite unnecessary. In other in- stances, adulterated and inferior guanos are sold by the ana- lysis of a genuine sample ; and sometimes an analysis is made to do duty for many successive cargoes of a guano which, though all obtained from one deposit, may differ excessively in composition. In the case of a Peruvian guano, a complete analysis is not necessary ; but an experienced chemist, by the application of a few tests, can readily ascertain whether the sample is genuine. Where the difference in value between different samples is required, a complete analysis is necessary, and this is indispensable in the case of inferior guanos." 608 QUANTITATIVE ANALYSIS. Analysis. The following points should be attended to in the chemical examination of this manure, the sample being mixed as uniformly as possible, and kept in a well-stoppered bottle : I. Determination of the Water. Dry 100 grains in the water- oven till it ceases to lose weight. A great difference is here found, even in genuine guanos, the margin extending from 7 to 20 per cent. ii. Determination of total amount of fixed constituents. In- cinerate 50 or 100 grains, at a low red-heat, in a porcelain or platinum crucible, till the weight remains constant ; the residue (which will, of course, vary greatly according as the guano is ammoniacal or phosphoric) should be white or grey ; a yellow or red colour indicates a probable adulteration with loam, brick- dust, or sand. in. Determination of the total amount of Ammonia. The alkalimetrical method of Peligot, gives excellent results ; it involves, however, the operation of distillation. This is ob- viated by the following modification of the process, which is particularly applicable to the analysis of guano, and ammonia- cal manures generally. A weighed quantity (say 20 grains) of the guano is placed in a shallow glass vessel, about four inches in diameter, standing in a plate filled with mercury ; on a glass triangle, laid across the dish containing the guano, is placed a second dish, containing a measured volume of standard sul- phuric acid ; a few drachms of water are then thrown on the guano, and then some milk of lime, by means of a pipette; the whole is now covered with a beaker, and allowed to remain for two or three days, or until reddened litmus-paper is no longer affected when introduced within the glass, The whole of the ammonia has now been expelled from the guano without the aid of heat, and has been absorbed by the acid, the quantity of which left free, is determined by means of a standard solution of soda, and the amount of ammonia calculated from the re- sult. IV, Determination of the total amount of Nitrogen. The fol- lowing method, though only an approximative one, recommends itself by the facility with which it is executed, and the rapidity with which it enables the operator to ascertain the comparative values of different specimens of guano. It is founded on the fact, that when guano is treated with a solution of chloride of lime, the nitrogen both of the organic matter and of the am- moniacal salts is evolved as gas. Instead of collecting and ANALYSIS OF MANURES. 609 Buring the gas evolved, which would be scarcely practicable on account of the violent effervescence, the volume of water which is expelled by the gas is ascertained by means of the simple ap- paratus represented in the figure ; it con- sists of a flask capable of containing about half a pint, provided with a narrow gas de- livery tube, bent twice at right angles. One limb, rather the shorter of the two. is passed air-tight through the cork of the flask, and bent upwards, to prevent as far as possible the escape of bubbles of gas. This tube descends nearly to the bottom of the flask. A second very narrow short tube is also passed through the cork, and serves for the escape of air when the cork is in" troduced. The longer limb of the delivery tube dips into a tall cylinder or tube, which *& is graduated to cubic centimetres or cubic inches. The flask is half filled with solution of chloride of lime, carefully prepared and kept in a dark place in a closed vessel. About fifteen or twenty grains of guano are then weighed in a small glass vessel which may be the end of a test tube in which a few small shot have been placed in order that it may float. With the aid of the iron- wire handle shown in the figure, the tube is let down so as to float upon the surface of the solution of chloride of lime, the cork with the tube is then tightly adjusted, the orifice of the smaller tube closed with wax, and the flask shaken so that the little vessel may fill and sink ; a volume of liquid equal to the nitrogen evolved from the guano, then flows into the gra- duated cylinder : when no more liquid passes over, the cylinder is depressed so as to bring the liquid to the same level as that in the generating flask. The wax plug is then removed, the cork withdrawn, and the liquid still contained in the de- livery tube is allowed to run into the cylinder where the whole is carefully measured. Fifteen grains of good Peruvian guano evolve between four and five cubic inches of gas. A more exact determination of the amount of nitrogen is obtained by burn- ing a known quantity of the manure with soda lime, accord- ing to Will and Varrentrapp's method as modified by Peligot (p. 555). Y. Determination of the constituents, soluble and insoluble in Water. -Digest 150 or 200 grains with several times its volume 610 QUANTITATIVE ANALYSIS. of water ; throw on a previously dried and weighed filter, and wash till a few drops of the wash-water collected in a test tube are not rendered turbid by the addition of ammonia and chloride of calcium ; the filter, with the washed guano, is then thoroughly dried in the water oven and weighed. The differ- ence, less the proportion of water found in (i.) gives the amount of soluble constituents. Burn the insoluble matter and weigh the ash which represents the amount of fixed insoluble salts. vi. Sand. Dissolve the ash in dilute hydrochloric acid : if it effervesce strongly, there is reason to suspect an adultera- tion of chalk ; a good guano only gives rise to a slight effer- vescence. The residue, after being well washed, dried, and gently ignited, is weighed as sand ; it should not exceed 2 per per cent. vn. Determination of Phosphate of Lime. On adding slight excess of ammonia to the hydrochloric solution of the ash, this salt falls ; it is filtered off, washed, dried, ignited, and weighed. The filtrate from the precipitate by ammonia should give, on the addition of oxalic acid, only slight indications of lime. If, however, a considerable precipitate occurs, the guano has certainly been adulterated with chalk, the amount of which must be determined. Should it be desired to ascertain the exact amount of phosphoric acid, the hydrochloric solution of the #sh should be nearly neutralized by ammonia, then oxalate of ammonia added, and the solution boiled, acetate of ammonia in excess being added to the boiling liquid. On standing, oxa- late of lime precipitates, and is removed by filtration, and the phosphoric acid in the clear filtrate is determined as pyrophos- phate of magnesia (sect. 159). vin. Determination of Alkaline Salts. For practical pur- poses it is only necessary to evaporate the filtrate from the phosphate of lime to dryness, and weigh the residue after gentle ignition, to expel the ammoniacal salts, calculating it as " al- kaline salts ; " when however chalk has been found as an adulteration, it must of course be previously removed by oxalic acid. By way of reference, we append some analyses of the highest and lowest qualities of some genuine guanos, on the authority of Professor Anderson ( e Elements of Agricultural Chemistry,' p. 210). ANALYSIS OF MANURES. 611 Water Angamos. Peruvian. Bolivian. Highest. Lowest. Highest. Lowest. Highest. Lowest. 12-60 65-62 10-83 7-50 3-45 7-09 50-83 8'70 1630 17-08 10-37 55-73 25-20 7-50 1-20 21-49 46-26 18-93 10-64 2-68 11-53 11-17 62-99 9-93 4-38 16-20 12-86 52-95 13-83 4-16 Organic matter and "1 ammoniacal salts j Phosphates Alkaline salts Sand Ammonia 100-00 25-33 100-00 17-15 100-00 18-95 100-00 14-65 100-00 1-89 100-00 2-23 The following is the composition of (so-called) Peruvian guano, which has evidently however been adulterated, and which is worth less than half the price of the genuine ma- nure : Water 12*06 Organic matter and ammoniacal salts . 34- 14 Phosphates . 22-08 Gypsum ll'OS Alkaline salts 12'81 Sand . . . 7-83 100-00 (2.) Superphosphate of Lime. " The deliberate adulteration of superphosphates," observes Professor Anderson, "that is, the addition to it of sand or similar worthless materials, I believe to be but little practised. The most common fraud consists in selling as pure dissolved bones, articles made in part, and sometimes almost entirely, from coprolites. Occa- sionally refuse matters are used, but less with the intention of actually diminishing the value of the manure, than for the purpose of acting as driers. It is said that sulphate of lime is sometimes employed for this purpose, but this is rarely done, because that substance is always a necessary constituent of superphosphate in very large quantities ; and as farmers look upon it with great suspicion, all the eiforts of the manufactu- rers are directed towards reducing its quantity as much as possible." 612 QUANTITATIVE ANALYSIS. There is no manure which requires greater vigilance on the part of the purchaser than superphosphate of lime, because of the great variations in quality ; though in consequence of in- creased competition, and the process of manufacture being better understood, it is much better than it was a few years ago. Analysis. The following are the points to be attended to in the chemical examination of this manure : i. Determination of the Water. 50 grains are dried in the air bath at a temperature of about 300 F. until the weight remains constant; the loss represents the amount of moisture absorbed by the superphosphate, and the water in the gypsum. ir. Determination of the relative proportion of soluble and insoluble constituents. 150 grains of the sample are boiled up with water, and allowed to settle ; the decanted fluid is thrown on a filter, and the residue again and again boiled with water, until there is no longer an acid reaction. The insoluble matter is then dried in the air bath at about 300 F., and weighed. Of course, any solid particles that may have been retained on the filter are added ; this is best done by drying and burning the filter, and adding the ash. in. Determination of the organic matter in the insoluble portion. Moisten with a little nitric acid, to peroxidize any iron that may be present, then ignite at a low red-heat in a platinum crucible, placed sideways, so as to allow air to have access ; when cold, weigh. The loss represents the amount of organic matter, plus that of any ammoniacal salts that may have been present. iv. Analysis of the insoluble matter. Boil for some time with dilute hydrochloric acid, dilute largely, and boil again; filter, wash, dry, ignite, and weigh the residue on the filter, which is sand and clay. Mix the filtrate and washings well together, and divide into three parts , b, c. In a de- termine the sulphuric acid (from the gypsum) by chloride of barium ; mix b and c together, add oxalate of ammonia, and considerable excess of acetate of ammonia, and boil. All the lime is hereby precipitated as oxalate ; it is collected on a filter, and estimated as sulphate. To the filtrate from the oxalate of lime add tartaric acid and ammoniacal sulphate of magnesia,* which precipitates the phosphoric acid as am- monio-magnesian phosphate. Combine the sulphuric acid found in a with the quantity of lime requisite to form gypsum * It is convenient to keep on hand a etock of iartariaed ammoniacdl ANALYSIS OF MANUKES. 613 (40 sulphuric acid + 28 lime) ; add the remainder of the lime to the phosphoric acid, and enter it as " insoluble phosphate." v. Analysis of the soluble portion. Divide into three equal parts , ft, c ; transfer a to a platinum dish, and evapo- rate gently on the sand bath, adding, a little at a time, thin milk of lime, until a piece of red litmus-paper is turned faintly but distinctly blue, showing an alkaline reaction ; continue the evaporation to perfect dryuess, transfer to the air bath, dry at 320 F., and then weigh ; ignite, and weigh again. The difference between the results of the two weighings expresses the quantity of organic matter in the aqueous solution. Boil the ignited residue with lime water, and then with distilled water for a considerable time ; remove the sulphuric acid from the filtrate by chloride of barium, and then the excess of the barium and calcium salts by carbonate of ammonia, and filter. The filtrate contains nothing but the alkalies, which are deter- mined in the usual manner. Determine the sulphuric acid in I by chloride of barium ; evaporate c to dryness in a platinum dish with excess of carbonate of soda, and add a little nitre ; ignite the residue, rinse into a beaker, and dissolve in hydro- chloric acid with the aid of a gentle heat. To the clear solu- tion add ammonia, and then acetic acid in excess. If any phosphate of sesquioxide of iron be here precipitated, filter it off, and divide the filtrate into two equal parts, in one deter- mine the phosphoric acid by the acetate of protoxide of ura- nium; the liquid should be boiled after the addition of this reagent, and the precipitate allowed to subside. Fresenius recommends the addition of a drop or two of chloroform imme- diately after the precipitation, and when the liquid has cooled a little, giving the mixture a vigorous shake, which materially assists the precipitation of the finer particles ; these are col- lected on a filter, the bulk of the precipitate being washed by i decantation ; the well-washed precipitate is dried and ignited ; 1 the filter is also dried and burnt, and its ashes added to the ; ignited precipitate, 100 parts of which contain 19*78 parts of phosphoric acid. sulphate of magnesia for phosphoric acid determinations. The following proportions may be used : 2 pints of water. 225 grains of tartaric acid. 80 anhydrous sulphate of magnesia. 250 chloride of ammonium. PART II. 2 S 614 QUANTITATIVE ANALYSIS. Those engaged in the frequent analysis of superphosphate of lime and other phosphatic manures, will find it convenient to estimate the phosphoric acid by one of the volumetric pro- cesses described in section 193. Liebig's method by perchloride of iron, as modified by E. Davy, gives excellent results ; on the whole, however, preference must be given to the acetate of uranium method (p. 455). Determine the lime and magnesia in the other portion of c, in the usual manner. vi. Determination of the Nitrogen.- The process of Will and Varrentrap, as modified by Peligot (p. 555), is to be followed : The composition of superphosphates must necessarily vary to a great extent, and depends not only on the materials, but on the proportion of acid used for solution. The following analyses by Dr. Anderson ('Elements of Agricultural Che- mistry ') illustrate the composition of some good samples made from different substances : Water Bones alone. Bone Ash. Chiefly Coprolites. Mixtures, containing Salts Ammonia. 7-74... 7-79 17-83... 21-69 13-18 .. 9-87 10-31... 21-17 46-00... 35-30 1-46... 0-94 3-48... 3-24 5-33.., 10-40 6-94... 4-92 *21'35... 23-09 5-92... 6-08 56-16... 47-78 traces 3-30... 4-30 5-90... 10-17 5-10... 4-13 *12-24. . 13-75 16-90. . 0-17 52-39. . 62-62 2-47. . 0-96 6-00. . 8-20 7-07... 15-: 9-87... 13- *17'63... 12-' 12-60... 8'* 49-77... 45-:< 0-06... !( 3-00... 2-1 Organic matter anch ammoniacal salts 3 *Biphosphateoflime Insoluble phosphates Sulphate of lime . . . Alkaline salts Sand Nitrogen 100-00... 100-00 2-11... 3-01 20-57... 15-39 100-00... 100-00 0-23... 0-31 33-33... 36-02 100-00... 100-00 0:11... 0-57 19-10... 21-43 100-00... HXH 1-28... 1-1 27-50... 19-1 *Equivalent to so-| luble phosphates) Superphosphates made from bones alone are generally dis- tinguished by a large quantity of ammonia, and rather a low percentage of biphosphate of lime. This is owing to the difficulty experienced in making the acid react in a satisfactory manner on bones, the phosphates being protected from its action by the large quantity of animal matter, which, when moistened, swells up, fills the pores, and prevents the ready access of the acid to the interior of the fragments. Super- ANALYSIS OF MANURES. 615 phosphates from bone ash, on the other hand, contain a mere trifle of ammonia, and when well made, a very large quantity of biphosphate of lime. Their quality differs very greatly, and depends, of course, on that of the bone ash employed, which can rarely be obtained of quality sufficient to yield more than 30 or 35 per cent, of soluble phosphates. Coprolites are seldom used alone for the manufacture of superphosphates, but are generally mixed with bone ash and bone dust. Mixtures containing salts of ammonia, flesh, blood, etc., are also largely manufactured, and some are now produced containing as much as 4 or 5 per cent, of ammonia, and the consumption of such articles is largely increasing. The following analyses (Anderson) illustrate the composi- tion of some inferior varieties of superphosphates, in the manufacture of which the quantity of sulphuric acid has been reduced, and consequently containing a smaller proportion of soluble phosphates ; these manures are sold in the market for much more than they are really worth : "Water .......... 21-60 5'37 7-19 Organic matter and ammoniacal salts 11'62 13*91 8'80 Biphosphate of lime 2'98 2'02 6'42 Insoluble phosphates 2570 15'80 14'03 Sulphate of lime 23'60 47'52 51-93 Alkaline salts 10-7Q >73 3'43 Sand . . . 3-80 11-65 8-20 100-00 100.00 100-00 Ammonia . 1-32 0-59 0-33 214. BONES. The manurial value of bones is dependent partly on their phosphates, and partly on the nitrogen they yield. The fol- lowing points should be attended to in estimating the value of of a sample : i. Determination of Water. A weighed quantity is dried in the water-oven, until the weight remains constant. ii. Determination of the total amount of fixed constituents. Introduce into a platinum crucible about 20 grains of the bone in a coarse powder ; place the crucible in an oblique posi- tion over the gas or spirit burner, and allow it to remain until the ash has become quite white ; the burning bone should be 2's2 616 QUANTITATIVE ANALYSIS. turned over from time to time with a platinum wire ; when cold, weigh the residue. in. Determination of the amount of Sand. Digest about 20 grains (or the residue of n.) with dilute hydrochloric acid ; throw on a filter ; wash, dry, ignite, and weigh the residue. iv. Determination of the Lime and Phosphoric Acid. To the filtrate from the sand (m.) add a few drops of ammonia, so as nearly to neutralize the liquor, then add oxalate of ammo- nia, boil, and then excess of acetate of ammonia ; allow the precipitated oxalate of lime to subside, then filter it off, con- vert it into sulphate, and weigh. To the filtrate from the oxa- late of lime, add tartaric acid, and then ainmonio-sulphate of magnesia, and well agitate ; filter off, after standing for a day ; wash the precipitate on the filter with ammoniacal water, dry, ignite, and weigh as pyrophosphate of magnesia. This method recommends itself by its extreme simplicity, though it is not absolutely accurate, as it takes no cognizance of the magnesia, which all bones contain; as, however, it exhibits the total quantity of phosphoric acid (the real fertilizing agent), it gives all the information required for estimating the manurial value of the sample. v. Determination of the Nitrogen. Ignite from 15 to 20 grains with soda lime, precisely according to the method re- commended for the analysis of ammoniacal salts. The result of the analysis will show the exact quantity of ammonia which the gelatine of the bone is capable of yielding by its decom- position. VI. Determination of the amount of Carbonic Acid, in combi- nation with Lime, and perhaps Magnesia. This is effected by the method of Fresenius and "Will. In estimating the value of a manure, the particular element required by the soil to which it is intended to apply it, must be taken into account. "Thus" (observes Anderson), "a farmer who finds that his soil wants phosphates, will look to the manure containing the largest quantity of that substance, and possibly not requiring ammonia, will not care to estimate at its full value any quantity of that substance which he may be com- pelled to take along with the former ; but will look only to the source from which he can obtain it most cheaply. It may be as well, therefore, to point out that ammonia is most cheaply purchased in Peruvian guano ; insoluble phosphates in copro- lites ; and soluble phosphates in superphosphates made from bone-ash alone." BONES. 617 The bones used in agriculture are chiefly those of cattle. The following composition of the bones of the cow represents very nearly that of a genuine sample : Organic matter (gelatine) r ; "v. ' f '' . 33'25 Phosphate of lime 55'50 Phosphate of magnesia 3*00 Carbonate of lime . . . .' ! ; i .'.;i' . 3'75 Soda and common salt . i Y ; : . t !i ,' f . 3*50 Chloride of calcium , .' '-*' \ - *' v . TOO 100-00 But bones are met with in commerce in other forms, in which their organic matter has been extracted, either by boiling or by burning ; the latter is very common in the form of the spent animal charcoal of the sugar refiners, which usually contains from 70 to 80 per cent of phosphate of lime. It was formerly considered that the efficacy of bones as a manure depends solely on the quantity of phosphate which they contain, but the benefit derived from the nitrogenous constituent, which in the soil eventually takes the form of ammonia, is now fully recognized. The gelatine or organic part of bones consists of Carbon 50'37 Hydrogen 6'33 Nitrogen 17'95 Oxygen 25'35 100-00 So that, supposing this animal substance to be decomposed in the soil, the quantity of it in 100 Ibs. of dry bone is sufficient to produce upwards of 6 Jibs, of ammonia, as much as is present in 20 Ibs. of sal-ammoniac, or in 30 Ibs. of crystallized sulphate of ammonia. Liebig has indeed advanced the opinion that the ammonia of the atmosphere will give nitrogen enough to the plant, provided the soil be sufficiently supplied with the mineral matters which it requires, but it is well known that the ex- hausted stiffening-liquor for calicoes, which, in Manchester, is largely made from bones by boiling them under pressure, has been applied as a liquid manure to grass lands, with the greatest success. Some years ago, at the suggestion of the late Professor Johnston, a series of experiments was made in Scotland to test the relative value of burnt and unburnt bones, and the general results were these : 618 QUANTITATIVE ANALYSIS. 1st. That whole bones under favourable conditions seldom fail in raising an average crop of turnips. 2nd. That burnt bones in equivalent quantity do not always succeed in raising an average crop. 3rd. That when to the burnt bones, a sufficiency of organic matter in the form , of farm-yard manure was added, then burnt bones produced the usual effects of whole bones. The farmer's surest reliance will therefore be on entire bones, especially if there be a deficiency of organic matter in his soil. 619 CHAPTER XII. ON THE ANALYSIS OF THE ASHES OF VEGETABLE AND ANIMAL SUBSTANCES. 215. ANALYSIS OF ASHES OF VEGETABLES. (1.) Method of Will and Fresenius (' Memoirs and Proceed- ings of the Chemical Society,' vol. ii. p. 179) : First preparation. Plants in a normal and healthy con- dition should be selected, unless the design be to study dis- eases and their causes. All foreign matter, such as dirt, dust, etc., should be carefully removed ; but the plants should not be washed, or certain soluble salts might be extracted. Plants which have been exposed to moist weather should, for the same reason, be rejected. The design of analysing the ashes of plants may be simply to ascertain the amount and nature of their inorganic consti- tuents ; or, the operator may have in view the discovery of the presence or absence of certain substances in the soil, such as alkalies, alkaline earths, and phosphates. For the latter pur- pose an examination of the ashes of all the varieties of plants growing upon the soil must be undertaken ; and when a know- ledge is hereby obtained of the composition of the inorganic constituents of both weeds and of cultivated plants, and when to this is added an acquaintance with the nature of the soluble constituents of the soil itself, the analyst is in a position to determine for what crops the soil in question is best adapted. Woods, herbs, and roots, after being perfectly dried, may be burnt upon a clean iron plate ; leaves, fruit, and seeds may be burnt in a Hessian crucible, by means of charcoal or coke : the crucible should be placed somewhat obliquely in the fire, in order to favour the access of air. During the operation of burning, and especially towards the end of the process, the ashes should be allowed to lie as lightly in the crucible as 620 QUANTITATIVE ANALYSIS. possible, in order that air might circulate freely through them ; they should not, therefore, be stirred together : by at- tending to this, a much whiter ash is procured than would other- wise be the case. After the ashes have been well burnt in the crucible, it is advisable to transfer them to a platinum dish, and to heat them to low redness over a gas or spirit lamp, with con- stant stirring ; they are generally left almost white, but some seeds require a higher temperature than others to rid the ashes entirely of charcoal : care must however be taken not to allow the heat to rise sufficiently high to fuse the alkaline salts, or it will be found afterwards almost impossible thoroughly to burn away the charcoal. The ash is rubbed to a fine powder, and transferred while warm to a well-stoppered bottle. Rose objects to this method of preparing ashes for analysis ; his observations on this subject will be referred to further on. The ash being prepared, the first step is to determine, by means of a qualitative examination, to which class the ashes belong, whether to the siliceous, or to the phosphoric, or to the carbonic class. A small portion is treated with concentrated hydrochloric acid, which generally dissolves it completely, un- less the ash abounds in silica ; if a strong effervescence accom- pany the solution in hydrochloric acid, carbonates of the alkalies or alkaline earths predominate, and the whole of the phosphoric acid present is probably combined with peroxide of iron ; if no (or only moderate) effervescence attend the solution, phosphates predominate ; and if complete solution in hydrochloric acid can- not be obtained, the ash belongs to the siliceous class. To the clear hydrochloric solution, acetate of ammonia is either at once added, or it is first neutralized by caustic ammo- nia, and free acetic acid afterwards added. In most cases a yellowish-white gelatinous precipitate is formed, consisting of phosphate of peroxide of iron. This precipitate is collected on a filter, and ammonia added in excess to the clear filtrate, by which means a fresh precipitate may be obtained ; if it be red, it is per- oxide of iron. The solution is well protected from the air, and allowed to stand for some time$ if no further precipitation take place, then it is known that the ash contains no other phosphate than that previously precipitated ; but should a white deposit gradually form, it consists of phosphates of lime and magnesia, and shows that the ash under examination contains more phos- phoric acid than is combined with the peroxide of iron. It is not necessary to proceed further with the qualitative examina- ANALYSIS OF THE ASHES OF VEGETABLES. 621 tion, unless the operator should wish to test for fluorine, oxide of manganese, iodine, bromine, or any other peculiar substance, the presence of which may be suspected, and in such cases separate portions of the ash must be used for each experiment, as also for determining quantitatively the amounts of carbonic acid in the ashes abounding in that principle, and of the alka- lies in the ashes belonging to the siliceous class. Determination of the quantity of Ashes yielded ~by a given weight of the Plant. This is a problem, the solution of which is of considerable importance : a quantity of solid matter is annually removed from, the soil in the crop taken from it, which loss should be repaired as nearly as possible by the judicious addition of manure. It is the nature of the manure to be fur- nished which the farmer seeks from the analytical chemist ; but the farmer must take his share in the inquiry by informing the analyst as to the weight of the crop which a given surface of soil should yield, and this he can in most cases do with suffi- cient accuracy. The vegetable substance under examination should be dried in the water-bath, or still better in a current of dry air produced by the efflux of water in the apparatus, de- picted in Fig. 65, page 252, till it ceases to lose weight. The quantity of substance employed in this experiment depends on the proportion of its inorganic constituents. Of herbs and seeds, which are in general rich in these matters, from 50 to 100 grains will be sufficient, whilst of woods ten times that amount must be taken. The combustion succeeds best in a muffle, the dried vegetable to be incinerated, being placed in a porcelain or platinum dish, which just fits into the muffle, which is gradu- ally heated, the temperature not being allowed to rise beyond dull redness. Large cast-iron muffles, two or three of which may be heated by one fire, are exceedingly useful methods of applying heat in a laboratory. With the aid of a damper the temperature may be regulated with great accuracy, and kept uniform for any length of time. The time required to prepare an ash fit for analysis is from ten to twelve hours. In cases where a muffle is in the first place inadmissible, a large Hessian crucible may be used. It should at first be covered, and a gentle heat only applied, and the mass being carbonized should be reduced to ash in the muffle. Those ashes which do not effervesce with acids, as the ashes of seeds, may be treated with nitric acid, and again ignited, by which treatment they will speedily be rendered quite white. 622 QUANTITATIVE ANALYSIS. If a very strong heat has been employed, the carbonic acid, and in those ashes which effervesce with hydrochloric acid, will be expelled, but it may be restored by moistening the ashes with solution of carbonate of ammonia, and afterwards again exposing them to gentle ignition ; so also, at a high temperature in con- tact with charcoal, the sulphates (if any be present) may be converted into sulphides, but the reconversion of those com- pounds into the original salts, may be effected by heating the ashes strongly, and for a considerable time, together with pure oxide of mercury. Although the above method of estimating the amount of ashes yielded by a plant is not altogether free from objection, it is, nevertheless, a sufficiently close approxi- mation to truth to answer every practical purpose ; indeed, it is doubtful whether more accurate methods would really be more valuable, since it is found that the amount of ashes yielded by the same plant is not constant. Analysis of Ashes completely soluble in Hydrochloric Acid : Determination of the Silica, Charcoal, and Sand. About 60 grains of the ashes, which have been found to be soluble in hy- drochloric acid, are treated with concentrated acid in a flask held obliquely so as to avoid any loss of the liquid during the evo- lution of the carbonic acid; a gentle heat is then applied until it is evident that everything is dissolved excepting the carbonaceous and sandy particles. The whole is now carefully removed into a porcelain basin, evaporated to dryness over a water-bath, and then heated somewhat more strongly, as is usual in separating silica (see section 195). The mass when cold, is moistened with strong hydrochloric acid, digested for about half an hour with a sufficient quantity of water, and boiled, after which the acid liquor is poured upon a stout filter, which has previously been dried at 212, and weighed. The silica remains on the filter, and if the ashes were not perfectly white and pure, some sand and charcoal also. The filter is washed and dried, and the substance carefully removed from it into a platinum or silver crucible without injury to the paper. This is effected without difficulty, if the matters be perfectly dry, the paper in most cases only retaining so much as to be slightly coloured by the charcoal. The powder is now boiled for half an hour with pure potassa-ley (free from silica), by which the whole of the silica, natural to the ash, will be gradually dissolved, leaving the sand and charcoal unacted upon. The insoluble matter is again collected on the same filter, and after being well washed, it is ANALYSIS OF THE ASHES OF VEGETABLES. 623 dried at 212 till it no longer loses weight. The increase upon the weight of the dried filter is to be estimated as charcoal and sand. The silica in the alkaline solution is determined by add- ing hydrochloric acid in excess, whereby it is precipitated ; the whole is evaporated to dryness, and the further treatment con- ducted in the usual manner. The acid solution originally filtered from the silica, sand, and charcoal, after being well mixed, is divided into three, or more conveniently into four equal portions, one portion being reserved in case of an accident happening with either of the other quan- tities. The division is best effected by means of an accurately graduated tube or cylinder ; the whole of the fluid is collected into the tube or cylinder, the measure of which thus represents the weight of ash experimented upon. The solution is now divided into three or four equal or known portions, the volume of each is noted, and they are labelled respectively with the letters a, b, c, and d. In a the peroxide of iron (oxide of manganese) and the alka- line earths are estimated. In b the alkalies. In c the sulphuric and phosphoric acids. (a.) Estimation of the Phosphate of Peroxide of Iron, Per- oxide of Iron, Oxide of Manganese, and Alkaline Earths. To the solution ammonia is added until the precipitate thereby pro- duced no longer entirely redissolves ; acetate of ammonia is next added, and sufficient acetic acid to render the solution strongly acid. From the form and appearance of the precipitate, it can easily be judged whether it contains phosphate of lime ; if this be the case, more acetic acid must be added. The yellowish- white precipitate which remains consists of phosphate of per- oxide of iron (Fe 2 3 ,P0 5 ) ; its separation from the fluid is as- sisted by gently heating, it is then well washed on the filter with hot water, ignited, and weighed. To the filtered solution, neutral oxalate of ammonia is added as long as a precipitate continues to be formed, and the amount of lime is determined in the usual manner. When it has been shown by the qualitative analysis that, besides phosphate of iron, the ash contains peroxide of iron or oxide of manganese (in which case the presence of the earthy phosphates is very rarely detected), the solution, pre- vious to the separation of the lime, should be supersaturated with ammonia, and precipitated by means of sulphide of am- monium, the two oxides being afterwards separated according 624 QUANTITATIVE ANALYSIS. to one of the methods given in section 169. If the ashes under examination contained earthy phosphates, the solution filtered from the oxalate of lime will contain free acetic acid, if other- wise, there will be free ammonia ; it is next somewhat concen- trated, rendered ammoniacal, treated with a solution of phos- phate of soda, and the precipitate formed, collected, and esti- mated as pyrophosphate of magnesia. (5.) Estimation of the Alkalies. The solution is treated with baryta water until it gives an alkaline reaction, it is then gently heated and filtered. By this means we get rid of all the sul- phuric and phosphoric acids, the peroxide of iron, the magnesia, and part of the lime. The precipitate is washed on a filter as long as the washings render turbid a solution of nitrate of silver. It is next warmed, treated with caustic and carbonate of am- monia, and allowed to stand until the precipitate becomes heavy and granular. The whole is now filtered, and the solid matter washed, after which the solution is evaporated to dryness, and the residue heated to redness in a platinum capsule to expel the ammoniacal salts. What remains consists of the chlorides of potassium or sodium, or more generally of a mixture of the two. The weight being noted, a little water is added, which generally leaves undissolved a trace of magnesia ; this is col- lected on a filter, its quantity subtracted from that of the supposed alkaline chlorides, and added to that of the magnesia, as previously ascertained. The quantity of potassa is deter- mined by means of chloride of platinum in the usual/ way, and that of the soda is calculated from that of the chloride of sodium indicated by deducting the weight of the chloride of potassium from that of the mixed alkaline chlorides ; or the amount of the two alkalies may be determined by the indirect method as di- rected in page 261. (c.) Estimation of the Sulphuric and Phosphoric Acids. Heat nearly to boiling, add chloride of barium in excess, allow the precipitated sulphate of baryta completely to subside, then filter it off, and after well washing, dry, ignite, and weigh. If the ash effervesce strongly on the addition of hydrochloric acid, alkaline or earthy carbonates predominate, and the whole of the phosphoric acid is probably in the state of phosphate of iron. Should there be no effervescence, or an inconsiderable one, then phosphoric acid may exist in some other form of com- bination, and it will be found in the solution filtered off from the phosphate of iron (a), and the process must be modified ANALYSTS OF THE ASHES OF VEGETABLES. 625 thus : having removed the phosphate of iron precipitated by ammonia and acetic acid, boil with oxalate of ammonia, filter off the oxalate of lime, and divide the filtrate into two parts ; in one determine the magnesia by the addition of ammonia and phosphate of soda, and in the other the phosphoric acid by the addition of ammonia, chloride of ammonium, and sulphate of magnesia. Estimation of the Chlorine. A fresh portion (about 15 grains) of the ashes is weighed out and exhausted with hot water, slightly acidulated with nitric acid ; the solution is precipitated with nitrate of silver, following the directions given at page 351. If the ashes should contain appreciable quantities of iodine and bromine, these bodies will be found in the precipi- tated silver salt ; for their 'quantitative estimation, however, a larger quantity of the ashes must be employed. .Estimation of the Carbonic Acid. The amount of this acid is determined with the apparatus of Freseuius and Will (p. 268). Analysis of Ashes abounding in Silica. Ashes of this kind are in general only partially soluble in acids. Their alkalies must therefore be determined in a separate portion of the ash. The chlorine and carbonic acid are determined in the same manner as when the ash is entirely soluble in acids. The quantity of chlorine found in ashes of this class is, however, probably always somewhat less than it should be, since the al- kaline chlorides, when ignited with silica and carbon, undergo a partial decomposition. Estimation of the Silica. Pure potassa or soda ley is poured upon about 60 grains of the ashes, and evaporated to dryness in a platinum or silver dish. The silicic acid compounds are, by this treatment, dissolved, leaving the sand unaffected. The heat should not be so great as to fuse the mass, or some of the charcoal might be oxidated at the expense of the water of the hydrated alkali. Diluted hydrochloric acid is poured upon the mass, the whole evaporated, and the silica, charcoal, etc., deter- mined in the manner already described. The acidulous solution filtered from the insoluble matter is divided into two parts : one is employed for the determination of the sulphuric and phosphoric acids, and the other for that of the peroxide of iron and the alkaline earths by the methods detailed above. Estimation of the Alkalies. A second portion of the ash, (about 50 grains,) is ignited in a platinum crucible with four times its weight of hydrate of baryta. The acid solution which 626 QUANTITATIVE ANALYSIS. remains after separating the silica, etc., is precipitated succes- sively with baryta water and carbonate of ammonia, the alkalies being then obtained in the state of chlorides. The further treatment has already been described. (2.) Rose's method. It has been mentioned above that the method of preparing the ashes of plants for analysis by igniting the vegetable in Hessian crucibles, and continuing the heat until all organic matter is destroyed, has been objected to by Rose, who observes (Chem. Gaz., vol. v. p. 158) that, if the fixed constituents of plants are examined according to the process which he adopted in examining the ashes of ox-blood, results dif- fering entirely from those yielded by the ash analyses hitherto published, may be obtained. The process adopted by the Berlin chemist in his analyses of the ashes of blood was this. The blood was exposed in a covered platinum crucible to a very faint red-heat, then extracted with cold water, and the colourless liquid evaporated to dryness. It was found to consist of alkaline chlorides and carbonates, with very minute quantities of alkaline sulphates and phos- phates. The charred mass, extracted with water, was now treated with hydrochloric acid : the filtered solution did not yield, with ammonia, a very considerable precipitate, which, though it looked almost like pure hydrated oxide of iron, con- tained some phosphoric acid as well as lime and magnesia. In the filtered solution a pretty considerable quantity of oxalate of lime was obtained with oxalate of ammonia, proving the pre- sence of carbonate of lime in the charred blood; and in the liquid separated there was also a small quantity of magnesia. The cinder, after treatment with water and hydrochloric acid, yielded a very considerable quantity of a red-coloured ash, on being burnt in an atmosphere of oxygen. It was in a semifused state, and contained peroxide of iron (which formed the chief part) and earthy and alkaline phosphates. These results differ considerably from those of M. Enderlin, whose method was (Liebig's 'Annalen' and Chem. Gaz., vol. iii. p. 229) to evaporate fresh" blood to dryness, and then to powder and incinerate the residue. The ash thus obtained dissolved in hydrochloric acid without effervescence, w r hence he concluded that the alkalinity of blood cannot be caused by an alkaline carbonate, and that there cannot exist in the blood any alkaline salts with organic acids. The salts found by M. Enderlin in the ashes of blood he states to be tribasic phosphate of soda ANALYSIS OF THE ASHES OF VEGETABLES. 627 (3NaO,P0 6 ), chlorides of sodium and potassium, sulphate of soda, phosphate of lime, phosphate of magnesia, and oxide, with some phosphate of iron, the alkalinity of the blood being pro- duced by phosphate of soda (3NaO,P0 5 ). If the phosphate of soda in the blood were the ordinary phosphate (2NaO,HQ, PO 5 ), it would, according to M. Enderlin, be converted by a red heat into the pyrophosphate (2JN"aO,PO 5 ), the third atom of base (HO) escaping ; but, according to Rose, the ordinary phosphate of soda (2NaQ,HO,P0 5 ) would, at a high temperature, decora- pose carbonate of soda and be converted into tribasic phosphate (3JN"aO,P0 5 ) ; and that, in consequence, the conclusions of M. Enderlin are erroneous. It has been shown also by Marchand (Jour, fur Prakt. Chem., April 6, 1846) that the circum- stance of no carbonates being found in the ash, does not in the least prove that they are are not contained in the blood, and that we are as little justified in admitting unconditionally the presence of tribasic phosphate of soda (3ISaO,PO 5 ) because that salt is found in the ash. Indeed Marchand declares that the admission of the presence of 3NaO,P0 5 in the blood abso- lutely requires the admission of that of carbonate of soda, since it has been proyed that 3NaO,PO 5 is converted, on exposure to a moist atmosphere containing carbonic acid, into 2JNaO,HO P0 5 , andNaQ,00 8 . The same objections which have been urged against the conclusions drawn by Enderlin with regard to the salts exist- ing in the blood, from the analysis of the ashes prepared at high temperatures, have been also applied by E-ose to the conclusions drawn from the results of the analyses of the ashes of plants prepared in the usual manner. It had previously been remarked by Erdmann (Liebig's ' Annalen,' vol. Ivi. and Ivii.) that the mode of preparation of the ashes for analysis has great influence on their apparent composition. His me- thod was to burn the plant or seeds in a muffle furnace, in which in the course of three or four hours, upwards of 200 grains of the most beautiful ash of corn may be obtained. The ashes so prepared, especially in those seeds that are difficult to reduce to ash, generally contain the phosphoric acid in a lower state of saturation than those prepared in a crucible. Thus the ash of rye, which usually contains an alkaline phosphate, yielding, with oxide of silver, a white precipitate, gives a yellow one when it has been prepared by long-continued ignition in a covered crucible ; and biphosphate of potassa, when ignited for a long 628 QUANTITATIVE ANALYSIS. time with carbonized sugar, is converted into a bibasic, and finally even into a tribasic salt. It is evident, therefore, that a reduction of the phosphoric acid has taken place, and the same is the case with the sulphuric acid. These facts induced Rose to undertake an examination of the fixed constituents of certain plants by the same method which he had adopted in the analysis of ox-blood, a method which, although it may be ob- jected to in that it takes more time, furnishes (he says) far more correct results, and gives satisfactory answers to several questions as to how, or in what combinations, the constituents found in the ash were contained in the organic substance. His method is as follows : The organic substance is car- bonized at a low red-heat in a capacious platinum or porcelain crucible, until a strong empyreumatic odour ceases to be per- ceived, and the carbonaceous mass yields no longer any yellow or brown substance to water. The remaining mass is now ex- hausted with hot water, as long as a few drops of the washings leave any considerable residue. The aqueous solution contains the alkaline salts (chlorides, sulphates, and phosphates) which existed as such in the organic substance, and frequently like- wise alkaline carbonate which either pre-existed in the sub- stance, or the alkali in it was combined with an organic acid, or some other organic body which acted the part of an acid to- wards the alkali ; during the process of carbonization this or- ganic acid was decomposed, carbonic acid being formed. The solution is evaporated nearly to dryness, diluted with water, allowed to stand for some time in order to separate the earthy salts (carbonates and phosphates of lime and magnesia) which had been dissolved in the salts of the alkalies ; the precipitate is then filtered off, the filtrate dried up, weighed, and the acids and bases contained therein estimated by the usual methods. The carbonized mass exhausted with water is now digested for some time with hydrochloric acid, and then washed with hot water, until a considerable quantity of the washings ceases to be precipitated by ammonia ; the acid solution contains the earthy phosphates which were present as such, and sesquioxide of iron ; these are precipitated by ammonia, and weighed to- gether with the earthy salts deposited from the aqueous solution. The phosphoric acid in this precipitate is determined by nitric acid and metallic mercury in the manner described (p. 460). The liquid filtered from the earthy phosphates still contains some lime and magnesia, which are successively precipitated as oxalate and phosphate. ANALYSIS OF THE ASHES OF VEGETABLES. 629 The carbonized mass exhausted by hydrochloric acid is yet rich in ash ; it is by degrees completely incinerated in a thin porcelain crucible, through the perforated cover of which oxygen is passed, and the weight of the ashes added to that of the con-. stituents extracted by water and hydrochloric acid. Subse- quently Eose moistened the carbonaceous mass with bichloride of platinum before incinerating : the ash-grey residue thus ob- tained, is ignited in a stream of hydrogen gas, in order entirely to decompose the double salts of the alkaline chlorides with bichloride of platinum ; then digested with hydrochloric acid, and the filtrate analysed in the same manner as the hydrochloric extract of the charcoal. The undissolved residue, consisting of platinum, sand, and silica, is either first boiled with carbonate of soda to remove the silica, and -the platinum and sand then separated by aqua-regia, or the platinum is first extracted with aqua-regia, and the residue is then treated with carbonate of soda. This method is exceedingly tedious and complicated ; it has moreover been shown by Strecker (Ann. Ch. Pharm., Ixxiii. 339), by Staffel (Ann. Pharm. (2) Ixiv. 1, 129), and by Weber (Pogg. Ann., Ixxxi. 402) that as regards chlorine it is incorrect, in consequence of which Rose (Pogg. Ann. Ixxx. 9) modified his original proofs by mixing the charred organic sub- stance with spongy platinum, incinerating it by small portions at a time in a platinum dish. The grey platiniferous mass ob- tained by incineration is heated in an air bath to 120 until its weight is constant. It is then treated with hot water, the aqueous solution evaporated to dryness, and the residue gently ignited and weighed. The dry residue is then divided into dif- ferent portions and analysed in the usual manner. That part of the ash (containing platinum) which remains after extrac- tion with water is treated with hot dilute nitric acid, and thoroughly washed with water acidulated with nitric acid.. The solution contains phosphates of lime, magnesia, and ses- quioxide of iron, together with nitrates of potassa, soda, lime, and magnesia, but no sulphuric acid or chlorine. It is treated with metallic mercury according to the method described (p. 460 for the determination of phosphoric acid, and the bases separated from the latter are determined in the ordinary way. The pla- tinum residue is boiled with solution of caustic potassa which dissolves the silica ; this is separated from the alkaline solution in the usual manner. PART II. 2 T 30 QUANTITATIVE ANALYSIS. (3.) Staffers method (Arch, der Pharm.,lxiv. p. 1). This is a modification of a method previously proposed by Wackenroder. I. The organic substances are carbonized in a crucible, the lid. of which is stuck on with starch paste. The lid has a hole through which the gases escape ; these are inflamed. The charred mass is exhausted with water, and the aqueous extract (through which, if very alkaline, a stream of carbonic acid is passed) is evaporated to dryness and weighed. The solution of the weighed residue is acidified with nitric acid, and the silica being separated in the usual manner, the chloride is determined by nitrate of silver. The excess of silver being removed by hydrochloric acid, the solution is supersaturated with ammonia, and if no precipitate of phosphate of lime take place, the am- moniacal liquor is mixed with oxalic acid, for the separation of the lime. The liquid filtered from the oxalate of lime is preci- pitated with chloride of barium. The mixed oxalate, phosphate, and sulphate of baryta, after being dried and calcined, is di- gested with very dilute nitric acid, which leaves the sulphate of Baryta undissolved. To the filtered liquid a slight excess of ammonia is added, and it is allowed to stand for some time. If any phosphate of baryta should be deposited, it is collected, dried, and calcined ; and the phosphoric acid calculated accord- ing to the formula (5BaO + 2PO 6 ) containing 27'16 per cent, of phosphoric acid. There is still the potassa and soda to de- termine in the liquid filtered from the oxalate, sulphate, and phosphate .of baryta. The excess of baryta is removed by car- bonate of ammonia, and the alkalies are separated and estimated in the usual manner. ii. The charred and exhausted mass is reduced to ash in a Hessian crucible, placed in a slanting position. The ashes are weighed, and extracted repeatedly with water. The solution is divided into several parts .: in one part, the chlorine is esti- mated; in another, (acidified with nitric acid) the silica; in a third the sulphuric and phosphoric acids are thrown down by chloride of barium, and separated as above described. The so- lution freed from the mixed baryti.c precipitates serves for the .determination of the alkalies. In a fourth portion the lime and inagnesia are estimated. in. The residue, insoluble in water, is digested with hydro- chloric acid. The solution is evaporated to dryness, and again dissolved in warm hydrochloric acid ; the silica and sand in the insoluble residue are separated by caustic potassa. The hydro- ANALYSIS OF THE ASHES OF VEGETABLES. 631 chloric solution is divided into two equal parts : in one, the alka- lies are determined; the other serves for the estimation of the earths, peroxide of iron, and phosphoric acid, which is effected according to the following plan : The solution is heated to boiL- ing in a flask, mixed with acetate of soda and some acetic acid ; the resulting precipitate, consisting of perphosphate of iron and phosphate of alumina, or of the first only, is collected on a filter, dried, calcined, and weighed. To separate the peroxide of iron from the alumina, the weighed precipitate is dissolved in hydrochloric acid, diluted with water, and heated to boiling for some time, with the addition of caustic potassa. The separated peroxide of iron is collected on a filter and calculated for phos- phate of iron according to the formula (Fe 2 3 , P0 5 ) proposed by Wackenroder. The alumina is precipitated from the fil- tered alkaline liquid by acetic acid as phosphate of alumina (A1 2 O 3 ,P0 5 ) dried, calcined, and weighed. In this case we must not omit to examine the mixed precipitate obtained by acetate of soda, for phosphate of lime, as frequently some is found in ashes which contain much phosphate of lime. If this preci- pitate be again dissolved in dilute hydrochloric acid, diluted with water, and heated to boiling with the addition of acetate of soda, it is obtained free from every trace of phosphate of lime. After the alumina and peroxide of iron have been precipitated by acetate of soda from the acid extract of the ash in the state of phosphate, the separated liquid is mixed with a solution of perchloride of iron (the amount of iron in which is known, and equal in weight to about the tenth or twentieth part of the entire weight of the ash), in order to determine the phosphoric acid still contained in it, the liquid itself is diluted with a con- siderable amount of water, heated to boiling in a flask with the addition of 6 or 7 grains of chlorate of potassa, nearly neu- tralized with carbonate of soda, and the whole of the iron added to the liquid thrown down as basic phosphate by the further addition of acetate of soda. The precipitate is collected on a filter, washed, dried, and ignited in a platinum dish after hav- ing been moistened with a few drops of nitric acid, and in this way a basic phosphate of iron is obtained perfectly free from protoxide. The liquid filtered from the basic phosphate of iron is heated to boiling in a flask with carbonate of soda, in order to precipitate lime, magnesia, and protoxide of manganese as car- bonates. The filtered liquid is mixed with some phosphate of soda and ammonia, and the crystalline precipitate of ammoiiior 2 IT 2 632 QUANTITATIVE ANALYSIS. phosphate of magnesia, obtained after twelve or twenty-four hours, added to that portion to be mentioned presently. The mixed precipitates of earthy carbonates are, after drying, heated to strong redness in a porcelain crucible, then dissolved in cold nitric acid. If any manganese be present it is left as Mn 3 O 4 . The separated liquid is mixed with oxalate of potassa, the oxalate of lime calcined after drying, and calculated from the amount of carbonate. After separating the oxalate of lime, the liquid is mixed with ammonio-phosphate of soda, and the crystalline precipitate obtained, after standing, is mixed with that previously obtained, calcined, and the weight of magnesia calculated from it. The other half of the liquid is examined for alkalies according to the usual method. (4.) Way and Ogston's method (' Journal of the Royal Agri- cultural Society of England,' vols. viii. and ix.) : I. Estimation of the Water in the vegetable. Grains are dried entire ; bulbs, leaves, etc., are cut into small pieces ; the tempe- rature employed is at first about 150 or 160 R, at which it is kept for a fortnight ; the vegetable is then weighed, and dried again for another week or fortnight ; the drying is com- pleted in the water oven at 212, till there is no longer a loss of weight. IT. Estimation of the percentage of Ash. From 100 to 200 grains of seeds, and from 60 to 80 grains of dried bulbs, leaves, etc., are burnt in a small platinum dish (watch-glass shape) over the air-flame of gas, the burning being continued till the ash is quite white ; with ashes that fuse but still con- tain little silica, the burning is stopped before fusion begins. The weighed ash is treated with dilute hydrochloric acid; the charcoal (which seldom amounts to T ^ of a grain) is col- lected, washed, dried, weighed, and deducted. In all specimens burnt for analysis, the estimation is checked by the larger quantity. ill. Burning the Ash for analysis. Prom 4000 to 6000 grains of seeds carefully cleaned, and from 1000 to 2000 grains of dried bulbs, leaves, etc., are burnt in a large platinum dish, over a Black's furnace, supported in the opening at the top, by a solid iron ring ; the temperature never exceeds dull red- ness ; there is perfect access of air, the dish being 8 or 10 inches broad, and 1 J inch deep only. The ash is stirred, when necessary, with a thick platinum wire ; the burning is discon- tinued as soon as the charcoal ceases to glow at a dull red-heat. ANALYSIS OF THE ASHES OF VEGETABLES. 633 The ash is in every case examined, in case of any trace of foreign matter having been overlooked ; it is then powdered, dried, and weighed. iv. Estimation of Chlorine and Sulphuric Acid. 15 or 20 grains are heated with water, and a small quantity of nitric acid added ; the solution is filtered, and precipitated with nitrate of silver for chlorine; the excess of silver being thrown down from the precipitated chloride of silver by hydrochloric acid, the sulphuric acid is precipitated from the filtrate by chloride of barium. v. Estimation of Carbonic Acid. This is effected in 15 or 20 grains of the ash by one of the well-known methods. (See CAKBONIC ACID.) vi. Estimation of the Silica. If the ash be siliceous, or con- tain much charcoal, 40 or 50 grains are mixed with exactly its own weight of finely powdered and pure nitrate of baryta ; the mixture is thrown by a platinum knife, by small portions, into a large platinum crucible gently heated by a spirit or gas lamp ; a gentle deflagration occurs without loss ; the ash becomes quite white, and is easily decomposed by acids ; it is treated with hydrochloric acid, evaporated to dryness, and the silica separated as usual ; correction is made for the sulphate of baryta present ; the silica is dissolved by boiling for half an hour with caustic potassa, filtered, and the loss of weight ascertained ; the silica is again separated by an acid and evaporation. vii. Estimation of the Alkalies. A certain measure Of the liquid is precipitated successively by caustic baryta, and carbo- nate of ammonia mixed with caustic ammonia ; evaporated to dryness, and the residue ignited and weighed. It is redissolved in a little water, treated with chloride of platinum, and the liquid evaporated to dryness on the water-bath ; alcohol (con- taining a little ether) is added, and the salt collected on a filter, washed with the same liquid, and dried at 212. VIII. Estimation of Perphosphate of Iron, Lime, and Mag- nesia. The remainder of the liquid is precipitated with dilute sulphuric acid ; the sulphate of baryta is collected and weighed ; it always contains some sulphate of lime, which admits of calculation from the quantity of nitrate of baryta employed. One portion of the filtered liquid is nearly neutralized by ammonia, acetate of ammonia added, warmed, and the phos- phate of iron collected ; the lime is precipitated by oxalate of ammonia, and the magnesia by phosphate of soda and am- 634 QUANTITATIVE ANALYSIS. inonia ; to the lime is added that present in the sulphate of baryta. ix. Estimation of the Phosphoric Acid. A known quantity of the solution is nearly neutralized with ammonia ; acetate of ammonia is then added, and from the filtrate now containing free acetic acid, the lime is precipitated by oxalate of ammonia ; the solution filtered from the oxalate of lime is divided into two parts ; in one, the magnesia is precipitated by ammonia, phosphate of soda being added if necessary ; in the other portion the phosphoric acid is determined by means of sulphate of mag- nesia, chloride of ammonium, and ammonia, in the usual manner. In ashes which contain but little silica or charcoal, the same method is followed, omitting the deflagration with nitrate of I aryta. With respect to Rose's method of preparing and analysing ashes, the conclusions at which Way and Ogston arrived after a laborious series of experiments may be summed up thus. It is impossible to burn a siliceous vegetable at any available temperature without a loss of carbonic acid ; and in the absence of silica, phosphates with less than three atoms of base will, to a greater or less extent, act upon the carbonates of the ash, and thus vitiate the results as far as this acid is concerned. The method proposed by Rose of treating the charred vege- table by water and estimating the alkaline carbonate in solu- tion can but partly remove this defect, for it leaves the insoluble earthy carbonates quite undetermined. In fact, we know of no method of burning plants so as to obtain estimations of the carbonic acid of their ashes which shall be at all indica- tive of the organic acids present in the vegetable substance. But a knowledge of the proportion of carbonic acid for prac- tical purposes is not required. In the ordinary method of burning, there is a liability to loss of sulphuric acid, wherever silica is present, and where silica is not present there is always a loss of sulphur not existing in the plants as sulphuric acid, which loss is very variable in different experiments, depending on a variety of circumstances over which the operator can exercise no distinct control. The presence of phosphates does not appear to have any action upon the sulphates of the ash, and therefore the sulphuric acid afforded by an ordinary burning of a non-siliceous plant must represent khefull quantity of that acid actually existing in the plant. There seems, at present, no method of distinguishing the sulphuric acid of the plant ANALYSIS OE THE ASHES OF VEGETABLES. 635 from its sulphur, but an exact estimation of the total sulphuf is obtained by digesting the plant with strong nitric acid, till the substance of it has been broken down into one uniform pulp, when the free acid is neutralized with pure carbonate of soda, and the mixture evaporated and carefully burnt. The ash is then introduced into a deep flask, and treated very gradually with strong nitric acid till it becomes acid. It is important to employ strong nitric acid, and a deep rather than an open vessel ; because without these precautions there is a very considerable liability to loss, sulphuretted hydrogen being evolved from the alkaline sulphides produced in the last stages of the burning. The sulphuric acid in the nitric solu- tion is estimated as sulphate of baryta in the usual manner. The ordinary methods of burning give perfectly accurate re- sults for phosphoric acid both in siliceous and in non-siliceous ashes, provided an intense heat be not employed, in which case it is just possible that phosphorus may be reduced and dissi- pated, though from the tenacity with which phosphoric acid retains its last equivalent of base,' such a result even then is more than questionable. Phosphates containing two atoms of a fixed base, with one of water, are not capable of decomposing chlorides at a high temperature, but phosphates with only one equivalent of a fixed base, and one or more of water, do decom- pose alkaline chlorides when heated with them, the chlorine escaping as hydrochloric acid ; but there is good reason for believing that in the burning of plants, this water is dissipated by the heat long before the mineral substances of the ash can have been brought in sufficient contact to enable them to act upon each other. Direct experiments proved that it was quite possible to obtain coincident results in regard to chlorine by careful burning, though when an unusually high temperature is employed a loss of greater or less extent occurs ; but this loss does not seem to be increased by the presence in the ash of a large proportion of silica or of phosphates. With respect to the alkalies, as they could only be lost in the form of chlo- rides, the same observations which apply to chlorine apply to those ingredients as plant ashes also. It is, after all, a matter 1 of temperature, and the authors think that the discrepancies in the results obtained by different ash-analysts may, no doubt, be referable in many cases to this cause. The burning of a plant for analysis is plainly a process requiring considerable care, a process that should, in no instance, be hurried forward ; 636 QUANTITATIVE ANALYSIS. but the authors submit that, on the whole, they have de- monstrated that (with the exception of sulphur, which is ob- tained by a separate process) their method of burning crops affords an ash really and correctly representing the mineral matter of the plants under investigation. (5.) StrecJcer's method. He first chars the substance at a very low red-heat in a muffle, until fumes are no longer given off'; it is then well mixed with a saturated solution of pure hydrate of baryta ; evaporated to dryness and ignited at a low red-heat. The baryta ash is digested with aqua-regia and eva- porated to dryness ; the sulphur is hereby converted into sul- phuric acid, and the phosphorus into phosphoric acid : it is then digested with hydrochloric acid and filtered : silica, sulphate of baryta, charcoal, and sand remain on the filter; the filtrate contains all the phosphoric acid, together with excess of baryta, lime and magnesia : it is treated with ammonia until a precipi- tate begins to fall, which is redissolved with a few drops of hy- drochloric acid. To the slightly acid solution a sufficient quan- tity of ammoniacal sulphate of magnesia and tartaric acid is added: no precipitate takes place, because the solution is acid; oxalate of ammonia, and abundance of acetate of soda are now added, by which the excess of baryta and lime are precipitated as oxalates, and are filtered off": the phosphoric acid remains in the form of ammonio-magDesian phosphate in the acetic solution, from which it is precipitated by ammonia, filtered, and estimated as usual. Arrangement of the results of analysis. Since the analysis of the ashes of a plant can convey no precise information as to the manner in which the several acids and bases found are actually combined in the living vegetable, the method recom- mended by Fresenius and Will, namely, that of enumerating the percentage weights of the acids and bases found, is pro- bably the best w r hich the present state of our knowledge ad- mits of; but, whether this or any other method be adopted, it is highly desirable that there should be uniformity on the subject, for it is only thus that the results of different analyses can be compared together, and so applied to the solution of certain interesting questions in physiology and in agriculture. The chlorine found in the analysis is always calculated as chlo- ride of sodium, or chloride of potassium should the quantity of sodium be insufficient, and the manganese as manganoso- manganic oxide (Mn 3 4 ). ANALYSIS OF THE ASHES OF BLOOD AND FLESH. 637 216. ANALYSIS OP ASHES OF BLOOD AND FLESH. i. Of Blood (Verdeil's process, Arm. Ch. Pharm., Ixix. p. 89). The blood is first charred in a porcelain crucible till no more empyreumatic vapour is evolved ; the porous coal is powdered and incinerated in a muffle, and lastly, the ash is heated with nitrate of ammonia, which is dissolved in water and added by degrees. The coal is thus entirely burnt, and the carbonates are converted into nitrates, by which means a complete sepa- ration is effected of the constituents soluble in water from those which are insoluble. The ash of blood thus prepared, when digested with water and filtered, yields a perfectly neu- tral solution, which contains alkaline phosphates, nitrates, and sulphates, as well as metallic chlorides and phosphate of mag- nesia, whilst phosphate of lime and sesquioxide of iron re- main behind. From the aqueous solution all the chlorine and phosphoric acid are precipitated by nitrate of silver. The pre- cipitate when treated with dilute nitric acid, leaves chloride of silver, dissolving the phosphate ; from the filtrate, the silver is precipitated by chloride of potassium, and the phosphoric acid is thrown down as ammonio-magnesian phosphate. From the solution which contains the alkalies, the sulphuric acid is removed by chloride of barium, the baryta and lime then pre- cipitated by carbonate of ammonia and ammonia, the filtrate evaporated, and the residue ignited. On treatment with water the magnesia is left, whilst the alkalies are dissolved, and are determined as usual. The sesquioxide of iron, the lime, and the phosphoric acid are also determined by the usual methods in the portion which is insoluble in water. For the determi- nation of the lime, the sulphuric and carbonic acids, separate quantities of ash are employed. ii. Of Flesh, and Animal Substances generally (Keller, Ann. Ghem. Pharm., Ixx. 91). The animal matter is boiled repeatedly with water, and the solution separated from the re- sidue by pressure. The liquid is evaporated, and the residue charred in a porcelain capsule ; the powdered coal is boiled with water, and the undissolved residue is completely incine- rated in the muffle. The residue of the flesh which has been boiled out and pressed is likewise dried and carbonized, and the coal, in a finely divided state, is treated for several days with strong nitric acid, after which it burns in the muffle very easily. The nitric solution is evaporated, and the ignited re- 638 QUANTITATIVE ANALYSIS. sidue repeatedly heated with nitrate of ammonia, till all the coal has disappeared. The ashes thus obtained are intimately mixed, and ignited with three or four times their volume of fused hydrate of baryta. From the aqueous solution of the ignited mass, the baryta is removed by sulphuric acid or car- bonate of ammonia, and the alkalies estimated in the filtrate. The portion insoluble in water, which contains all the phos- phoric acid, is dissolved in the least possible quantity of nitric acid : the sulphate of baryta which separates, is weighed, and the phosphate of sesquioxide of iron precipitated by ace- tate of ammonia: from the acetic solution, the total amount of phosphoric acid is separated by acetate of lead. After the de- composition of the washed phosphate of lead by sulphate of ammonium, the phosphoric acid is precipitated from the filtrate by sulphate of magnesia. In the solution filtered from the phosphate of lead, the baryta and oxide of lead are precipitated by sulphuric acid, the filtrate (from which the last traces of lead are precipitated by sulphuretted hydrogen) is evaporated, and the lime and magnesia estimated as usual. QUANTITATIVE ANALYSIS. 639 APPENDIX. TABLE I. For the Calculation of Analyses. IN the following Table the numbers have been re- calculated on the equivalent numbers assigned to the elementary substances in the Table, page 255, which is from the latest authorities. The application of this Table, and its convenience in quantitative analysis, may be illustrated by the following ex- ample : Suppose as the result of an experiment to determine the amount of sulpfair in a certain substance, 13*75 grains of sul- phate of baryta have been obtained. On turning to the Table we find under the head of " required sulphur" found, "sulphate of baryta," a series of figures arranged in nine columns, those in the first column indicating the amount of sulphur in a unit of sulphate of baryta, and those in the other columns the mul- tiples of that number by 2, 3, 4, 5, 6, 7, and 9. In the case we are considering, we require to learn the amount of sulphur in 13'75 grains of sulphate of baryta: we turn to Column 1, where we find the number 0*13734 j we put down this number, shift- ing the decimal point one figure further to the right ; we next turn to Column 3, where we find the number 0*41202 ; we place this number under the former, preserving the decimal point un- altered; the next number is 7, under which column we find the number 0*96138 ; we place this number under the last, shift- ing the decimal point one figure further to the left ; in like manner the last number being 4, we write down the number found under that column, viz. 0'6870, shifting the decimal point two figures further to the left : the whole will consequent!^ stand thus : 1-373400 0*412020 0096 138 0-006867 1-888425 13'75 grains of sulphate of baryta contain) therefore, 1'888 grams of sulphur. 640 QUANTITATIVE ANALYSIS. QUANTITATIVE ANALYSIS. 641 8 CO ~co~ 00 5 2 N CD rH \fi> CO CO 00 rH (M rH 00 1O rH 00 g 8 ! ~co to" -H. rH 6 H ~ O 1^ CO g | N (M CO ~r~s~ iO tO (M t>- rH rfl 00 l> 6 6 o_ GO O5 I 5 CO ^ 6 rH S i OC "^ ^ O O SH 3cq ^o h M |s PQ O o^o.lo^o^d ? S M U.^jfig* o^ O^^-^^-pO-riC) - 6600 642 QUANTITATIVE ANALYSIS. I 00 I QUANTITATIVE ANALYSIS. 643 rH rH $ 00 OS cq cp rH Q^ V^^ p-HJ ,^-H] ^ ^o ggte' 3 ic?id'|cf 4 |i8;t:*S IB 1*1*1** PM" C^ F o |o S- . 6> | o ij^ jp SJ g 3 "o!^ ^W SM o^i ^ rH ' ' L_J IT* L^ P4 g ^ JZJ . J2j isr, J I ^PM 2 PM PM 2PM 644 QUANTITATIVE ANALYSIS. s Ol CO ^ 6_ 9 S S I I ~* l^f 2 O" | O" -E 5 ? Q 1 f3 ^w IM ^M -SM - PM ro k5 o o P : .S ffi O S SO SO P-i QUANTITATIVE ANALYSIS. 645 o o 6 6 Chloride o NaCl 4 So"- O ^rg 66 loride JlaOi w , ^o 05 ^ (D O2 H d 1 o l * |o |o f^Jo So -go" ^o ^|^^ |fl|p|^^ alQaQHo2OpClO 3 *oO VI |o1 |i | "^ O2 QQ 02 02 * ' ^ SX 9 t -r t i 02 O I ^o 173 o 'go So d 3 o fl - o 02 O 02 _g S3 'p N QJ M O O "2~F PAET II. 646 QUANTITATIVE ANALYSIS. TABLE II. French Measures of Weight, with their English Equivalents. In English grains. In troy ounces = 480 grains. f|t * t- 3 It In cwts. = 112 Ib. = 784000 grains. Tons = 20 cwts. = 15680000 grs. M$igramme ... Centigramme . . . Decigramme ... Gramme 0-01543 0-15432 1-54323 15-43235 154-32349 1543-23488 15432-34880 154323-48800 0-000032 0-000322 0-003215 0-032151 0-321507 3-215073 32-150727 321-507267 0-0000022 0-0000220 0-0002205 0-0022046 0-0220462 0-2204621 2-2046213 22-0462126 o-ooooooo 0-0000002 0-0000020 0-0000197 0-0001968 0-0019684 0-0196841 0-1968412 o-ooooooo o-ooooooo o-ooooooi o-oooooio 0-0000098 0-0000984 0-0009842 0-0098421 Dekagramme ... Hectogramme.. [Kilogramme . . Myriogramme.. 1 grain = 0*064799 gramme ; 1 troy ounce = 31 '103496 grammes ; 1 Ib. avoir. =P 0-453595 kilogrs. ; 1 cwt. = 50'802377 kilogrs. TABLE III. French Measures of Length, with their English Equivalents. In English inches. In English feet. In English yards. In English miles. Millimetre ... Centimetre ... Decimetre . . . Metre 0-03937 0-39371 3-93708 39-37079 0-003281 0-032809 0-328090 3-280899 0-0010936 0-0109363 0'- 1093633 1-0936331 0-0000006 0-0000062 0-0000621 0-0006214 Dekametre ... Hectometre . . . Kilometre . . . Myriometre . . . 393-70790 3937-07900 39370-79000 393707-90000 32-808992 328-089920 3280-899200 32808-992000 10-9363310 109-3633100 1093-6331000 10936-3310000 0-0062138 0-0621382 0-6213824 6-2138244 1 inch = 2-539954 centimetres ; 1 yard = 0-9143835 metre. 1 foot = 3-0479449 decimetres ; 1 mile = 1-6093149 kilometre. TABLE IV. French Measures of Surface, with their English Equivalents. In English square feet. In English square yards. In English acres=43560 square feet. Centiare, or square metre .... 10-764299 1076-429934 107642-993418 1-196033 119-603326 11960-332602 0-0002471 0-0247114 2-4711431 Are, or 100 square metres Hectare, or 10,000 square metres 1 square inch = 6*4513669 square centimetres ; 1 square yard = 0-83609715 square metres. 1 square foot = 9*2899683 square decimetres ; 1 acre = 0-40467102 hectare. QUANTITATIVE ANALYSIS. 647 * I I I C^ -M C* j3 IP. I "fSi 1-1 IO d co .2 \O rH W O GO XO CO Oi .IT 1 * lO rH O! O GO "^ lO O1 !> lO rH Ol O GO "^ i> lO rH CQ O GO O} t>- 1O rH G>1 O O3 i> VO rH G<> O Ol !> lO rH OOOOOW3 COrHGOt^CO^rHTft OOOrHt^COOt^ rH X> CO O CO CO GO r-H 1> ^ CO rH CO ^O GO O CO O CO rH CO VO GO O CO XO CO rH CO K3 SCO 1O CO rH CO, O CO lO CO 6 6 6 6 6 co St^ rH 1O (M XO COlOrHiO COrHOOl>OOrH pCOrHN01l>pp OOCOrHOC^li^O CO rH O O3 t^ CO rH O ^53 CO tilfJill ! 1 1;i 1 1 % 1 S 11 1 1 I * 5 L ? t L 1 5 * 1 1 1 i 1 1 1' I 1 1 1 1 1 1 1 1 ^OPr^PWMr^ 648 QUANTITATIVE ANALYSIS. TABLE VI. For the Comparison of the Centigrade and Fahrenheit Thermometers. Cent. Fahr. Cent. Fahr. Cent. Fahr. Cent. Fahr. -SO -580 O -5 23-0 40 104-0 85 185-0 _49 -562 -4 24-8 41 105-8 86 186-8 _48 -54-4 -3 26-6 42 107-6 87 188-6 -47 -52-6 -2 28-4 43 109-4 88 190-4 -46 -50-8 -1 30-2 44 111-2 89 192-2 -45 -49-0 32-0 45 113-0 90 194-0 _44 _47'2 + 1 33-8 46 114-8 91 195-8 _43 -45-4 2 35-6 47 116-6 92 197-6 -42 -43-6 3 37-4 48 118-4 93 199-4 -41 -41-8 4 39-2 49 120-2 94 201-2 -40 -40-0 5 41-0 50 122-0 95 203-0 -39 -38-2 6 42-8 51 123-8 96 204-8 -38 -36-4 7 44-6 52 1256 97 206-6 -37 -34-6 8 46-4 53 127-4 98 208-4 -36 -32-8 9 48-2 54 129-2 99 210-2 -35 -31-0 10 50-0 55 131-0 100 212-0 -34 -29-2 11 51-8 56 132-8 101 213-8 -33 -27'4 12 53*6 57 134-6 102 215-6 -32 -25-6 13 55-4 58 136-4 103 217-4 -31 -23-8 14 57-2 59 138-2 104 219-2 -30 -22-0 15 59-0 60 140-0 105 221-0 -29 -20-2 16 60-8 61 141-8 106 222-8 -28 -18-4 17 62-6 62 143-6 107 224-6 -27 -16-6 18 64-4 63 145-4 108 226-4 -26 -14-8 19 66-2 64 147-2 109 228-2 -25 -13-0 20 68-0 65 149-0 110 230-0 -24 -11-2 21 69-8 66 150-8 111 231-8 -23 - 9-4 22 71-6 67 152-6 112 233-6 -22 - 7'6 23 73-4 68 154-4 113 235-4 -21 - 5-8 24 75-2 69 156-2 114 237-2 -20 - 4-0 25 77-0 70 158-0 115 239-0 -19 - 2-2 26 78-8 71 159-8 116 240-8 -18 - 0-4 27 80-6 72 161-6 117 242-6 -17 + 1-4 28 82-4 73 163-4 118 244-4 -16 3-2 29 84-2 74 165-2 119 246-2 -15 5-0 30 86-0 75 167-0 120 248-0 -14 6-8 31 87-8 76 168-8 121 249-8 -13 8-6 82 89-6 77 170-6 122 251-6 -12 10-4 33 91-4 78 172-4 123 253-4 -11 12-2 84 93-2 79 174-2 124 255-2 -10 14-0 35 95-0 80 176-0 125 257-0 - 9 15-8 36 95-8 81 177-8 126 258-8 - 8 17-6 37 98-6 82 179-6 127 260-6 - 7 19-4 38 100-4 83 181-4 128 262-4 - 6 21-2 39 102-2 84 183-2 129 264-2 QUANTITATIVE ANALYSIS. TABLE VI. (continued.) 649 Cent. Fahr. Cent. Fahr. Cent. Fahr. Cent. Fahr. 130 266-0 179 354-2 228 442-4 277 530-6 131 267-8 180 356-0 229 444-2 278 532-4 132 269-6 181 357-8 230 446-0 279 534-2 133 271-4 182 359-6 231 447-8 280 536-0 134 273-2 183 361-4 232 449-6 281 537-8 135 275-0 184 363-2 233 451-4 282 539-6 136 276-8 185 365-0 234 453-2 . 283 541-4 137 278-6 186 366-8 235 455-0 284 543-2 138 280-4 187 368-6 236 456-8 285 545-0 139 282-2 188 370-4 237 458-6 286 546-8 140 284-0 189 372-2 238 460-4 287 548-6 141 285-8 190 374-0 239 462-2 288 550-4 142 287-6 191 375-8 240 464-0 289 552-2 143 289-4 192 377-6 241 465-8 290 554-0 144 291-2 193 379-4 242 467-6 291 555-8 145 293-0 194 381-2 243 469-4 292 557-6 146 294-8 195 383-0 244 471-2 293 559-4 147 296-6 196 384-8 245 473-0 294 561-2 148 298-4 197 386-6 246 474-8 295 563-0 149 300-2 198 388-4 247 476-6 296 564-8 150 302-0 199 390-2 248 478-4 297 566-6 151 303-8 200 392-0 249 480-2 298 568-4 152 305-6 201 393-8 250 482-0 299 570-2 153 307-4 202 395-6 251 483-8 300 572-0 154 309-2 203 397-4 252 4856 301 573-8 155 311-0 204 399-2 253 487-4 302 575-6 156 312-8 205 401-0 254 489-2 303 577-4 157 314-6 206 402-8 255 491-0 304 579-2 158 316-4 207 404-6 256 492-8 305 581-0 159 318-2 208 406-4 257 494-6 306 582-8 160 320-0 209 408-2 258 496-4 307 584-6 161 321-8 210 410-0 259 498-2 308 586-4 162 323-6 211 411-8 260 500-0 309 588-2 163 325-4 212 413-6 261 501-8 310 590-0 164 327-2 213 415-4 262 503-6 311 591-8 165 329-0 214 417-2 263 505-4 312 593-6 166 330-8 215 419-0 264 507-2 313 595-4 167 332-6 216 420-8 265 509-0 314 597-2 168 334-4 217 422-6 266 510-8 315 599-0 169 336-2 218 424-4 267 512-6 316 6008 170 338-0 219 426-2 268 514-4 317 602-6 171 339-8 220 428-0 269 516-2 318 604-4 172 341-6 221 429-8 270 518-0 319 606-2 173 343-4 222 431-6 271 519-8 320 608-0 174 345-2 223 433-4 272 521-6 175 347-0 224 435-2 273 523-4 176 348-8 . 225 437-0 274 525-2 177 350-6 226 438-8 275 527-0 178 352-4 227 440-6 276 528-8 650 QUANTITATIVE ANALYSIS. TABLE VIL Showing the Quantity of Oil of Vitriol of Specific Gravity 1'8485, and of Anhydrous Acid, in 100 Parts of dilute Sulphuric Acid, of different Densities, according to Dr. Ure. Liquid. Specific Gravity. Dry. Liquid. Specific Gravity. Dry. 100 1-8485 81-54 50 1-3884 40-77 99 1-8475 80-72 49 1-3788 39-95 98 1-8460 79'90 48 1-3697 39-14 97 1-8439 79-09 47 1-3612 38-32 96 1-8410 78-28 46 1-3530 37-51 95 1-8376 77-46 45 3440 36-69 94 1-8336 76-65 44 3345 35-88 93 1-8290 75-83 43 3255 35-06 92 1-8233 75-02 42 3165 34-25 91 1-8179 74-20 41 3080 33-43 90 1-8115 73-39 40 2999 32-61 89 1-8043 72-57 39 1-2913 31-80 88 1-7962 71-75 38 1-2826 30-98 87 1-7870 70-94 37 1-2740 30-17 86 1-7774 70-12 36 1-2654 29-35 85 1-7673 69-31 35 1-2572 28-54 84 1-7570 68-49 34 1-2490 27-72 83 1-7465 67-68 33 1-2409 26-91 82 1-7360 66-86 32 1-2334 26-09 81 1-7245 66-05 31 1-2260 25-28 80 1-7100 6523 30 1-2184 24-46 79 1-6993 64-42 29 1-2108 23-65 78 1-6870 63-60 28 1-2032 22-83 77 1-6750 62-78 27 1-1956 22-01 76 1-6630 61-97 26 1-1876 21-20 75 1-6520 61-15 25 1-1792 20-38 74 1-6415 60-34 24 1-1706 19-57 73 1-6321 59-52 23 1-1626 18-75 72 1-6204 58-71 22 1-1549 17 ; 94 71 1-6090 57-89 21 1-1480 17-12 70 1-5975 57*08 20 11410 16-31 69 1-5868 56-26 19 1-1330 15-49 68 1-5760 55-45 18 1-1246 14-68 67 1-5648 54-63 17 1-1165 13-86 66 1-5503 53-82 16 1-1090 13-05 65 1-5390 53-00 15 1-1019 12-23 64 5280 52-18 14 1-0953 11-41 63 5170 51-37 13 1-0887 10-60 62 5066 50-55 12 1-0809 9-78 61 1-4960 49-74 11 1-0743 8-97 60 4860 48-92 10 1-0682 815 59 4760 48'H 9 1-0614 7-34 58 4660 47-29 8 1-0544 6-52 57 1-4560 46-48 7 1-0477 571 56 1-4460 45'66 6 1-0405 4-89 55 1-4360 44-85 5 1-0336 4'08 54 1-4265 44-03 4 1-0268 3-26 53 1-4170 43-22 3 1-0206 2-446 52 1-4073 42-40 2 1-0140 1-63 51 1-3977 41-58 1 1-0974 0-8154 QUANTITATIVE ANALYSIS. 651 TABLE VIII. Showing the Quantity of Seal or Anhydrous Nitric Acid in 100 Parts of Liquid Acid, at different Densities, according to Dr. Ure. Specific Gravity. Eeal Acid in 100 Parts of the Liquid. Specific Gravity. Eeal Acid in 100 Parts of the Liquid. 1-5000 79-700 1-2947 39-850 1-4980 78-903 1-2887 39-053 1-4960 78-106 1-2826 38-256 1-4940 77-309 2765 37^459 1-4910 76-512 2705 36-662 1-4880 75-715 2644 35-865 1-4850 74-918 2583 35-068 1-4820 74-121 2523 34-271 1-4790 73:324 1-2462 33-474 1-4760 72-527 1-2402 32-677 1-4730 71-730 12341 31-880 1-4700 70-933 1-2277 31-083 1-4670 70-136 1-2212 30-286 1-4640 69-339 1-2148 29-489 1-4600 68-542 1-2084 28-692 1-4570 67-745 1-2019 27-895 1-4530 66-948 1-1958 27-098 1-4500 66-155 1-1895 26-301 1-4460 65-354 1-1833 25-504 1-4424 64-557 1-1770 24-707 1-4385. 63:760 1-1709 23-900 1-4346 62-963 1648 23-113 1-4306 62-166 1587 22-316 1-4269 61-369 1526 21-519 1-4228, 60-572 1465 20-722 1-4189 59-775 1403 19-925 1-4147 58-978. 1345 19-128, 1-4107 58-181 1286 18-331 1-4065. 57-384 1227 17-534 1-4023 56-587 1-1168 16-737 1-3978 55-790 1109 15-940 1-3945 54-993 1051 15-143 1-3882 54:196 1-0993 14-346 1-3833 53-399 1-0935 13-549 1-3783 52-602 1-0878 12-752 1-3732 51-805 1-0821 11-955 1-3681 51-068 1-0764 11-158 1-3630 50-211 1-0708 10-361 1-3579 49-414 1-0651 9-564 1-3529 48-617 1-0595 8-767 1-3477 47-820 1-05^0 7-970 1-3427 47-023 1-0485 7-173 1-3376 46-226 1-0430 6-376 1-3323 45-429 1-0375 5-579 1-3270 44-632 1-0320 4-782 1-3216 43-835 1-0267 3-985 1-3113 43-038 1-0212 3-188 1-3110 42-241 1-0159 2-391 1-3056 41-444 1-0106 1-594 1-3001 40-647 1-0053 0-797 652 QUANTITATIVE ANALYSIS. TABLE IX. Shelving the Quantity of A bsolute Alcohol in Spirits of different Specific Gravities, according to Lowitz. 100 Parts. Specific Gravity. 100 Parts. Specific Gravity. Alcohol. Water. At 68. At 60. Alcohol. Water. At 68. At 60. 100 0-791 0-796 49 51 0-917 0-920 99 1 0794 0-798 48 52 0-919 0-922 98 2 0797 0-801 47 53 0-921 0-924 97 3 0-800 0'804 46 54 0-923 0-926 96 4 0-803 0-807 45 55 0-925 0-928 95 5 0-805 0-809 44 56 0-927 0'930 94 6 0-808 0-812 43 57 0-930 0-933 93 7 0-811 0-815 42 58 0-932 0-935 92 8 0-813 0.817 41 59 0-934 0-937 91 9 0-816 0-820 40 60 0-936 0-939 90 10 0-818 0-822 39 61 0-938 0-941 89 11 0-821 0-825 38 62 0-940 0-943 88 12 0-823 0-827 37 63 0-942 0-945 87 13 0-826 0-830 36 64 0-944 0-947 86 14 0-828 0-832 35 65 0-946 0-949 85 15 0-831 0-835 34 66 0-948 0-951 84 16 0-834 0-838 33 67 0-950 0-953 83 17 0-836 0-840 32 68 0-952 0-955 82 18 0-839 0-843 31 69 0-954 0-957 81 19 0-842 0-846 30 70 0-956 0-958 80 20 0-844 0-848 29 71 0-957 0-960 79 21 0-847 0-851 28 72 0-959 0-962 78 22 0-849 0-853 27 73 0-961 0-963 77 23 0-851 0-855 26 74 0-963 0-965 76 24 0-853 0-857 25 75 0-965 0-967 75 25 0-856 0-860 24 76 0-966 0-968 74 26 0-859 0-863 23 77 0968 0-970 73 27 0-861 0-865 22 78 0-970 0-972 72 28 0-863 0-867 21 79 0-971 0-973 71 29 0-866 0-870 20 80 0-973 0-974 70 30 0868 0-872 19 81 0-974 0-975 69 31 0-870 0-874 18 82 0-976 0-977 68 32 0-872 0-875 17 83 0-977 0-978 67 33 0-875 0-879 16 84 0-978 0-979 66 34 0-877 0-881 15 85 0-980 0-981 65 35 0-880 0-883 14 86 0-981 0-982 64 36 0-882 0-886 13 87 0-983 0-984 63 37 0-885 0889 12 88 0-985 0-986 62 38 0-887 0-891 11 89 0-986 0-987 61 39 0-889 0-893 10 90 0-987 0-988 60 40 0-892 0-896 9 91 0-988 0-989 59 41 0-894 0-898 8 92 0-989 0'990 58 42 0-896 0-900 7 93 0-991 0-991 57 43 0-899 0-902 6 94 0-992 0-992 56 44 0-901 0-904 5 95 0-994 55 45 0-903 0-906 4 96 0-995 54 46 0-905 0-908 3 97 0-997 53 47 0-907 0-910 2 98 0-998 52 48 0-909 0-912 1 99 0-999 51 49 0-912 0-915 100 1-000 50 50 0-914 0-917 QUANTITATIVE ANALYSIS. 053 TABLE X. Showing the Quantity of Absolute Alcohol by Weight in Mix- tures of Alcohol and Water of different Specific Gravities, according to Mr. Drink water.* Sp. Gr. Alcohol Sp. Gr, Alcohol Sp. Gr. Alcohol Sp. Gr. Alcohol at 60 P. per cent, by weight. at 60 F. per cent, by weight. at 60 F. per cent, by weight. at 60 F. per cent, by weight. 1-0000 o-oo 9959 2-22 9918 4-64 9877 7-30 9999 0-05 9958 2-28 9917 470 9876 7-37 9998 O'll 9957 2-34 9916 4-76 9875 7-43 9997 0-16 9956 2-37 9915 4-82 9874 7-50 9996 0-21 9955 2-45 9914 4-88 9873 7-57 9995 0-26 9954 2-51 9913 4-94 9872 7-64 9994 0-32 9953 2-57 9912 5-01 9871 7-71 9993 0-37 9952 2-62 9911 5-07 9870 7-68 9992 0-42 9951 2-68 9910 5'13 9869 7-85 9991 0-47 9950 2-74 9909 5-20 9868 7-92 9990 0-53 9949 2-79 9908 5-26 9867 7-99 .9989 0-58 9948 2-85 9907 5'32 9866 8-06 9988 0-64 9947 2-91 9906 5'39 9865 8-13 9987 0-69 9946 2-97 9905 5-45 9864 8-20 9986 074 9945 302 9904 5-51 9863 8-27 9985 0-80 9944 3-08 9903 5-58 9862 8-34 9984 0-85 9943 3-14 9902 5-64 9861 8-41 9983 0-91 9942 3-20 9901 5-70 9860 8-48 9982 0-96 9941 3-26 9900 5-77 9859 8-55 9981 1-02 9940 3-32 9899 5-83 9858 8-62 9980 1-07 9939 3-37 9898 5'89 9857 8-70 9979 1-12 9938 3-43 9897 5-96 9856 8-77 9978 118 9937 3-49 9896 6-02 9855 8-84 9977 1-23 9936 3-55 9895 6-09 9854 8-91 9976 1-29 9935 3-61 9894 6-15 9853 8-98 9975 1-34 9934 3-67 9893 6-22 9852 9-05 9974 1-40 9933 3-73 9892 6'29 9851 9-12 9973 1-45 9932 3-78 9891 6-35 9850 9-20 9972 1-51 9931 3-84 9890 6-42 9849 9-27 9971 1-56 9930 3-90 9889 6-49 9848 9-34 9970 1-61 9929 3-96 9888 6'55 9847 9-41 9969 1-67 9928 4-02 9887 6'62 9846 9-49 9968 1-73 9927 4-08 9886 6-69 9845 9-56 9967 1-78 9926 4-14 9885 6-75 9844 9-63 9966 1-83 9925 4-20 9884 6'82 9843 970 9965 1-89 9924 4-27 9883 6-89 9842 9-78 9964 1-94 9923 4-33 9882 6-95 9841 9-85 9963 1-99 9922 4-39 9881 7-02 9840 9-92 9962 2-05 9921 4-45 9880 7-09 9839 9'99 9961 2-11 9920 4-51 9879 7-16 9838 10-07 9960 2-17 9919 4-57 9878 7-23 The absolute alcohol, with which the experiments which furnished the data for the above Table were made, was prepared as follows : Carbonate of potassa was exposed to ' Memoirs of the Chemical Society,' vol. iii. p. 454. 2 x 654 QUANTITATIVE ANALYSIS. a red-heat to deprive it of water ; and, when sufficiently cool, was pulverized and added to ordinary alcohol of specific gravity '850 at 60 F. till it ceased to dissolve any more : the whole was then allowed to digest 24 hours, being frequently agitated, when the alcohol was carefully poured off. As much fresh-burned quicklime as was considered sufficient, when powdered, to absorb the whole of the alcohol, was introduced into a re- tort, and the alcohol added to it ; after digesting 48 hours, it was slowly distilled in a water-bath at a temperature of about 180 F. The alcohol thus obtained was care- fully redistilled, and its specific gravity at 60 F. was found in two experiments to be '7946 and '7947. It was subsequently digested a second and a third time for several days with recently ignited quicklime, and redistilled at 172 F. The mean of several de- terminations gave the number '79381, which the author thinks may be regarded as expressing the true specific gravity of absolute alcohol at 60 F. TABLE XI. Tension of Aqueous Vapour at different Temperatures. 11 !- a a Tension of the aqueous vapour expressed in millimetres. Temperature in degrees (Centigrade). Tension of the aqueous vapour expressed in millimetres. 1st HI 14 Tension of the aqueous vapour expressed in millimetres. 4-525 14 11-882 28 28-148 1 4-867 15 12-677 29 29-832 2 5-231 16 13-519 30 31-602 3 5-619 17 14-409 31 33-464 4 6-032 18 15-351 32 35-419 5 6-471 19 16-345 33 37-473 6 6-939 20 17-396 34 39-630 7 7-436 21 18-505 35 41-893 8 7-964 22 19-675 36 44-268 9 8-525 23 20-909 37 46-758 10 9-126 24 22-211 38 49-368 11 9-751 25 23-582 39 52-102 12 10-421 26 25-026 40 54-969 13 11-130 27 26-547 INDEX. Page Acetates, comportment of, with re- agents 177 Acetic Acid as a reagent ..." . 44 Acid, Acetic, general characters of . 177 Benzoic, general characters of . . 175 Boracic, general characters of . . 148 Salts of, comportment of with reagents 148 quantitative determination of . 463 separation of, from bases . . . 463 determination of in Silicates . 464 separation of, from Phosphoric Acid 465 Bromic, general characters of . .156 Carbonic, general characters of . 139 Salts of, comportment with re- agents 139 quantitative determination of in aqueous solution 474 in the gaseous state 475 in aerated waters 475 separation of, from Sulphurous Acid 480 from all other acids 480 Chloric, general characters of . . 170 Chromic, general characters of . . 152 Salts of, comportment of, with reagents 153 Citric, general characters of ... 173 Formic, general characters of, and comportment with reagents . . 177 Gallic, general characters, of, and comportment with reagents . . 176 Hydriodic, general characters of . 157 Hydrobromic, general characters of 155 Hydrochloric, general characters of 154 Hydrocyanic, general characters of 159 detection of in organic mixtures 161 Hydro ferrocyanic . . . . .164 Hydroferridcyanic 165 Hydrofluoric, general characters of 151 Salts of, comportment of with re- agents 151 Hydrosulphuric 166 Hydrosulphocyanic 164 Hypochloric, characters of . . .170 Hyposulphurous Salts of, comport- ment with reagents 141 "luric general characters . 143 Page Acid, Ipdic, general characters of . . 158 Malic, general characters of ... 174 Meconic, general characters and recognition of 178 Mtric, general characters of. . . 167 estimation of free, by Hydrate of Baryta 510 by acidiinetry 510 determination of in combination with bases 511 indirect method of estimating . 513 estimation of, by conversion into Ammonia 520 Pugh's method 521 Nitrous, general characters of . . 169 Oxalic, general characters of . . 171 Paratartaric 173 Perchloric, general characters of . 144 Phosphoric, general characters of . 144 tribasic, comportment of Salts of, with reagents 145 Phosphoric estimation of, as Pyro- phosphate of Magnesia . . . 452 by Molybdate of Ammonia . . 460 by Nitrate of Silver 453 by Nitrate of Uranium .... 454 by Nitrate of Bismuth . . . .457 by metallic Tin 457 volumetric estimation of, by Ace- tate of Uranium 455 by Perchloride of Iron .... 458 in soils, estimation of .... 459 by Nitric Acid and metallic Mer- cury 460 separation of from Phosphorous Acid 461 Selenic, comportment of, with re- agents 144 Phosphorous, general characters of 147 Silicic, general characters of . . 149 quantitative estimation of . . 465 separation of, from bases . . . 465 Succmic, general characters of . . 175 Sulphuric, general characters of . 142 Salts of, comportment with re- agents 143 estimation of as Sulphate of Baryta 438 volumetric estimation of ... 439 Sulphurous, estimation of . . . .439 656 INDEX. Page Acid, Sulphur, analysis of a mixture of compounds of, with Oxygen . . 440 Sulphurous, general characters of 140 Salts of, comportment of with reagents 140 Tannic, general characters of, and comportment with reagents . . 176 Tartaric, general characters of . . 172 Uric, general characters and recog- nition of 178 Acids, examination for 205 inorganic, reagents for grouping . 138 organic, reagents for grouping . .138 Aconitina, general characters of . . 184 Air-bath, Karnmelsberg'a .... 250 Taylor's 251 Alcohol, as a reagent 58 Alkalimetry, Gay-Lussac's method of 263 Mohr's method of 264 Fresenius and Will's method of . 268 Alkaloids, poisonous, detection of in organic mixtures 185 Aluminum of Sesquioxide, general cha- racters of 69 Salts of, comportment of with re- agents 69 Sesquioxide, quantitative estima- tion of 285 separation of, from Lime .... 286 from Magnesia 286 from Lime and Magnesia . . . 287 Amalgams, analysis of ...... 374 Ammonia, as a reagent 46 general characters of . ... 63 Sesquicarbonate of, as a reagent Oxalate of, as a reagent . . . and Soda, Phosphate of, as a re agent ......... Molybdate of, as a reagent . . volumetric estimation of . . . Schlaesing's method of determin ing . . alker's Walker's method of determining determination of in rain water Boussingault's method . . Lawes and Gilbert's method separation of from Potassa and Soda 275 Ammoniacal Salts, analysis of ... 272 Ammonium, Chloride of, as a reagent 46 Sulphide of, as a reagent .... 46 comportment of Salts of with re- agents 64 estimation of, as Chloride . . . 271 as double Chloride of Platinum and Ammonium 271 Analysis qualitative, operations to be performed, and apparatus re- quired qualitative, systematic 181 quantitative, operations in ... 215 Antimony, Teroxide of, general cha- racters of 11' comportment of Salts of with re- agents .... 11 Page detection of, in organic mixtures . 112 Antimony, quantitative estimation of, as Sulphide 404 as Antimonious Acid .... 406 ores, assay of 406 Arsenic, detection of, in paper hang- ings 122 Acid, general characters of ... 124 comportment of with reagents . . 124 estimation of by Protoxide of Lead 407 as Sulphide 407 as Arsenate of Magnesia and Am- monia 408 separation of, i'rom Antimony . . 409 Behren's method 410 Fresenius' method 411 Hofmann's method 411 Mayer's method 412 Arsenious Acid, general characters of 113 comportment of with reagents . 114 detection of in organic com- pounds 123 separation of, from Arsenic Acid . 409 Atropia, general characters of . . .184 Balance philosophical, description of . 213 requisites for a good 214 sensibility and stability of ... 215 Captain Eater's 218 Kobinson's 216 Ladd and Oertling's 218 Barium Chloride of, as a reagent . . 53 comportment of Salts of with re- agents 64 Oxide of, quantitative determina- tion of 277 as Sulphate 277 as Carbonate 277 separation of, from the Alkalies . 278 Baryta, Hydrate of, as a reagent . . 53 Nitrate of, as a reagent .... 53 general characters of 64 Bases, examination for 196 Benzoates, comportment of solutions of with reagents 175 Benzoic Acid. See Acid, Benzoic. Bismuth, Oxide of, general characters of 100 quantitative determination of 374 comportment of Salts of with re agents 100 separation of from Oxides of Lead 375 from Oxide of Silver . . , 376 Blood, analysis of ashes of ... 637 Blowpipes, Black's 26 Planner's 27 Plattner's lamp 27 Blowpipe supports 30 Griffin's, for reductions and fu- sions 31 charcoal 32 spoon and forceps 32 Bone earth, analysis of 462 Bones, analysis of 614 Boracic Acid. See Acid, Boracic. INDEX. 657 Page Brass, analysis of 392 Bromates, comportment of solutions of, with reagents 157 Bromic Acid. See Acid, Bromic. Bromides, comportment of solutions of, with, reagents 155 Bromine, estimation of, as Bromide of Silver 500 direct estimation of 501 indirect estimation of 501 separation of, from Iodine and Chlorine 502 Pisani's method 502 Lyte and Field's method . . .502 Brucia, general characters of, and re- actions with reagetots- .... 184 Burette, tray-Lussac's 242 Binks's 243 Mohr's 243 graduation of 244 Cadmium, Oxide of, general charac- ters of 101 comportment of solutions of Salts of, with reagents 101 quantitative determination of, as Oxide 377 as Sulphide 377 separation of, from Oxide of Lead 378 from Oxide of Silver ... 378 from Oxide of Mercury . . 379 from Oxide of Bismuth . . 379 from Oxide of Zinc .... 379 Calcium, Chloride of, as a reagent . 55 comportment of Salts of, with re agents 66 Oxide of, quantitative determina- tion of 279 as Sulphate 279 as Carbonate 279 separation of, from Baryta . . . 280 from Strontia 280 from Baryta and Strontia . . 280 from the Alkalies 280 Camphor vapour, determination of specific gravity of 238 Carbon, separation of, from Nitre and Sulphur 481 calculation of, in organic com- pounds 542 determination of total amount of, in soil 601 Carbonates, analysis of 476 Cellulose, its conversion into Humin, U Imin, and organic acids in the soil 599 Chemical analysis, general remarks on 1 Chlorates, general characters of . . 170 analysis of 522 Chloric Acid, quantitative estimation of 522 Chloride of Lime, determination of the value of 491 Gay-Lussac's method .... 491 by Arsenious Acid 493 by Protosulphate of Iron . . . 494 by Yellow Prussiate of Potash . 495 Page Chlorides, solutions of, comportment of, with reagents 154 soluble, analysis of 485 insoluble, analysis of 486 Chlorimetry 491 Chlorine volumetric determination of 488 quantitative estimation of free . . 489 Chlorine water as a reagent .... 57 Chrome ores, analysis of . . . . 292 Chromium Sesquioxide of, general characters of 73 comportment of Salts of with re- agents 73 Sesquioxide of, estimation of . . 290 estimation of, in its compounds . 290 Sesquioxide of, separation of from Alumina 293 from the alkaline earths . . . 294 from the Alkalies 294 Citrates, comportment of with re- agents 174 Clays, analysis of , . 604 Cobalt, Oxide of, general characters of 77 comportment of Salts of with re- agents 77 Oxide of, estimation of .... 302 reduction of to the metallic state . 302 Protonitrate of, as a reagent . . 57 Oxide of, separation of from Oxide of Nickel. . / 304 from Oxide of Zinc 307 from Oxide of Chromium ... 307 from Alumina 307 from the alkaline earths . . . 308 ores, analysis of 308 and B iekel, extraction of from the ore 308 Condensation of products of distilla- tion, Woulfe's apparatus for . . 39 small apparatus for 40 Condenser, Liebig's . 87 ordinary 37 Conia, general characters of, and reac- tions with reagents 181 Copper, Oxide of, general characters of 102 comportment of Salts of with re- agents 102 poisoning by 103 detection of, in organic mixtures . 103 Sulphate of, as a reagent . . c . 54 determination of, as Oxide . . . 380 as Subiodide 381 estimation of, in ores, Mohr's me- thod 381 Levol's method 382 Cassaseca's method ..... 383 volumetric method of estimating, Pelouze's method 383 C. Mohr's method 385 Schwartz's method 385 Fleitmann's method .... 386 Terreil's method 386 Parker and Mohr's method . . 386 Streng's method 387 Plessy and Moreau's method . 388 658 INDEX. Page Copper, volumetric method of estima- ting, E. O. Brown's method . . .388 Kunsel's method 388 Oxide of, separation of from Oxide of Lead 389 from Oxide of Silver . . . 389 from Oxide of Mercury . 390 from Oxide of Bismuth . 390 from Oxide of Cadmium . 391 from Oxide of Zinc . . .392 from Oxide of Nickel . .393 commercial analysis of . . . 393 Oxide of, separation of from Oxides of Lead, Bismuth, Silver, Mer- cury, and Cadmium 395 Crenic Acid, its formation in the soil . 599 Crucibles, Cornish 24 Hessian 24 Black Lead 24 Berlin ware 25 Platinum 25 Crystals efflorescent, preparation of, for analysis 247 deliquescent, preparation of, for analysis 247 Cupellation 360 Cyanides, comportment of solutions of, with reagents 160 analysis of 504 Cyanogen, estimation of as Cyanide of Silver 503 by the action of Iodine . . . 505 separation of from Chlorine, Bro- mine, and Iodine 506 Desiccation 246 in vacuo 252 Distillation 35 general process of 36 precautions to be observed during 37 Equivalents, chemical table of ... 255 Ether, as a reagent 58 Evaporation 34 precautions to be observed during 34 Ferridcyanides, comportment of with reagents 165 Ferrocyanides, comportment of with reagents 164 Filtering, operation of 9 Filtering paper 7 purification of 8 Filters, method of making .... 8 Filtration 7 Flame, structure of 27 oxidizing 29 reducing 29 Flesh, analysis of ashes of , ... 638 Fluorides, comportment of solutions of with reagents 151 analysis of by Sulphuric Acid . . 507 Wohler's process for .... 608 separation of from Phosphates . 509 Fluorine, estimation of as Fluoride of Calcium . ...... 507 Page Fluorine, conversion of into Fluoride of Silicon 508 Furnaces, Faraday's Black Lead . . 23 Aikin's 23 Black's . 24 Gas burners, gauze 1 Bunsen's 13 Griffin's 1 Normandy's 19 Gases, measurement of 240 Georgina paper, as a reagent ... 42 Glucinum. Oxide of, general charac- ters of . , 70 comportment of Salts of with re- Oxide of, quantitative estimation of 287 separation of from Alumina . . 287 from Lime, Magnesia, and the Alkalies 288 Gold, Teroxide of, general characters of . . . 128 behaviour of solutions of Salts of, with reagents 129 precipitation of by Protosulphate offron ../..... .424 by Oxalic Acid 424 estimation of by standard solution, Henry's process 425 analysis of alloys of 427 purification of by cementation . . 429 by fusion with Sulphide of Anti- mony 429 separation of from Platinum . . 427 from Antimony, Tin, and Arsenic 427 Guano, analysis of 606 Gunpowder, analysis of 481 Heat, sources of 12 Hydrochloric Acid as a reagent . . 506 quantitative determination of . . 485 Hydrocyanic Acid, determination of the amount of in medicinal Hy- drocyanic Acid 506 in bitter-almond water . . . 506 in cherry-laurel water .... 506 Hydrofluosilicic Acid as a reagent . . 56 Hydrogen, calculation of inorganic compounds 542 Hydrometers 230 Beaume's 231 Neubauer's 231 Hydrosulphuric Acid as a reagent . 47 quantitative estimation of ... 442 Humic group of acids, determination of in soH 601 Humus Coal, determination of in soil 601 Hypochlorites, general characters of . 170 Imperial pint, its contents .... 240 Indigo, solution of as a reagent . . 42 lodates, comportment of solutions of with reagents 159 Iodides, comportment of solutions of with reagents 157 INDEX. 659 Page Iodine, estimation of, as Iodide of Silver .......... 496 as Protiodide of Palladium . . 496 as Protiodide of Copper . . . 496 Pisani's method of estimating . .497 indirect method of estimating . . 497 estimation of free ...... 497 estimation of, in Kelp ..... 498 in water ......... 498 separation of, from Chlorine . . 497 detection of, in the presence of Ni- tric Acid ......... 499 volumetric estimation of .... 499 Iridium, Binoxide of, general charac- ters of ......... 128 of Salts of, with reagents . . .128 extraction of, from Platinum ores 402 Iron, Protosulphate of, as a reagent . 55 Sesquichloride of, as a reagent . 56 Protoxide of, general characters of 80 comportment of Salts of, with re- agents ......... 81 Sesquioxide of, general charac- ters of ......... 81 comportment of Salts of, with re- agents ......... 82 Oxides, estimation of 309 Protoxide of, separation of, from Peroxide ........ 310 separation of,from Oxides of Nickel and Cobalt ........ 313 from Oxide of Manganese . . 313 from Oxide of Zinc ..... 314 from Oxide of Chromium . . 314 from Yttria ....... 315 from Alumina ...... 315 from Lime and Magnesia . . . 316 ores, dry assay of ...... 316 humid assay of ...... 318 Marguerite's method of determin- ing ........... 321 Penny's method of determining .323 pig, analysis of ....... 325 cast, determination of carbon in . 328 Kupfernickel, analysis of ..... 301 Lead paper as a reagent ..... 43 neutral Acetate of, as a reagent . 55 basic Acetate of, as a reagent . . 55 Protoxide of, general characters of 84 comportment of Salts of, with re- agents .......... 85 poisoning by, detection of, in or- ganic mixtures ...... 86 action of water on ...... 88 recognition and estimation of, in water .......... 91 estimation of, as Sulphate . . .343 volumetric estimation of, by Sul- phide of Sodium ...... 343 by Bichromate of Potassa . .344 quantitative estimation of, as Oxide 453 precipitation of, as Oxalate . . . 346 as Chloride ....... 346 Sulphate of, separation of from Sulphate of Baryta 347 Lime water as a reagent 55 Sulphate of, as a reagent .... 55 Superphosphate of, analysis of . . 611 Limestones, analysis of 603 Lithia, general characters of, and tests for 63 Litmus paper, blue, as a reagent . . 42 red, as a reagent 42 Magnesium, comportment of Salts of with reagents 67 Oxide of, quantitative estimation of 281 as Sulphate 281 as Pyrophosphate 281 separation of from Baryta and Strontia 281 from Lime 282 from the Alkalies 282 Malates, comportment of solutions of with reagents 174 Manganese, Oxide of, general charac- ters of 78 comportment of Salts of with re- agents .79 Protoxide of, estimation of . 333 Eed Oxide, estimation of . . 333 Sulphate, estimation of . . 333 Sulphide, estimation of . . 334 separation of from Oxides of Co- balt and Nickel ...... 334 from Cobalt 335 from Iron and Nickel . . . .336 from Oxide of Zinc 336 from Oxide of Chromium . . .336 from Alumina 336 Magnesia and Lime 336 Binoxide of, determination of com- mercial value of 336 Mannite, analysis of 541 calculation of the formula of . . 542 Manures, analysis of 606 Measuring 240 Mercury Chloride of, as a reagent . . 57 Subnitrate of, as a reagent ... 57 Suboxide of, general characters of 94 comportment of Salts of with re- agents 94 Oxide of, general characters of . . 95 comportment of Salts of with re- agents 96 poisoning by 97 detection ot in organic mixtures . 97 quantitative determination of as metallic Mercury in the dry way 366 in the wet way 369 Oxides of, separation of from Oxide of Lead . 372 Saboxide of, separation of from Oxide of Lead 372 from Oxide of Silver .... 373 660 INDEX. Page Mercury, Suboxide of, separation of, from Oxide of Silver 373 Suboxide of, separation of from Oxide of 374 Metaphosphates, comportment of so- lutions of with reagents .... 144 Molybdate of Lead, analysis of ... 435 Molybdenum, Oxides of, general cha- racters of 135 Binoxide of, comportment of with reagents 136 quantitative estimation of ... 434 Molybdic Acid, general characters of 136 behaviour with reagents . . . 137 Morphia, general characters of, and reactions with reagents . . . 181 Mudeseous Acid, its formation from vegetable matters in the soil . . 600 Narcotina, general characters and re- cognition of 182 Nickel Oxide of, general characters of 76 comportment of Salts of, with re- agents 76 quantitative estimation of . . 298 separation of from Oxide of Zinc 299 from Oxide of Chromium . . 300 from Alumina 303 from Magnesia 301 from the alkaline earths . . 301 Nicotina general characters of, and reaction with reagents .... 180 Nitrates, detection of Nitric Acid in solutions of 168 Nitre valuation of, Pelouze's method 513 -remarks of Abel and Bloxam on 516 Fresenius's modification of . . 517 Gay-Lussac's method .... 518 Abel and Bloxam' s method . .518 Nitric Acid as a reagent 43 Nitrogen, volumetric determination of 274 determination of, in organic com- pounds 547 determination of, in a soil . . . 601 Nitromuriatic Acid as a reagent . . 44 Nitrotoluide, analysis of 561 Organic analysis, gas as a fuel for . . 561 combustion-furnace for .... 536 gas-furnace for 561 calculation of the results of . . .541 Organic bodies, elementary analysis of 524 Front's method 525 Gay-Lussac and Thenard's me- thod 525 Liebig's method 527 preliminary testing of 527 testing of for Nitrogen .... 527 testing for and estimation of Sul- phur Ruling's process ....... 528 Kemp's process ...... 529 Weidenbusch's process . . .529 Heintz's process 530 by Acetate of Lead 530 Page Organic bodies, testing for, and esti- mation of Phosphorus in ... 531 Organic compounds containing Chlo- rine, analysis of 558 containing Sulphur, analysis of . 559 determination of the atomic weight of 559 Organic non-volatile solids not contain- ing Nitrogen, analysis of . . . 531 Organic volatile compounds not con- taining Nitrogen, analysis of . . 545 Organic substances containing Nitro- gen, analysis of 547 estimation of Nitrogen in a gaseous state, Dumas' method . 548 Liebig's method 552 Bunsen's method 553 Simpson's method 651 Will and Varrentrap's method . 555 Peligot's method 555 Organic substances in the soil, their conversion into organic acids . 598 Osmium, Oxides of 109 comportment of Salts of with re- agents 109 extraction of from Platinum ores, Berzelius's process ... . 400 Fritzsche's process ... . 401 Oxalates, comportment of with re- agents . 171 Oxalic Acid as a reagent ... .44 quantitative estimation of . . 484 Oxides metallic, their comportment with reagents ....... 59 analytical classification of ... 61 metallic, of the first and second groups, general remarks on the . 68 metallic, of the third group, general remarks on . 74 metallic, of the fourth group, gene- ral remarks on 83 metallic, of the fifth group (Section A), general remarks on . . . 109 metallic, of the fifth group (Section B), general remarks on ... 137 Palladium Oxide of, general charac- ters of comportment of Salts of, with re- agents estimation of separation of, from Copper . . . Phosphates, tribasic, alkaline charac- ters of different forms of, with agents Phosphoric Acid. See Acid, Phos- phoric. Phosphorus, estimation of in inor- ganic compounds Pipettes . Pitchblende, analysis of Platinum Bichloride of, as a reagent Binoxide of, general characters of comportment of solutions of Salts of with reagents ...... 107 108 396 396 1-JG 531 2 15 SSI r>6 127 INDEX. 661 Page Platinum, precipitation of as Ammo- nio-chloride 420 as Potassio-chloride .... 421 as Sulphide 421 separation of from Antimony, Ar- senic, and Tin 422 ores, analysis of 422 Potassa, Hydrate of, as a reagent . 45 general characters of 62 Carbonate of, as a reagent ... 45 Chromate of, as a reagent ... 52 Sulphate of, as a reagent ... 52 Bisulphate of, as a reagent ... 52 Bimetantimoniate of, as a reagent 53 Nitrate of, as a reagent .... 54 Bitartrate of, as a reagent ... 52 Potassium, Sulphide of, as a reagent 47 Ferrpcyanide of, as a reagent . .51 Ferridcyanide of, as a reagent . . 52 Cyanide of, as a reagent .... 52 Iodide of, as a reagent .... 55 comportment of Salts of with re- agents 62 quantitative estimation of as Sul- phate 25? as Nitrate 257 as Chloride 258 as Platinp-chloride 258 Precipitates, different forms of ... 5 washing of 9 drying of 11 Precipitation, meaning of the term . 5 Press for making blowpipe clay basins 33 Pyrophosphates, comportment of so- lutions of with reagents . . . 144 Reagents, their preparation and uses 41 Ehodium, Oxide of, general characters of 108 comportment of Salts of with re- agents 108 estimation of 398 separation of, from Copper . . . 398 extraction of, from Platinum ores 399 Kuthenium, extraction of from Plati- num ores ... . 400 Saltpetre grottoes in Ceylon .... 597 Seleniateof Lead, analysis of . . .431 Selenic Acid, estimation of .... 431 Selenious Acid, general characters of 130 comportment of with reagents . .130 Selenium, quantitative estimation of 480 separation of from other metals . 431 Silicates, natural quantitative analysis of 465 decomposition of, by acids . . . 468 by fusion with alkaline Carbo- nates 470 by Hydrofluoric Acid .... 471 by fusion with Carbonate of Lime 473 Silver, Nitrate of, as a reagent ... 54 Ammonio-nitrate of, as a reagent . 54 Oxide of, general characters of . 93 PABT II. Page Silver, comportment of Salts of with reagents 93 quantitative estimation of as Chlo- ride 351 as Sulphide 352 as Cyanide 353 estimation of in alloys, wet method 354 dry method 360 determination of in argentiferous galena 353 Soda, Carbonate of, as a reagent . . 45 Phosphate of as a reagent ... 53 Biborate of as a reagent .... 54 and Ammonia, Phosphate of, as a reagent 54 Hydrate of, general characters of 63 separation of from Potassa . . .259 estimation of in crude Potassa . . 260 indirect separation of, from Potassa 261 Sodium, quantitative estimation of . 258 as Sulphate 258 as Carbonate 259 as Chloride 259 comportment of Salts of with re- agents 63 Soils, mechanical analysis of .... 587 determination of the density of . 591 determination of the absolute weight of 591 chemical analysis of 591 determination of Phosphoric Acid in 592 organic constituents of .... 596 Solution, two kinds of 2 apparatus for 3 Specific gravity, definition of the term 234 of a solid, determination of . . . 225 of a powder 227 of a liquid, determination of . . 228 of a gas, determination of ... 232 of a vapour, determination of . . 235 of camphor vapour 238 Spirit lamps, Berzelius's 21 Russian 21 common 22 Sainte-Claire Seville's .... 22 Starch-paste, preparation of, as a re- agent 43 Steel mortar for blowpipe analysis . 33 Strontium, comportment of Salts of, with reagents 65 Oxide of, quantitative estimation of 278 as Sulphate 278 as Carbonate 278 separation of from Baryta . . . 278 from the Alkalies 279 Strychnia, general characters of, and reactions with reagents . . . 182 Sublimation, apparatus for .... 40 Succinates, comportment of solutions of with reagents 175 Sugar, its conversion into Humin, Ul- miu, and organic acids in the soil 598 Sulphides comportment of, with re- agents 166 2 Y 662 INDEX. Page Sulphocyanides, comportment of solu- tions of, with reagents .... 164 Sulphur, estimation of, in Iron and Copper Pyrites 447 in Sulphide of Lead .... 448 in Sulphide of Silver .... 448 in Sulphide of Bismuth . . . 448 separation of from metals, by Chlo- rine gas 448 in Nitrogenous organic com- pounds 450 in volatile organic compounds . 450 in organic compounds .... 528 Sulphuretted Hydrogen, general cha- racters of 166 quantitative estimation of ... 442 Sulphuric Acid as a reagent .... 43 Sulphurous Acid as a reagent ... 56 Tartaric Acid as a reagent .... 45 Tartrates, comportment of solutions of, with reagents 172 Tellurium, estimation of as Sulphide . 432 separation of from Selenium . . 433 from Selenium and Sulphur . . 433 Tellurous Acid, general characters of 131 comportment of solutions of with reagents 131 Thallium, properties of 350 Protoxide of, general characters of 347 comportment of Salts of with re- agents 349 Thorinum, Oxide of, general charac- ters of 71 comportment of Salts of with re- agents 72 Oxide of, separation of from Mag nesia and Lime Tin, Protochloride of, as a reagent Oxide of, general characters of 125 comportment of Salts of with re agents 125 Peroxide of, general characters of 126 comportment of solutions of with reagents 126 quantitative estimation of as Per- oxide 412 volumetric estimation of, Qaultier de Claubry's method .... 413 Stromeyer's method .... 414 Protoxide of, separation from Per- oxide 414 alloy with Copper, analysis of . 414 separation of from Antimony, Ber- thier' s method 415 Level's method " . 415 Eose's method 416 Tookey's method 418 from Arsenic 419 = from Arsenic and Antimony . 419 Titanic Acid, general characters of . 73 comportment of with reagents . . 73 estimation of in Silicates .... 436 separation of from metals of the fifth group 437 Page Titanic Acid, separation of, from Oxide of Iron 637 Titanium, estimation of as Titanic Acid 436 estimation of, in pig Iron . . . 437 Tungstates insoluble, analysis of . . 434 Tungsten Binoxide of, general cha- racters of 133 Acid of, general characters of . . 133 quantitative estimation of ... 433 Tungstic Acid, estimation of ... 433 Turmeric Paper as a reagent ... 42 Uranium, quantitative estimation of . 329 separation of from other metallic Oxides S29 from Nickel, Cobalt, and Zinc . 330 from Oxide of Iron 331 Sesquioxide of, general characters of 83 comportment of Salts of with re- agents 83 Urinorneter, Neubauer's 231 Vanadic Acid, general characters of . 134 comportment of solution of with reagents 135 Vanadium, Oxides of, general charac- ters of 133 comportment of Binoxide of with reagents 134 Vapour, determination of specific gra- vity of 235 Gay-Lussac's method .... 235 Dumas' method 236 Vegetables, analysis of ashes of . .619 method of Will and Fresenius . 619 method of Rose 626 method of Staffel 630 method of Way and Ogston . . 632 method of Strecker . . . .636 Vitelline, analysis of 557 Wash-bottles 9 Water, distilled 58 Water-oven and bath 249 Waters mineral, qualitative analysis of 565 quantitative analysis of .... 566 estimation of free Carbonic Acid in 567 estimation of Iodine in . . 572 estimation of Bromine in . . . 572 estimation of Nitric Acid in . . 573 estimation of Ammonia in . . 573 estimation of Lithia in . . 574 estimation of Phosphoric Acid in . 574 estimation of Fluorine in . . . . 574 estimation of organic acids in . . 675 determination of the hardness of natural 578 arrangement of the results of the analysis of 580 determination of the volume of gases in 583 of Purton, analysis of .... 581 INDEX. C63 Page Weighing 213 rules to be observed in .... 221 Weights, English 223 French ; . . 223 Wolfram, analysis of 434 Yttrium, Oxide of, general characters 71 comportment of Salts of with re- agents 71 Oxide of, quantitative estimation of 288 separation of from Alumina and Glucina 288 from Magnesia and the Alkalies 289 Zinc, Oxide of, general characters of . 75 comportment of Salts of with re- agents 75 Page Zinc, Oxide of, quantitative estimation of 294 precipitation of as basic Carbonate 294 as Sulphide 294 separation of from Oxide of Chro- mium 295 from Alumina 295 from Magnesia 296 from the alkaline earths and Al- kalies 296 ores, volumetric analysis of ... 296 Zirconium, Oxide of, general characters of 72 comportment of Salts of with re- agents 72 Oxide of, quantitative estimation of 289 separation of from Alumina . . . 289 THE END. JOHN EDWABD TAYLOR, PBINTEB, MTTEE QUEEN STBEBT, LINCOLN'S INN FIBLBS. 14 DAY USE RETURN TO DESK FROM WHICH BORROWED LOAN DEPT. This book is due on the last date stamped below, or on the date to which renewed. Renewed books are subject to immediate recall. ..1 ^OAH^kO^-'^ y mbi - ~'n LD i.r HWS 1 *- 'JUt i 01QR r-la'^TW L ,x^ VU^-S Rec'o LD ^N6 '64- REC'D LD MAY 19 '64 -11 AM LD 2l (B6221slO)476B General Library University of California Berkeley YB 16886 UNIVERSITY OF CALIFORNIA LIBRARY