£)tafe College of Agriculture at Cornell mnibetsitp Slliaca. M. B- ILibtavp ®P.t.-. Cornell University Library QD 101.0541922 A text-book of quantitative chemical ana 3 1924 002 984 478 .. Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924002984478 . A TEXT-BOOK OF QUANTITATIVE CHEMICAL ANALYSIS BT GRAVIMETRIC, ELECTROLYTIC, VOLUMETRIC AND GASOMETRIC METHODS WITH SEVENTY-FOUR LABORATORY EXERCISES THE ANALYSIS OF PURE SALTS, ALLOYS, MINERALS AND TECHNICAL PRODUCTS J. C. OLSEN, A.M., Ph.D. Professor of Chetnical Engineering, Polytechnic Institute, Brooklyn, N* Y,; Secretary, American Institute of Chemical Engineers; formerly Professor of Chemistry, Cooper Union, and Fellow, Johns Hopkins University; Editor. "Van Nostrand*s Chemical Annual." FIFTH EDITION, REVISED AND ENLARGED TOTAL ISSUE 12,000 NEW YOEK D. VAN NOSTRAND COMPANY Eight Warren Street T 7711] Copyright, 1908 ' BY D. VAN NOSTRAND COMPANy Copyright, 1916 BT D. VAN NOSTRAND COMPANY PREFACE TO THE FIFTH EDITION. In the preparation of this edition, all atomic and molecular weights as well as factors of weighable precipitates have been recalculated by the 1916 atomic weights. The quantitative methods described have been revised where recent investigations have indicated improvement in the procedure, and where trial in the laboratory has demonstrated the advantage over older methods. Among the new methods inserted may be mentioned Allen and Bishop's method of precipitation of sulphuric acid and oxidation of sulphur in sulphides, the standardization of potassium perman- ganate solutions by means of Sorensen's sodium oxalate, the general revision of the methods of analysis of water in conformity with the methods given in the Standard Methods of Water Analysis, and also the introduction of the determination of dissolved oxygen in water, which has come to be. recognized as one of the most significant criteria of the purity of water. The table of solubility of oxygen in fresh and salt water by Geo. C. and M. C. Whipple has been added to the Appendix. J. C. Olsbn. Cooper Union, New Yoke, October 2, 1916. ill PREFACE TO THIRD EDITION. Two editions of this text-book haying been exhausted since its publication, four years ago, the author has talcen advantage of the printing of the third edition, to make such corrections and additions as his own experience and that of a number of its users, who have been kind enough to suggest improvements, has shown to be advisable. The author has endeavored in this revision to adhere to the original plan of the work, namely, to describe no method which has not been thoroughly tried in his own as well as other laboratories and found to be satisfactory, believing that only after continued use by many workers are all the conditions neces- sary to the successful carrying out of a method discovered, so that it can be accurately described. The new methods which have been introduced in this edition are principally the following. Penfield's methoc^ of determining water has been described because it is believed that this simple and accurate method should be more widely known. The precipi- tation of nitric acid by means of nitron is important as the only gravimetric method for determining this common acid. The rota- tion of the anode or cathode hi electrolytic determinations has proved so advantageous that its use is fully described, as well as the use of electric and water motors for this purpose. The methods for analyzing water have been revised and brought into better accord with the standard methods adopted by the Ameri- can Public Health Association. The simple bacteria count in water and the test for coli have also been described, as the prep- aration of the culture media requires considerable chemical skill. A description of various methods of determining specific gravity as well as the determination of the calorific value of fuels has also iv PREFACE. V been introduced, since chemists are so often called upon to make such determinations. A number of typical chemical problems have been introduced throughout the book in order to develop the ability to make chemical calculations, which seems so difficult for many students to acquire. A number of less important additions have been made through- out the book, among which may be mentioned: Rules for using and cleaning platinum ware; volumetric methods for determining antimony, copper, and testing shellac for rosin. The method of determining the Polenske value, the flash point, and the viscosity of fats and oils, as well as the separation of mineral from animal and vegetable fats and oils. All atomic and molecular weights as well as the table of factors have been recalculated by the 1908 table of atomic weights. The description of methods has been improved wherever experience has shown the necessity. While the author has given the precautions necessary to obtain correct results with all methods described, he has endeav- ored to avoid producing a book which can be followed mechanic- ally with the production of good analytical results, but poor chem- ists. In writing the exercises he has assumed that the student has studied the text-book and knows the reason for each step in an analysis. The author is of the opinion that the student can- not successfully perform these exercises unless he has carefully studied and understood the text. It is undoubtedly for this rea- son that the book has not been found adapted for use in large laboratory classes where a sufficient number of competent instruc- tors has not been provided. The author has been very much encouraged by the favorable reception accorded the book in the past and hopes for a still larger field of usefulness in the future. He takes this opportunity of thanking all those who have been kind enough to point out errors or suggest improvements. He desires especially to express his obligation to Mr. Albert Seeker for the revision of the chapter on oils and fats and to Mr. Andrew Mayer, Jr. chemist with the National Lead Co. for the revision of the chapters on the analysis of alloys. •■• Pb(N0,)2 R. R. R. R. R. D. R. R. R. R. 36.5 76.9 20.3 52.3 119.7 37.2 35° 51.3 7.39 75° 104.6 127 Magnesium. . Manganese. . . Mercury MgS04.7H,0 671.2 K2S04.MgSO,.6H20. . . MnSO,.(NH,),S04.6H,0 HgCL 6.73 53.96 Metal Nickel NiS04.(NH,),S04.6H,0 KNO3 '13.3 28.5 122 35^ 2.5 8.1 31.2 34.7 227 85° 39.5 247'"' KCl 56.6 AeNO, 940 Metal Sodium NaCl P. R. P. R. R. 35.7 36.0 35.8 39.8 Na^CsOa 6.33 Strontium. . . SrCO, Tin SnCl4.2NH4Cl ZnS04.(NH4)2S04.6H20 10 14.5° 33 Zinc 17.3 80° 67.8 In the column headed " Method of Purificationj" the letters have the following signification: R., recrystallization; D., distillation; S., sublimation; I., ignition of nitrate, and P., precipitation. 32 GE-JEliAL OPERATIONS. COMPOUNDS OF THE METALS AND ACIDS— {Continued). AQIDS. Compounds. ^1 Parts of the Salt Dissolved by 100 Parts of Water at 0° 20° 100° Hydriodic. . . I s. R. R. R. P. KI 127.9 35.7 16° 1.53 144.2 17° 5.4 64.5 36.0 209 KHaOs)^ KBr Hydrobromic Hydrochloric. 102 NaCl 39.8 Nitric HNO3 KNO, R. R. R. R. R. R. R. 13.3 31.2 11.1 247 Oxalic HAO4.2H2O (NHJ.CA KH3(CA).-2H,0 . . Na2HP04.12H,0. . . NaNH4HP04.4H,0. KH,P04 350 2.2 15° 4.2 Phosphoric. . 6.3 23.4 16 236.8 100 Sulphuric. . . . H2SO, K,S04 R. 8.5 10.9 26.2 In the column headed "Method of Purification," the letters have the following signification: R., recrystallization; D., distillation; S., sublima- tion; I., ignition of nitrate, and P., precipitation. metal by passing hydrogen sulphide through the acid solution, filtering off any precipitate, and then adding ammonia and again filtering. To the solution thus purified, ammonium carbonate solution is added with vigorous stirring. The precipitated car- bonate is then thoroughly washed and dried. This method can- not be used for the preparation of the pure carbonates of the other metals, because these elements do not form stable carbonates of definite composition. NOTES. Ammonium Chloride as purchased varies greatly in purity. Phosphoric arsenic, sulphuric, and sulphocyanic acids may be present as well as heavy metals and earths and aniline derivatives. The pure salt is best prepared by neutralizing C.P. hydrochloric acid by a stream of ammonia- gas produced by heating pure concentrated ammonium hydrate. When the acid is neutralized the solution is made slightly acid with hydrochloric acid and evaporated to dryness in a porcelain dish and dried on the water-bath. Ammonium Oxalate is examined in the same manner as oxalic acid. It does not lose water of crystallization as readily as the acid. PURE SALTS 33 Ammonium Sulphate may be made by the method given for ammonium chloride, but must be dried at 120° and the solution must be neutral or alkaline before evaporating to dryness. Arsenlous Oxide may be tested for sulphide by heating 1 gram in a test-tube. The first sublimate which forms must be pure white. At the end of the operation no residue must remain in the tube. Barium Carbonate must be prepared by precipitation of the C.P. chloride with ammonia and ammonium carbonate. The washed precipitate is gently ignited in a platinum dish. Barium Chloride may be tested for metals by passing hydrogen sulphide through the acid solution and then rendering it alkaline with ammonia. Five grams of the salt are dissolved in 200 c.c. of water, acidified with a drop or two of dilute hydrochloric acid, heated to boiling, and the barium precipitated with sulphuric acid. After digesting for some time, the precipitate is filtered off and the filtrate evaporated to dryness in a platinum dish. The alkalies, lime, etc., will be found in the residue, which should be less than .1% in the C.P. salt. Bismuth Oxide may be prepared pure by dissolving the commercial article in nitric acid, diluting largely with water, filtering and washing the precipitate, which may be dried and ignited in a platinum dis"h. A still purer product may be obtained by dissolving the basic nitrate in pure nitric acid, precipitating with ammonia and ammonium carbonate, washing and igniting. Cadmium Iodide may be tested for tin, lead, copper, zinc, and other metals by dissolving 2 grams in water and adding nitric acid. A white pre- cipitate indicates tin. To one-half of the solution a large excess of ammonia is added. The solution must remain clear and colorless. The other part of the solution is diluted with water and filtered after adding excess of potash solution. Pass hydrogSn sulphide through the filtrate and acidify. No precipitate must be produced either in the alkaline or acid solution. Calcium and Strontium Carbonates are prepared by the method given for barium carbonate. During the drying, especially of the calcium car- bonate, the platinum dish must not be allowed to become even faintly red. It is advisable to support the dish about an inch above a wire gauze, which is heated by the Bunsen burner. A very pure form of calcium carbonate may be purchased in the form of Iceland spar. Copper prepared electrolytically and of a high percentage of purity may be readily obtained. Copper Sulphate. — This salt must almost invariably be recrystalUzed and carefully dried in order to insure the presence of the theoretical amount of water. The dry crystals on exposure to the air effloresce. The loss of water is indicated by the appearance of white spots in place of the clear blue of the hydrated salt. Ferrous Ammonium Sulphate, frequently called Mohr's salt, may be purchased remarkably pure. The crystals should be small and free from 34 GENERAL OPERATIONS. ferric iron. A gram of the salt dissolved in boiled water which has been allowed to cool out of contact with the air must show only a faint red with thiooyanate. Hydrochloric, Sulphuric, and Nitric Acids may be easily obtained almost absolutely free from impurities. The percentage of acid present may be ascertained from the tables pp. 479-486 after carefully taking the specific gravity. Iodine must always be purified by sublimation. For this purpose the commercial article is mixed with one-fourth or one-fifth of its weight of potassium iodide by grinding them together in a porcelain mortar. The material is placed in a watch-crystal or crystallizing-dish, which is placed on a piece of asbestos board having a hole cut in the centre. The iodine ij sublimed by gentle heat from a Bunsen burner, the first portion of the vapor being allowed to escape as the material is frequently quite moist. Another watch-crystal or crystallizing-dish is now placed over the first one, the upper one being kept cold by means of moist filter-paper or cloth. The upper dish is removed before all of the iodine is sublimed. It is best to sublime the iodine immediately before use as it is very hygroscopic. Iron. — The soft iron wire, which is sold for standardizing purposes, con- tains from. .2 to .4% of carbon. For careful work the percentage of iron or carbon must be determined gravimetrically. The DOUBLE AMMONIUM SALTS of MANGANESE, COBALT, NICKEL, and ZINC must be prepared by the general methods given. Mercury. — On being shaken in a clean dry bottle the surface must remain perfectly bright and the metal must not adhere to the surface of the glass. Mercuric Chloride may be tested for mercurous chloride by treating a gram of the finely ground sample with 15 c.c. of pure ether. It must dis- solve completely. The filtrate from the^ hydrogen-sulphide precipitate must leave no weighable residue. On testing this precipitate with dilute ammonia, filtering and acidifying the filtrate with hydrochloric acid, no yellow precipitate or color must be produced indicating absence of arsenic. Oxalic Acid must leave no residue on ignition. An alkaline residue indicates the presence of sodium or potassium. Sulphuric acid is tested for by dissolving 5 grams in 100 c.c. of water, adding a few drops of dilute hydrochloric acid and barium chloride. No precipitate must form after standing several hours in a hot place. Ammonia and ammonium sulphide must produce no coloration. On heating 2 grams in a test-tube with caustic- soda solution no ammonia must be given off. A very pure aci'd may be made by dissolving the impure material in a mixture of equal parts of alcohol and ether, filtering, evaporating off the alcohol and ether after the addition of water, and recrystallizing the product. Potash Alum. — Recrystallization and carefully drying may be necessary to insure the correct percentage of water. PURE SALTS 65 Potassium-acid lodate may be purchased remarkably pure. Meineke found two samples 100.001 and 100.010% pure. Potassium Bromide and Iodide can usually be obtained quite pure. The presence of bromates or iodates is shown by the blue color produced by added starch solution and a little dilute hydrochloric acid to the dilute solution and in the case of the bromide a little potassium iodide. The heavy metals and sulphuric acid are tested for in the usual manner. The salts must be dried at 100°. Potassium Bichromate is generally obtained very pure. The com- mercial samples frequently contain sulphates or chlorides. The C.P. or recrystallized salt may be dried by heating until the salt is fused. It should not be heated higher, and all organic matter must be carefully excluded. Potassium Nitrate is also sold in a very pure condition. Manufacturers generally guarantee a maximum impurity of sodium chloride of about \ part in 10,000 or .01%. Potassium Permanganate, which is very nearly 100% pure, may be readily obtained. Chlorides, sulphates, and nitrates, which are present in the commercial article, are absent from the C.P. salt. The latter, however, invariably contains small quantities of manganese dioxide. Potassium Sulphate must give a clear neutral solution (1 to 20) and must give no reaction with hydrogeij-sulphide water, ammonium oxalate, potassium carbonate, or silver nitrate. It may be dried at 100°. Potassium Tetroxalate may be prepared by making a hot saturated solution of oxalic acid, neutralizing one-fourth of it with pure potassium carbonate and adding this with stirring to the remainder of the solution. The product is still further purified by recrystallization. Silver Nitrate. — Good commercial samples have been found to contain from .01 to .03% of impurity. The crystals should be pure white. Metallic Silver may be obtained of a high state of purity guaranteed by government assay. Sodium Oxalate. — For tests and purification, see page 259. Sodium Phosphate and Sodium-ammonium Phosphate must be neutral to phenolphthalein and free from sulphates, chlorides, and nitrates. If the crystals are not transparent, but show signs of efflorescence, the salt must be recrystallized and carefully dried. Sodium and Potassium Chlorides may be purchased pure enough for most quantitative work. The method of purification of sodium chloride given in Experiment 6 yields a very pure product. Zinc and Magnesium Sulphates must always be recrystallized and care- fully dried. The double sulphate of magnesium and potassium is pre- pared as directed in Exercise V. Many of the tests given in this section have been quoted from "Testing of Chemical Reagents," by Dr. C. Krauch. The student is referred to this book or to "Chemical Reagents, their Purity and Tests," by E. Merck, for the testing of other salts. 36 QUASTITATIVE ANALYSIS. EXERCISE 4. Preparation of Pure Copper Sulphate, CuSOj-gH^O. Weigh out about 150 grams of crystallized copper sulphate. Crush the large crystals in a porcelain mortar. Dissolve in about 150 c.c. of hot distilled water. If the solution is not perfectly clear, filter, using a folded filter-paper and a hot-water funnel. The clear filtrate must be vigorously stirred as soon as crystals begin to separate. Cool the solution by placing the beaker in cold tap-water or in ice-water. Continue the stirring until no more crystals sepa- rate out. Filter off the crystals immediately, using a funnel of suitable size with a platinum cone or a perforated porcelain dish in the bottom of the funnel. Connect the funnel by means of a rubber stopper to a filter-flask and remove as much of the mother-liquor as possible by means of strong suction from the filter-pump. Press the crystals into a compact mass by means of a spatula. Pour about 50 c.c. of ice-water over the crystals and suck it through as rapidly as possible with the suction. When no more liquid comes through, spread out the crystals by means of a spatula on an unglazed porcelain plate or on several folds of filter-paper. When no more liquid seems to be absorbed , spread the crystals out in an even layer. Cover them with a piece of filter-paper and allow them to dry in the air for several hours with occasional stirring. Test for moisture from time to time by placing a few crystals on a clean and dry porcelain or glass surface. Dry crystals will not adhere to such a surface. As soon as dry, transfer to a clean dry bottle with ground -glass stopper. Care must be taken in removing from the filter-paper not to detach shreds of filter-paper with the crystals. If the crystals have been exposed too long to the air, portions will appear white from loss of water of crystallization. If the work has been carried on in a room containing hydrogen sulphide or ammonium sulphide in the atmos- phere, black spots will appear on the crystals. If such crystals cannot be removed the work must be repeated. EXERCISE 5. Preparation of Potassium-magnesium Sulphate, K2SO4.MgSO4.6H2O. Weigh out about 163 grams of magnesium sulphate and dissolve in 115 c.c. of water. Dissolve 87 grams of potassium sulphate in 350 c.c. of water. If necessary, filter the solutions separately, using the hot-water funnel and folded filter-papers. Pour the hot solution of potassium sulphate rapidly and with vigorous stirring into the hot magnesium solution. Stir con- stantly while the double salt crystallizes out, and finally cool the solution by surrounding the beaker with cracked ice. When no more salt crystallizes out, filter the crystals off in a funnel with a platinum cone or perforated porce- lain plate in the bottom. Suck the mother-liquor off by means of the filter-pump, pressing the crystals together with a porcelain spatula. Wash the crystals with 50 c.c. of ice-cold distilled water, sucking them dry with the GENERAL OPERATIONS. 37 pump. Transfer the crystals to an unglazed porcelain plate or several folds of filter-paper. Remove the motlier-Iiquor by pressing the crystals on the porcelain plate with a porcelain spatula or by pressing them with the hand between the folds of the filter-paper. Finally, spread them out to dry in the air, testing them for moisture from time to time by placing them on a clean and dry porcelain or glass surface. The dry crystals will not adhere to such a surface. When dry, transfer to a clean and dry bottle. EXERCISE 6. Preparation of Pure Sodium Chloride. About 70 grams of common salt are weighed out and dissolved in 200 c.c. of distilled water. If necessary, the solution is filtered. Gaseous hydrochloric acid is now passed into the solution. It is generated in the flask A, which Fig. 7. contains about 150 grams of common salt. 300 grams of concentrated sul- phuric acid are poured into 220 c.c. of distilled water and allowed to cool somewhat. This acid is transferred to the dropping-funnel and allowed to drop slowly on the salt. 5 is a wash-bottle containing concentrated hydro- chloric acid to cat3h any sodium sulphate which might spatter over. The 38 QUANTITATIVE ANALYSIS. thistle-tube C touches the salt solution in the beaker. Toward the end ol the operation the flask may be heated so as to expel the hydrochloric acid. The stream of hydrochloric acid is allowed to pass until no more precipita- tion is noticed in the salt solution. The crystals are filtered off and washed with concentrated c.p. hydrochloric acid. When no more liquid can be sucked out by the pump, the sodium chloride is transferred to a porcelain dish and heated -syith the Bunsen burner until dry and free from hydro- chloric acid. The pure salt is then transferred to a clean and dry glass- stoppered bottle. RULES FOR USING AND CLEANING PLATINUM UTENSILS. Platinum when hot easily forms alloys with other metals. For this reason metals must never be heated in platinum. The same rule applies to all compounds of easily reducible metals, such as silver, copper, lead, and bismuth. Even stannic oxide is sometimes reduced to metallic tin, which alloys with the platinum. Heat platinum only with a clear blue flame. A deposit of carbon is injurious. The vessel should also be placed beyond the inner cone of the flame, as reducing gases may pass through the hot platinum. As phosphorus and sulphur combine with platinum, making it brittle, compounds of these elements should not be heated in platinum under reducing conditions. Caustic alkalies should not be fused in platinum as it is oxi- dized and dissolved. Nickel, iron, silver, or copper dishes may be used for this purpose. The carbonates of the alkalies may be fused in platinum, although a little platinum is dissolved, especially when potassium nitrate is present. Platinum vessels may be cleaned by digestion with hydro- chloric or nitric acids, but not with a mixture of these acids. Basic impurities not removed in this manner may be dissolved by fused acid potassium sulphate, while acid impurities such as silica may be removed by fused sodium carbonate. The surface of platinum should be kept smooth and bright by polishing with sea sand which is free from sharp particles. Platinum vessels change weight very slowly, losing a few tenths of a milligram after strong ignition, a little more after a fusion, especially with an alkaline oxidizing mixture, and still more on being polished. CHAPTER III. DETERMINATION OF WATER. 35. Conditions in which Water is Held. — Water is present in the great majority of substances submitted to chemical analysis, and the amount present must frequently be determined. It may be held mechanically or combined chemically as water of crystallization, or it may be an inherent part of the molecule. In considering its determination, the manner of separating it from the other constituents present must be studied as well as the method of weighing it. The simplest method of determining it involves heating the compound to the temperature at which the water volatilizes. This temperature differs greatly, ranging from the ordinary temperature of the air to that of the blast-lamp. Where the water is held mechanically, as in most minerals, as well as metals and non-crystalline substances, a temperature of 100° or a few degrees above that point is sufficient to dry the substance completely. In the case of salts which crystallize with water, the temperature at which it is complete^ given off varies from 100° centigrade to a red heat. In some cases a definite number of molecules of water of crystallization will be driven off at one temperature, while a higher temperature is required to completely dehydrate the substance. The silicates, as a class, hold water with great persistence. Glass maj'- be heated to a red heat for months, and still give off a trace of water. In this case the water is probably combined with the silica, forming a more or less complicated silicic acid. 36. Methods of Determining Water.— Whenever a substance can be dried completely by simply heating it to a temperature at which no other constituent is volatilized, the amount of water may be determined by weighing the substance before and after drying it. In this case the amount of water present may also 39 40 QUANTITATIVE ANALYSIS. be determined by allowing the water to be absorbed by a weighed amount of some hygroscopic substance, like calcium chloride or concentrated sulphuric acid. Such a determination would give the total amount ' of water present. This water may be present either as hygroscopic water or as water of crystallization or of CONSTITUTION. If it is necessary to determine the amount of each, the sub- stance must first be heated to a temperature at which only the hygroscopic water is given off. At a temperature of 105° the hygroscopic water will certainly be expelled in every case, but as at that temperature many crystals lose a part or all of their water of crystallization, a lower temperature must sometimes be used. In such cases, the substance is placed in a desiccator in which the air is dried by sulphuric acid or calcium chloride. Even under these circumstances many crystals lose water. By diluting the sulphuric acid, however, an atmosphere may be obtained having a tension of water vapor equal to or greater than that produced by the water of crystallization. Long standing in such an atmosphere will free the crystal of adherent or hygroscopic water. 37. Sources of Error. — ^When substances lose other constitu- ents than water on heating, the error may frequently be avoided by adding an anhydrous substance which is capable of retaining the volatile substance. The oxides of calcium, lead, and bis- muth have been used as retainers for fluorine, chlorine, sulphur, etc. Not only must loss of volatile constituents on heating be guarded against, but a possible increase in weight from the ab- sorption of oxygen, carbon dioxide or air must be kept in mind. Metals in their lower state of oxidation may absorb oxygen, and pass to the higher state of oxidation. A familiar example of this is ferrous sulphate, which will absorb oxygen even at the ordinary temperature. Sulphides of some metals easily take up oxygen, becoming sulphates. Some basic oxides of the metals absorb carbon dioxide, forming carbonates. Some minerals, such as the zeolites, have been found to absorb air after the water has been expelled. DETERMINATION OF WATER. 41 EXERCISE 7. Determination of Water of Crystallization in Copper Sulphate, CUSO4.5H2O. 38. Weighing the Salt. — A watch crystal of convenient size is weighed, a one-gram weight is added to the right-hand pan and recrystalhzed copper sulphate is placed on the watch-crystal by means of a spatula until equi- librium is restored. Exactly one gram of copper sulphate will then have been weighed out. In placing the salt on the watch-crystal by means of the spatula, the latter should not be allowed to touch the watch-crystal nor the scale-pan. The spatula with a moderate amount of the salt in it is held in the right hand so that by tapping it gently with the left hand the crystals will fall on the watch-glass. The arms supporting the beam are slightly lowered by the central thumb-screw, so that, by releasing by means of the left hand the spring supporting the scale-pans, the pointer will indi- cate when enough salt has been added by remaining at zero or moving to the right. After a little experience the quickness with which the pointer moves will indicate the amount of salt to be added or taken away. When the pointer moves slowly to the left, only a few small crystals at a time are allowed to fall on the watch-crystal until the pointer ceases to move. The swings may then be taken if necessary and the final adjustment made. If too much has been added, a considerable amount should be very care- ' fully taken off, and added in small amounts as before. When exactly a gram is weighed out, the calculation of percentage is very simple, as the number of milligrams of water lost is equal to the tenths of per cent. At first it may seem easier to calculate from an amount weighed out which is only approxi- mately one gram than to endeavor to obtain exactly one gram, but it is found that an experienced worker can weigh more readily than make the required calculations. 39. Determination of Four Molecules of Water. — The watch-glass with copper sulphate is now transferred to an air-oven, heated to 115°. The temperature is regulated by adjusting the flame of the Bunsen burnei" from time to time until the thermometer indicates a constant temperature After about one hour the copper sulphate is transferred to a desiccator, and after ten minutes weighed. It is again heated in the air-bath one-hal hour and cooled and weighed. This is repeated until the weight is con stant. In most analytical work a weight is considered constant if the change is not greater than 0.3 of a milUgram, since errors of weighing may be as great as this. Frequently a change of weight of 1 milligram warrants a discontinuation of the heating, since the next heating would probably produce a much smaller change in weight. As the watch-crystal may have lost weight on heating, the copper sulphate is brushed off and the watch-crystal weighed. The difference between this weight and the last weight of the dried copper sulphate and the watch-crystal gives the weight 42 QUANTITATIVE ANALYSIS. of the dried copper sulphate. The difference between the weight of the dried copper sulphate and of the crystals taken is the weight of water expelled. The percentage of water lost in this manner will be found to be very nearly that calculated for 4 molecules of water or 28.85%. The correct weight of the watch-crystal may also be obtained by drying it in the air-bath at 115° until its weight is constant before placing the copper sulphate on it. 3oa. Determination of the Fifth Molecule of Water. — If the air-bath can be heated to 210°-215°, the fifth molecule of water may be determined in the same manner as the first four. For this purpose either another gram of the crystals may be weighed out and the total amount of water deter- mined or the portion from which the 4 molecules have been expelled may be heated to constant weight at 210° to 215° and the percentage correspond- ing to the fifth molecule determined. If the air-bath cannot be heated to 215°, the watch-crystal containing the copper sulphate may be placed on a sand-bath which consists of an iron plate filled with sand and placed on a tripod so as to be heated by means of a Bunsen burner. Until the bulk of the water has been expelled, the thermometer-bulb should be immersed in the sand so as to touch the watch-crystal, which should be Fig. 8. pressed down firmly in the sand. When the copper sulphate has been heated about one hour in this manner, the bulb of the thermometer after being freed from sand, should be placed on the copper sulphate in a slantino- position. The watch-crystal should be covered with another and the Bunsen burner regulated until the thermometer indicates 210° to 215° and the heating continued for one-half hour, when the copper sulphate is carefully brushed off the thermometer and the former cooled in the desicca- tor and weighed. The heating with the thermometer on the copper sulphate is repeated until the weight is constant. The watch-crystal is now weighed DETERMINATION OF WATER. 43 after removing the copper sulphate and the percentage of water calculated. The theoretical percentage for 5 molecules is 36.07. EXERCISE 8. Determination of Water of Crystallization in Barium Chloride, BaCl2.2H20. A crucible, preferably of platinum, with its lid, is placed on a pipe-stem triangle and heated with the Bunsen burner for a few minutes, cooled in a desiccator, and weighed. Two grams of pure crystallized barium chloride are weighed out and transferred to the crucible. The crucible is placed on the pipe-stem triangle and heated gently by a small Bunsen burner flame, the Hd being on the crucible. The temperature is gradually raised until the crucible attains a low red heat, at which it is maintained for about ten minutes. It is then cooled in a desiccator and weighed. The heating and weighing is repeated until the weight is constant. Theoretical per- centage of water of crystallization in crystallized barium chloride is 14.74. 40. Penfield's Method of Determining "Water. — A very simple and accurate method of determining water, especially when it is present in email amount, is that of Penfield.* The apparatus used consists simply of a glass tube closed at one end and having one or two bulbs blown on the middle of the tube If a high tem- perature is required to expel the water, a hard glass tube must be used, otherwise ordinary soft glass tubing may be used. Tubes 20 cm. long and 6 mm. internal diameter will be found a convenient size. When the amount of water given off is large, a bulb should be blown on the middle of the tube. The tube is first dried by placing in an air-bath heated to a little above 100° and sucking out the moist air by means of a smaller tube. The tube is then cooled and weighed and the substance to be analyzed is carefully transferred to the bulb at the closed end of the tube, which is again weighed, the difference between this weight and the weight of the empty tube giving the weight of the substance analyzed.- A short piece of glass tubing drawn out to a capillary is attached to the Penfield tube by means of a short piece of rubber tubing. This simple device prevents the loss of water by circulating air currents. Moist filter-paper or cloth is wrapped around the bulb in the middle of the tube, and the bulb at the end containing the * Am. Jour. Sci., 3d series. Vol. XL VIII, p. 31 . 1894. 44 QUANTITATIVE ANALYSIS. substance being analyzed is heated high enough to expel the moisture. This may generally be accomplished with the Bunsen burner, although some substances require the blast-lamp. The water condenses on the cold middle portion of the tube and col- lects in the bulb, which prevents it from running back to the heated portion and cracking the glass. When all of the water has been expelled the flame is moved so as to melt off the bulb containing the substance being analyzed; when this has been accomplished the tube is cooled and weighed after removing the capillary tube at the end. The tube is now placed in the air- bath and dried as in the beginning of the experiment. It is then cooled and again weighed, the loss of weight being equal to the amount of water obtained. If other volatile substances besides water are given off, a retainer must be used. The oxides of calcium, lead, or bismuth have been used for this purpose. The retainer need not be weighed, but it ipust be thoroughly dried before being placed in the tube, and enough should be added after the weighed substance to be analyzed has been introduced to fill the bulb and form a layer about two centimeters deep. The retainer must be heated first and then the flame is increased so as to also heat the bulb. When all of the water has been expelled the part of the tube con- taining the retainer is melted off and the determination com- pleted as already described, when no retainer is required. The Penfield method is well adapted to the determination of water in minerals and substances which require high heat in order to expel the moisture. The results are accurate if the operation is careiully carried out, although only moderate amounts of water can be weighed in this manner. 41, Efficiency of Various Drying-agents for Gases. — ^When the water is to be weighed directly, it must be absorbed in a weighed amount of some dehydrating agent. The substances commonly used for this purpose are phosphorus pentoxide, concex- TRATED sulphuric ACID, and fused calcium chloride. Phos- phorus pentoxide is the most efficient drying-agent for gases known. It is doubtful if any water at all remains in a gas which has been dried over this reagent. On the contrary, gases dried over concentrated sulphuric acid give up moisture to (Jiosphorus DETERMINATION OF WATER. 45 pentoxide, while gases dried with fused calcium chloride give up moisture both to concentrated sulphuric acid and to phosphorus pentoxide. One litre of air dried with calcium chloride at 15° C. gives up 1 milligram of water to concentrated sulphuric acid, while phosphorus pentoxide is capable of absorbing .002 milli- gram of water per litre after the air has been dried over con- centrated sulphuric acid. Equally correct results, however, may he obtained by the use of the less efficient dehydrating agents if care is taken to dry the air when it enters the apparatus to the same extent as when it leaves. This object is best secured by using the same dehydrating agent for both purposes. Some moisture will be present in the air which passes over the hydrated substance, but as exactly the same amount of water is present in the air on leaving the second set of drying-tubes the water actually absorbed by these tubes will be that taken up within the apparatus only. 42. Drying Properties of Fused Calcium Chloride. — When cal- cium chloride is used care must be taken to keep all of the tubes in a series of the same temperature, as this salt differs in its efficiency at different temperatures. At 0° centigrade, 1 litre of air after being dried over calcium chloride contains .3 milli- gram of water; if dried at 15° it contains 1 milligram of water, while if dried at 30°, 3.3 milligrams of water remain. In one ex- periment where air was passed through a series of U-tubes (A, B, C, D, E) filled with calcium chloride, which were main- tained at different temperatures, the gains and losses in the weights noted below were observed. Before entering the first tube. A, the air was dried by passing through a tube of calcimn chloride maintained at the temperature of A. Tubes. Temperatures. Changes in Weight. A 16.9° .0000 gram B 0.0° .0452 " C 30.2° .1565 " D 0.0° .1569 " E 16.9° .0442 " Total change in weight 0014 gram It is extremely difficult to obtain fused calcium chloride -which is entirely free from calcium oxide. It will, therefore, be found to absorb a small amount of carbon dioxide. To neutralize the 46 QUANTITATIVE ANALYSIS. calcium oxide the calcium chloride should be kept in an atmo- sphere of carbon dioxide or hydrochloric acid for sopie time and then the acid gas displaced by air. This is best accomplished by filling tubes with it and passing a slow stream of the acid gas through them, and finally drawing air through. When moist air passes over calcium chloride some of it dissolves in the water absorbed, and this liquid mass on the surface of the lumps after- ward solidifies. The so-called Marshand tube is, therefore, best adapted for use with calcium chloride. This U-tube has a bulb blown in the exit tube on one side. The moist air should be led in at this point, so that the water may condense in the bulb, which may be kept cold by moist filter-paper wrapped around it. The calcium chloride in the half of the tube nearest the bulb should be in large lumps, so that the tube shall not be clogged by the dissolved calcium chloride. The other half of the U-tube should be filled with finer material, so as to dry the gas thoroughly. A plug of glass wool should be inserted above the fine calcium chloride, so as to prevent small particles from being carried out by the current of air. 43. Drying Properties of Concentrated Sulphuric Acid. — Con- centrated sulphuric acid- is about equally efficient at 0° and 30° centigrade. It should be free from sulphur dioxide. If pres- ent, this gas may be expelled by passing a current of air through the acid. If the acid is at all dark or brown in color from the presence of organic matter it should not be used, as sulphur diox- ide will result from the slow oxidation of the organic material. It absorbs carbon dioxide to an appreciable extent, but this is wholly expelled by passing air through it. U-tubes should first be partly filled with glass beads and then sulphuric acid added until the beads are thoroughly moistened and the liquid just forms a seal at the bottom of the tube. Care must be taken that sulphuric-acid U-tubes are alwaj^s held upright, so that the acid shall not come in contact with the stoppers or enter the exit tubes. It is therefore most convenient to attach a platinum wire, so that they may be suspended during the weighing and at other times. This wire should be hft on the tube. 44. Drying Properties of Phosphorus Pentoxide. — Phosphorus Dentoxide is seldom used_, because of the difficulty of manipu- DETERMINATION OF WATER. 47 lating tubes containing it. When moist air comes in contact with it, a very viscuous. liquid is produced which verj^ quickly prevents the further passage of gases through it. U-tubes are therefore filled with alternate layers of glass wool and the drying material. The phosphorus pentoxide must not contain any of the trioxide. If the latter is present, the material nmst be distilled in a stream of oxygen. 45. Making Impervious Connections. — Either cork or rubber stoppers may be used in closing U-tubes after being filled, unless they are provided with glass stoppers. Both cork and rubber are by no means impervious to moisture and carbon dioxide. The ex- posed surface must therefore be coated with sealing wax, paraffine, or shellac varnish. If the latter is used, a sufficient time should be allowed for the escape of the alcohol before weighing the tube. When the tubes are connected, short pieces of rubber tubing should be used, but the glass parts of the apparatus should be brought into contact, and the rubber tubing wired or tied to the glass tube. The portion of the rubber tube between the wire or string should be shellacked. Only in this manner can an abso- lutely tight and impervious joint be secured. When weighed U-tubes are disconnected from the other tubes, the ends should be closed with short pieces of rubber tubing closed at one end by glass plugs. These stoppers should be removed while weighing the tubes. EXERCISE 9. Determination of Water of Crystallization in Magnesium Sulphate, MgSO,.7H30. Fill three Marshand tubes with calcium chloride as directed in the pre- ceding section. Shellac the exposed parts of the rubber or cork stoppers. Connect these tubes with a hard-glass bulb as shown in Fig. 9, and support the series so that the bulb-tube may be heated by means of a Bunsen burner. A Bunsen filter-pump or a large bottle filled with water having an exit at the bottom for the water, or a siphon extending to tlie bottom of the bottle and a second tube passing through the stopper at the top for entrance of air, should be provide'd to draw air through the apparatus. A wash-bottle containing concentrated sulphuric acid should be inserted between the last U-tube and the aspirator. All joints made by rubber tubing should have the glass tubes in contact and the rubber tubes should be wired or tied to the glass and the part between should be shellacked. After the 48 QUANTITATIVE ANALYSIS. apparatus has been set up the first calcium-chloride tube should be closed by a rubber tube and glass plug. Gentle suction should be applied so Fig. 9. that on disconnecting with the aspirator, the acid in the wash-bottle shall rise in the long glass tube, showing a partial vacuum in the apparajjius. The height of this acid should be carefully noted, and if an appreciable change occurs in five minutes, a leak exists in some of the joints, which must be found and closed before proceeding with the analysis. When the apparatus has been made air-tight, light the Bunsen burner and heat the bulb-tube to redness and draw air slowly through the apparatus for about one-half hour. The U-tubes to the right of the bulb-tube are now discon- nected and weighed after being carefully cleaned. In the meantime, the bulb-tube must be closed with a rubber stopper. The magnesium sulphate should be weighed out in a glass tube of about 5 mm. internal diameter and about 8 cm. long, which has been closed at one end. This tube is weighed, about one-half gram of the salt placed in it, and again carefully weighed. Another piece of glass tubing of the same diameter and of sufficient length is taken, and the weighing-tube containing the salt fastened to it by means of a short piece of rubber tubing of suitable size. By means of this handle, the weighed tube is introduced into the bulb, and by tapping the glass-tube the salt is deposited in the bulb, as shown iq Fig. 10. The tube is carefully withdrawn so as not to drop any Fig. 10. adhering particles of salt, the weighed tube carefully detached from the rubber connector and again accurately weighed with whatever magnesium sulphate is left adhering to it. The difference between the last two weights gives the weight of magnesium sulphate deposited in the bulb-tube. The DETERMINATION OF WATER. 49 weighed U-tubes are connected in position again and the apparatus once more tested for leal^s. If none are found, the bulb-tube is gently heated by means of the Bunsen burner, while a slow current of air is drawn through the tubes. If the U-tubes become appreciably warm from the heat of the Bunsen burners, they must be shielded from the heat by means of asbestos boards or other non-conducting material. Finally, when the moisture which condenses in the bulb has been removed by the current of air, the bulb is heated to dull redness for about fifteen minutes. The burner is then removed, the weighed U-tubes disconnected and weighed again. As soon as the U-tube is removed, the bulb-tube is closed with a rubber stopper. It is then replaced and the heating continued for about half an hour and the U-tubes again weighed. This process is repeated until the U-tubes no longer increase in weight. The percentage of water is calculated from the weight of magnesium sulphate taken and the increase in weight of the U-tubes. DETERMINATION OF METALS. CHAPTER IV. DETERMINATION OF METALS AS OXIDE. 46. Properties of Weighable Precipitates. — In Chapter II mention was made of some of the properties of a salt or other combination of an element which determine its availability for the separation of that element from the solution. The discussion was limited to the requirements of complete precipitation and thorough wash- ing. In other words, all of the element to be determined must be present in the precipitate, and all other substances must be removed from the precipitate. It is then in condition to be weighed as such. Generally the element forms part of a more' or less complex chemical compoimd — oxides, chlorides, and sul- phates being common forms in which elements are weighed. This compound must be of depiite composition, so that the percentage of the element to be determined is fixed and invariable. The weighed substance must be non-volatile, not strongly hygroscopic, and in general unacted on by the atmosphere or the fumes liable to be present in the laboratory. Other things being equal, a precipitate of simple composition is apt to be of constant compo- sition, and therefore suitable for weighing. 47. Properties of the Metallic Oxides. — The oxides of the metals have been found in many cases to possess all the properties essential to a weighable compound. The affinity of most ele- ments for oxygen is so strong that even on ignition this element is not expelled. The oxygen of the air will even reoxidize most metals if they have become reduced by carbon or reducing-gases. The almost universal solvent in which precipitations are made is water which is composed of oxygen to the extent of about 89%. If a metal is to be weighed as an oxide, it may be precipitated as 50 DETERMINATION OF METALS AS OXIDE. 51 HYDROXIDE, CARBONATE, or OXALATE. In tUe case of many metals the salts of volatile acids may be ignited, leaving the metal as ■oxide. This method is especially applicable to nitrates and organic salts. In some cases the salts of non-volatile acids may be converted into oxides by heating with ammonium carbonate or mercuric oxide, the acid being driven off in combination as the volatile mercury or ammonium salt. These various methods win now be studied in detail, the metals being considered in groups in which the method of analysis is nearly identical for each group. IRON, ALUMINIUM, AND CHROMIUM GROUP. METAL PRECIPITATED AS HYDROXIDE BY AMMONIUM HYDROXIDE AND WEIGHED AS OXIDE. 48. Precipitation of the Hydroxides of Iron, Aluminium, and Chromium. — The hydroxides of these three metals are almost entirely insoluble in water, either hot or cold. They differ mark- edly in their solubilities in alkali. Iron is almost absolutely insoluble in ammonia or the fixed alkalies, while aluminium is easily soluble in the fixed alkalies and slightly so in ammonia. Chromium occupies an intermediate position in this respect between iron and aluminium, being almost insoluble in ammonia and moderately soluble in the fixed alkalies. Iron may, therefore, be completely precipitated by sodium, potassium, or ammonium hydroxide. It is found, however, that ferric hydrate carries down some of the potassium or sodium which cannot be washed out. Ammonium hydroxide is, there- fore, always used in precipitating iron, which for this purpose must be in the ferric condition. Any ferrous iron may be converted into ferric by boiliiig fhe solution after the addition of a few drops of nitric acid or bromine-water. As chromium and aluminium are soluble in sodium hydroxide, these metals are also precipitated by ammonium hydroxide. As aluminium hydroxide is soluble in even a shght excess of ammo- nia, the solution must he made as nearly neutral as possible. Ammonium chloride greatly weakens the alkaline properties of ammonia, and for this reason a considerable amount of this salt 52 QUANTITATIVE ANALYSIS. i's added to the solution of aluminium. This is unnecessary in the case of chromium and iron, because of their insolubility in ammonium hydroxide. Chromium hydroxide dissolves slightly in ammonium hydrox- ide in the cold, but is completely precipitated on boiling. Chro- mium must be in the chromic or basic condition. Chromates are reduced to chromic compounds by adding sodium sulphite to the solution acidified with hydrochloric acid, or, still better, by passing in gaseous sulphur dioxide. 49. Washing. — The hydroxides of these three metals are gelat- inous, and therefore there is no danger of their passing through the paper. They clog the pores, however, so that rapid filtration and washing can be accomplished only by using suction or a hot- water funnel. The washing must be thorough, as any ammo- nium chloride left in the precipitate reacts with the hydroxides, liberating ammonia and forming a chloride of the metal which volatilizes so that the result obtained is low. When these hydrox- ides dry they shrivel and crack so that if the washing is discon- tinued for a short time fissures are produced in the precipitate, through which the water passes without taking the impurities out of the precipitate. If the washing must be discontinued, the stem of the funnel should be closed with a piece of rubber tubing and a glass plug and the funnel filled with water. Many of these difficulties may be overcome by using paper pulp, as described on page 27. 50. Ignition of the Filter-paper.— When the filter-paper is burned neither the aluminium nor the chromium hydroxides are reduced by the hot carbon. Ferric hydroxide, however, is partly reduced to the magnetic oxide of iron, Feg04, while the remainder is converted into ferric oxide, Fefi^. As much of the precipitate as possible should therefore be detached from the paper, so that the error from this source may be negligible. If the filter-paper, with whatever ferric hydroxide cannot be removed, is burned in the platinum crucible, the iron which is reduced is apt to alloy with the platinum and be difficult of removal. It should there- fore, be burned in the flame of the Bunsen burner while held by a platinum wire. If a more open platinum dish, such as is used in milk analysis, is at hand, the moist precipitate may be DETERMINATION OF METAL AS OXIDE. 53 placed in the dish with the apex of the filter-paper on top. The side of the dish is now heated so that the paper dries and finally burns. The access of the air to the paper is so good that little iron is reduced. SI. Contamination of the Precipitate with Silica. — It is very difficult to obtain precipitates of these three metals which are free from silica. The ammonia which is used as a precipitant almost invariably contains silica which has been derived from the reagent bottle, most of which is precipitated and can be filtered off. Appreciable amounts of silica are obtained by the disinte- gration of the glass of the beaker in which the metal is precipi- tated. This silica is filtered off and weighed with the precipitate. The digestion of the alkaline solution should, therefore, he as brief as possible, and the excess of ammonia as small as possible. If too much ammonia is added it should be neutralized with dilute hydrochloric acid. It is also better to conduct the precipitation in a porcelain dish, since porcelain is less easily attacked by alkalies than glass. If a platinum dish is available, this source of error is entirely avoided. If glass is used the error from this source may be reduced, by working rapidly, to almost negligible quantities of .1 or .2%. If Jena beakers are used the error is still less. If the highest degree of accuracy is desired the percentage of silica must be determined and a correction made on the weight of the original precipitate. This determination is carried out by dissolving the precipitate in acid, evaporating to dryness on the water-bath, redissohdng in acid and water, and filtering off the silica, which is washed, ignited, and weighed. For this purpose the iron precipitate may be dissolved by digesting the powdered material with hydrochloric acid. Aluminium oxide may be dis- solved in the same manner, but the action is very slow unless paper pulp is used during the precipitation, as directed on page 27. It dis- solves more easily if first heated with eight parts of concentrated sulphuric acid and three parts of water, and then water added. It is also rendered soluble by fusion with potassium bisulphate. This method is especially to be recommended for both iron and aluminium, because the silica may be rendered insoluble by dissolving the melt in dilute sulphuric acid and evaporating until fumes of sulphuric acid are evolved. Chromium oxide is best dissolved by fusion in the platinum crucible with sodium carbo- 54 QUANTITATIVE ANALYSIS. nate and potassium nitrate, potassium chromate being formed. This solution must be acidified with hydrochloric acid and evap- orated to dryness and the silica determined as in the other oxides. 52. Removal of Organic Material. — These metals are not completely precipitated if organic matter is present in the solu- tion. It may be detected and also removed by evaporating the solution to dryness and igniting. This is not applicable if chlo- rides are present. In that case boiling with concentrated nitric acid or potassium chlorate and hydrochloric acid must be resorted to. The solution may also be treated with excess of sulphuric acid and evaporated to dryness and ignited. If the amount of organic matter is small, the ignition will frequently be unnecessary, as the charred material will be entirely destroj-ed by the concentrated sulphuric acid produced by the evaporation. A few drops of nitric acid followed by heating the solution will render it clear more quickly. EXERCISE 10. Determination of Aluminium in Potash Alum, K,S0,.AL(S0,)3.24H,0. Weigh carefully 2 grams of pure recrystallized potash alum. If bal- anced watch-crystals are used, place them on the scale-pans and ascertain if the pointer indicates zero when the beam and pans are released. If it does not, balance the watch-crystals by means of a small weight or the rider. Place a 2-gram weight on the right-hand watch-crystal and weigh out 2 grams of the alum as directed in Exercise 4. Transfer to a 500-c c. be'aker, brushing off the watch-crystal with a clean camel's-hair brush, and dissolve in 200 c.c. of water. Add about 50 c.c. of ammonium chloride solution, which should be filtered if not perfectly clear. The ammonium chloride may also be introduced together with paper pulp by adding 15 c.c- of concentrated hydrochloric acid to two or three ashless filter-papers * and 25 c.c. of dilute ammonia. Warm on a hot plate or an asbestos board or wire gauze over a Bunsen burner. Add filtered dilute ammonia with con- stant and \'igorous stirring of the solution until the odor of ammonia can be detected after vigorous stirring. Red litmus-paper should be only slowly turned blue when held over the solution. If too much ammonia has been added it should be nearly neutralized with dilute hydrochloric acid. * The weight of the ash of this paper is stated on the package obtained from the dealers in chemical apparatus and is usually small enough to be neglected. The paper should be kept in a box or other receptacle where it can be kept per- fectly clean. DETERMINATION OF METAL AS OXIDE. 55 53- Washing and Transferring the Precipitate to the Filter-paper.— After the solution has thus been made faintly alkaline, heat it nearly to the boiling-point for a few minutes. Fit a funnel with an ashless filter-paper preparatory to filtering by suction as directed in Chapter II. An 11 or i2i cm. filter-paper should be used and the funnel should be of such a size that a space at least a quarter of an mch is left free above the paper. Allow the precipitate to settle a few minutes and then decant the clear hquid through the paper, using suction. Guide the stream of water by holding the stirring-rod against the lip of the beaker. Add about 75 c.c. of hot water from the wash-bottle,* which for this purpose is kept on a steam-bath or hot plate. Stir the precipitate well with the stream of water. For this purpose the jet of water must be quite large and care must be taken not to spatter the precipitate out of the beaker. Allow the precipitate to settle and decant through the funnel as before. Repeat the washing by decantation twice a.iJ then transfer the main precipitate to the funnel. Holding the stirring-rod against the beaker with the left hand, sweep the precipitate from the beaker by means of a stream of water from the wash- bottle, which is held in the right hand as shown in Fig. 11. When the Fig. 11. bulk of the precipitate has been transferred in this manner, the particles adhering firmly to the glass must be removed by means of a so-called police- man, which is made by inserting the end of a rather large-sized stirring-rod * A so-called hot-water wash-bottle is prepared by fastening asbestos paper, a sheet of cork, or other insulating material arouncl the neck of an oi-diiiary wash- bottle by means of pieces of wire or string. A thick cord closely wound around- the neck of the flask will serve the same purpose. The water should not be allowed to boil or steam will enter the mouth when blowing out the water. To overcome this difficulty, a piece of rubber tubing should be placed on the glass tube which enters the mouth. By closing this tube with the teeth when not blowing, steam is kept out of the mouth. 56 QUAXTITATIVE ANALYSIS.. into a short piece of rubber tubing. The rubber tube should be left pro- jecting slightly beyond the end of the glass tube and sealed together with a little bicycle cement. The bottom as well as the sides of the beaker must be rubbed clean with this policeman. Both the stirring-rod and the policeman must finally be freed from particles of the precipitate. If portions of the precipitate cannot be rubbed off the glass they should be dissolved in a few drops of dilute hydrochloric acid and reprecipitated by neutralizing with ammonia and warming the solution. When all of the precipitate has been transferred to the funnel, it is washed with hot distilled water until free from chlorides. The exposed portion of the filter-paper must also be well washed. Test the wash-water from time to time with nitric acid and silver nitrate, finally using sevej-al cubic centimeters. In washing this precipitate, care must be taken that all of the water is not sucked off the precipitate so that it cracks, producing fissures through which the wash-water passes without dissolving the ammoniima chloride. If for any reason the washing must be discontinued, the stem of the funnel should be closed with a short piece of rubber tubing and a glass plug and the funnel filled with water. 54. Ignition and Weighing of the Precipitate. — During the washing of the precipitate a platinum crucible with the lid is supported on a clean pipe- stem or platinum triangle resting on a tripod and heated to redness ^^'ith a Bunsen burner. The air-vents of the Bunsen burner are adjusted so that the flame is entirely colorless. The crucible must be kept well above the inner cone. While still red-hot the crucible is seized -mth a clean pair of forceps or crucible tongs and placed in a desiccator, which, for this purpose, is provided A'sdth a pipe-stem or glass triangle. After fifteen minutes it is weighed, the weight being recorded on the right-hand side of the note- book, the page being fully labelled and dated. The left-hand page is reserved for a brief description of the process of analysis, and especially any procedure which led to disaster or failure and the remedy adopted. All of this work should be done during the intervals when the liquid is running through the funnel. It will run just as fast if not watched. The moist precipitate \vith the paper is taken out of the funnel and transferred to the weighed platinum crucible. If there is any delay, so that the precipitate is partially dried, it must be placed in the steam-oven and completely dried. It then shrinks to a very small bulk. The lid of the crucible must in this case be very carefully placed on the crucible, as a little moisture left in the hard lumps of the precipitate forms steam and scatters the precipitate with considerable force. When burning the moist precipitate, the paper is placed in the crucible with the apex up, so that spattering during drying will not throw the precipitate out of the crucible. A\iv portions of the precipitate left on the funnel are wiped off with small pieces of cjuantitative filter-paper and placed in the crucible. The co\'ered crucible is then placed on the triangle and heated \A-ith a very small flame DETERMINATION OF METAL AS OXIDE. 57 until steam and volatile matter cease to escape. The heat is gradually increased until the water and most of the volatile matter are expelled from the precipitate and the paper. The lid is now taken off the crucible, which is placed on its side and heated with the full flame of the Bunsen burner. As the carbon burns, the crucible is turned to expose fresh portions to the action of the air. The crucible is finally heated with the blast-lamp until the aluminium oxide is perfectly -white. While still red-hot, the crucible is transferred to the desiccator and after fifteen minutes, weighed. The crucible is again heated in the blast-lamp, cooled in the desiccator, and weighed. This is repeated until the weight is constant -v^ithin .2 or .3 milligram. The percentage of ALOj found is calculated by dividing the weight of the precipitate by the weight of alum taken and multiplying the result by 100. The theoretical percentage of Al^Oj in potash alum is 10.76. If a result much higher than this is obtained, the precipitate may be contami- nated with silica, which is determined as directed in the next exercise, with the exception that grinding the precipitate will be unnecessary if paper pulp has been used during the precipitation. EXERCISE II. Determination of Iron in Soft-iron Wire.* In this exercise the determination should be carried out in duplicate, the two analyses being carried on side by side. Clean the mre thoroughly bj' rubbing with emerj^-cloth or sandpaper until all rust is removed and finally wiping vnih. filter-paper. The wire should be wound in a spiral around a lead-pencil and should not be touched with the fingers after being wiped with the filter-paper. Weigh out from 200 to 300 milligrams of the wire and transfer to a 200-c.c. beaker. Add 5 c.c. of concentrated nitric acid and a httle water. Cover the beaker with a watch-crystal and warm on the hot plate under the hood. When the iron is all dissolved, rinse off the watch-crystal with a little distilled water and evaporate to dryness on the hot plate. Add 10 c.c. of dilute hydro- chloric acid and again evaporate to dryness on the water-bath. After the ferric chloride is dry,heat for one-half hour on the water-bath, dissolve :n a few cubic centimeters of dilute hydrochloric acid, add a little water, and filter off the silica on a small filter-paper. Wash the paper free from iron with hot distilled water to wliich a little hydrochloric acid has been added. The volume of the filtrate should be about 200 c.c. It is best to precipitate the iron in a platinum dish. If such a dish is not available, an evaporating-dish of Berlin porcelain should be used or a * This wire -is such as is used in standardizing potassium permanganate or other oxidizing solutions and is sold for use as piano-strings, and to florists for making wreaths. 58 DETERMINATION OF METALS. Jena beaker. The solution is neutralized with filtered ammonia, and brought to a boil. After allowing the precipitate to settle for a few minutes, the clear liquid is decanted through a 12J or 15-cm. ashless filter-paper arranged for suction as described in Chapter II. The precipitate is washed twice by decantation and then transferred to the funnel and washed with hot water, observing the precautions given in Exercise 10. Toward the end of the washing the stream of water from, the wash-bottle should be so directed as to loosen the precipitate from the sides of the paper and wash it down into the point of the funnel. When the precipitate has been washed free jrom chlorides it is dried in the steam- or air-oven. The tem- perature of the latter should not be allowed to rise much above 100°Centi- grade. 55. Ignition of the Paper on Platinum Wire. — A piece of glazed paper, preferably of a light color, about 8X10 inches is spread out on the desk. A small camel' s-hair brush, a piece of platinum wire about 4 inches long fastened to a glass or other handle, a Bunsen burner with a long blue flame, the ignited and weighed platinum crucible and the dried precipitate in the funnel are brought together. The glazed paper is carefully brushed clean with the camel's-hair brush. The crucible is placed in the center of the paper. The paper is taken out of the funnel and the precipitate loosened with the platinum wire and dropped into the crucible. Very little of the precipitate should be left on the paper. Rubbing the two inner surfaces of the filter-paper together will sometimes loosen a considerable amount of the precipitate. If any portion of the precipitate is found adhering to the funnel it should be rubbed off with a small piece of ashless filter-paper which has been moistened with a little dilute nitric acid. The filter-paper is now folded as if to be inserted into a funnel. It is rolled up tightly on one of its edges and the platinum wire is ^vrapped around the upper part of the paper which is always freest from portions of the precipitate. It is now held by means of the handle of the platinum wire over the crucible and the point touched to the Bunsen burner flame and ignited. The charred paper is completely burned by holding it in the oxidizing flame of the Bunsen burner and then exposing the red-hot carbon to the oxygen of the air. The ash is allowed to drop into the crucible. The wire is now brushed off, the crucible is taken in the left hand and any particles which have fallen on the paper are brushed into it. The crucible is placed on its side on a ]iipe-stem triangle and heated with the Bunsen burner until any unburned carbon of the filter-paper is oxidized. If a platinum dish is available the moist precipitate with the apex up may be placed in the dish which is gently heated on one side with the Bunsen burner. When the paper is completely charred the dish may be heated to redness until the carbon is completely burned. The precipitate is then heated with the blast lamp for a few minutes, cooled in the desiccator n,nd weighed. It is re-ignited and weighed until the weight is constant. DETERMINATION OF METALS AS OXIDE. 59 56. Calculation. — The weight of Ye^O, multiplied by tlie percentage of iron in Fefi^ (70 per cent) gives the weight of iron found. This number multiplied by 100 and divided by the weight of iron taken gives tlie percentage of iron in the wire which is usually 99.6 to 99.8 per cent. The result may also be computed by logarithms. The log. of the weight of the precipitate is added to the log. of the percentage of iron in Ye.O,. From this sum is subtracted the log. of the weight of iron wire taken. This log. + 2.0 gives the log. of the percentage of iron in the wire. The dupli- cates in this analysis should agree within .1 to .2 per cent. If they vary more than this amount a third or even a fourth analysis must be made. 57. Determination of Silica. — If the percentage found is greater than 99.8 per cent, the error is probably due to silica, especially if the precipita- tion was carried out in a beaker. Instead of making another determination, the percentage of silica should be determined and a correction made on the percentage found. For this purpose the precipitate is ground quite fine in an agate mortar. Loss by spattering should be avoided by covering the mortar with a piece of paper through which the pestle passes. When as much of the precipitate as can be readily removed from the crucible has been ground, the pulverized material is returned to the crucible which is then weighed. The difference between this weight and the weight of the crucible gives the weight of the iron oxide taken for the silica determination. The loose material is transferred to a 100-c.c. beaker, 15 c.c. concentrated hydro- chloric acid and 10 c.c. water added. The beaker is digested on the hot plate until the iron is dissolved. A little acid is also added to the crucible which is also heated until the iron is dissolved. This solution is added to that in the beaker. The beaker is now placed on the water-bath and the solution evaporated to dryness and heated dry for one-half hour. Dilute hydrochloric acid is added and the heating continued until the iron is again dissolved. The powdered precipitate, after being weighed in the crucible, may also be fused with acid potassium sulphate. The burner should be turned down until the heat is just sufficient to keep the material fused. When the iron is all dissolved the crucible is allowed to cool, transferred to a beaker and the melt extracted with dilute sulphuric acid and the solution evaporated until fumes of sulphuric acid are evolved. The solution obtained by either method is filtered through an ashless filter-paper, the silica m the beaker being transferred to the paper by means of a policeman as described in Exercise 10. The silica is washed a-id, after burning the paper, is ignited, finally for a few minutes v.ith the blast-lamp. The crucible is cooleid in the desiccator and weighed. Tlie contents of the crucible should be pure white, any tinge of red indi- cating iron due to imperfect washing or undissolved ferric oxide. On adding a little pure hydrofluoric acid the precipitate should be entirely volatilized on heating. Any residue which, on repeated treatment with 60 DETERMINATION OF METALS. hydrofluoric acid, fails to volatilize is not silica. As only part of tlie original iron precipitate was taken for this analysis the amount of siUca in the entire precipitate must be calculated and deducted to give the true weight of FcjOj from which the percentage of iron in the iron wire must be re-calculated, as in the following illustration. 0.3350 gram of the iron wire was weighed out. 0.4820 gram of the ferric oxide was obtained. The percentage of iron calculated from this weight of oxide by the method already given is 100.7. After grinding the precipitate for the silica deter- mination, 0.4153 gram remained in which .0043 gram Si02 was found. The amount present in the entire precipitate was obtained from the pro- portion 0.4135 : .0043 :: 0.482 : x where a; = .0050. On subtracting the silica from the weight of the impure precipitate we obtain 0.4770 as the weight of ferric oxide. On recalculating the percentage of iron we obtain 99.68. DETERMINATION OF GOFFER, MANGANESE, NICKEL, AND COBALT BY PRECIPITATION WITH CAUSTIC ALKALI. 58. Precipitation and Washing. — A number of metals cannot be precipitated as hydroxide with ammonia, because of their solubiUty in ammonium salts. Some of these metals form hydrox- ides which are insoluble in sodium and potassium hydroxides, and may be precipitated by either of these reagents. Consider- able difficulty is experienced in completely washing out the excess of alkali from the precipitate and, unlike ammonia, it is not expelled by ignition. On digesting the ignited precipitate with water the alkali can frequently be detected by its alkaline reac- tion. By filtering, washing, and reigniting, a considerable loss in weight is frequently noted. The solution should be dilute and only a slight excess of alkali added. The liability of contamination of the precipitate with silica from the action of the alkali on the glass, which was met with in the determination of iron, aluminium, and chromium is still more pronoimced when thf. stronger alkalies are used. The precipitation should therefore never be carried out in glass. Por- celain or, wherever possible, platinum dishes should be used. The precipitate should be tested for the presence of silica by solu- tion in hydrochloric acid, evaporation to dryness, again dissolv- DETERMINATION OF METALS AS OXIDE. 61 ing in acid, filtering, washing, and weighing the residue as directed under iron. These oxides dissolve with ease in acid. 59. Composition of the Ignited Oxides. — Considerable diffi- culty is experienced in obtaining precipitates of constant compo- sition. CuPRic HYDROXIDE is Converted on ignition into cupric oxide, CuO, but if reduced by the filter-paper or reducing gases from the flame to metallic copper it reoxidizes to cuprous oxide, CUjO. This is converted into cupric oxide by moistening with nitric acid, evaporating to dryness and igniting. ^Manganese hydroxide is oxidized by the oxygen of the air even at the ordinary temperature. It is not reduced either by the carbon of the filter-paper or by the reducing gases of the flame. On ignition in presence of air it is converted into man- ganoso-manganic oxide, MujOi. It is impossible, however, to obtain a precipitate in which the ' percentage of manganese cor- responds to the theoretical for Mn30^. Pickermg has shown that the amount of manganese in the ignited oxide varies from 69.688 to 74.997%, according to the temperature to which the oxide has been heated and other undetermined conditions. For this reason only small amounts of manganese should be weighed as MnjO,. When coBALTOus hydoxide is heated in contact with the solution from which it was precipitated, it is gradually converted by the oxygen of the air into cobaltoso-cobaltic hydroxide. On drying, it absorbs more oxygen, but on strong ignition oxygen is driven off, leaving cobaltous oxide OoO. On cooling in contact with the air, oxygen is again absorbed, the light brown cobaltous oxide changing more or less completely to the black cobaltoso- cobaltic oxide C03O4. If the air is excluded by a current of car- bon dioxide this action is prevented, and the pure cobaltous oxide is obtained. This oxide is readily reduced to metalHc cobalt by ignition in a stream of hydrogen and the alkali which could not be completely removed by washing the hydroxide can now be quite readily extracted with hot water. NiCKELOUS hydroxide is converted on ignition into nickelous oxide NiO, which is unaltered on heating in the air. It is readily reduced to metallic nickel on ignition in a stream of hydrogen or of carbon monoxide. 62 DETERMINATION OF METALS. EXERCISE 12. Determination of Copper in Crystallized Copper Sulphate, CuSO,.sHjO. Weigh out 1 gram of the pure salt, transfer to a platinum or porcelain dish, and dissolve in about 200 c.c. of distilled water. Make a dilute solution of caustic soda by dissolving about 1 gram in 25 c.c. distilled water. Heat the copper solution nearly to boiling and add the caustic-soda solution with constant stirring until no further precipitation occurs. Allow to settle for a few minutes, then decant through a paper fitted into a funnel for suction. Wash the precipitate twice by decantation with hot water, then transfer to the funnel and complete the washing with hot water until the wash-water is perfectly neutral. Dry the precipitate in the funnel at 100° C. Remove the precipitate from the paper as directed in Exercise 11 and place it on a watch-crystal. Burn the paper on the platinum wire,, being careful to wrap the wire around a portion of the paper on which there is the least amount of precipitate. Let the ash fall into a porcelain cruci- ble. Add a few drops of concentrated nitric acid, place the lid on the crucible, and evaporate off the excess of acid with a very small flame of the Bunsen burner, finally heating quite strongly. If any carbon from the filter-paper remains unburned, repeat the treatment with nitric acid. When the crucible is cold set it on the glazed paper and transfer the main portion of the precipitate to the crucible, brushing the watch-crystal and paper thoroughly. The moist paper with the precipitate may also be placed in the crucible and the paper burned as directed in Exercise 10. The precipitate must be moistened with nitric acid as directed for the ash of the filter-paper. Finally heat the crucible with the full flame of the Bunsen burner for ten or fifteen minutes, cool in the desiccator, and weigh. As copper oxide is hygroscopic, the weight should be taken rapidly. The crucible is reheated and when weighed the second time the weights should be placed on the right-hand pan before the crucible is placed on the left-hand pan so as to make the final adjustment with the rider or by taking the swings of the pointer before the precipitate gains appreciably in weight. The percentage of copper found in copper sulphate should be within .1 or .2 per cent of the theoretical 25.46 per cent. The method gives high results. EXERCISE 13. Determination of ■ Nickel in Nickel-ammonium Sulphate, (NHJjSO^.NiSO^.eHjO. Weigh out 2 grams of crystallized nickel-ammonium sulphate and trans- fer to a porcelain dish. Dissolve in about 200 c.c. distilled water. Heat nearly to boihng and add, vnth stirring, a dilute but freshly made solution of caustic soda until precipitation is complete. Continue heating the solu- tion for ten or fifteen minutes. Wash by decantation two or three times, then wash thoroughlv with hot water on the filter-paper. Dry the pre- DETERMINATION OF METALS AS OXIDE. 63 cipitate, detach from the paper, and place it on a watch-crystal. Burn the filter-paper over the porcelain crucible. Treat the residue with a drop of concentrated nitric acid, evaporating off the excess. Add the remainder of the precipitate and heat with the Bunsen burner to constant weight. The precipitate should now be examined for alkali by digesting with hot water and testing with litmus. If alkali is present it is removed by repeated extraction with hot water, the precipitate being finally heated to constant weight again. Calculate the percentage of nickel found in the nickel-ammonium sulphate. Theoretical percentage is 14.86. The nickel oxide may also be converted into metallic nickel by heating m a stream of hydrogen. For this purpose the perforated lid and porce- lain stem cf the Rose crucible are most convenient. In place of this an ordinary clay pipe may be used. The hydrogen from a Kipp generator should be freed from arsine by passing it through a mercuric chloride solu- tion and dried by passing through calcium chloride or concentrated sul- phuric acid. The apparatus is set up as shown in Fig. 12. The stream of hydrogen should be passed for a few minutes to displace the air from the wash-bottle and the U-tube. An explosion will not pass through the Fig. 12. apparatus if a plug of cotton or glass wool is piaced in the glass tube or limb of the U-tube nearest the crucible. The crucible is then gently heated with a small flame from a Bunsen burner, which is turned on full after a few minutes. The crucible is finally allowed to become nearly cold in the stream of hydrogen. It is then transferred to the desiccator for a few minutes before weighing. The ignition in the stream of hydrogen is repeated until the weight is constant. The precipitate is then tested for silica by dissolving the metal in a little dilute nitric acid. If any silica is found, it is filtered off, washed, and weighed. CHAPTER V. DETERMINATION OF METALS AS OXIDE. PRECIPITATION OF LEAD, BISMUTH, CALCIUM, BARIUM, AND STRONTIUM BY AMMONIUM CARBONATE. 60. Precipitation by Ammonium Carbonate. — Some of those metals whose oxides or hydroxides are soluble in water or in solutions of the alkali hydroxides may be precipitated by ammo- nium or sodium carbonate. Ammonium carbonate cannot be used with some of the metals because the precipitate dissolves, forming a double salt with the ammonium compounds produced by the neutralization of the mineral acid with which the base was combined. Ammonium carbonate is used in preference to sodium or potassium carbonate, because of the greater ease of freeing the precipitate from the ammonium salts by washing or volatilization during ignition. On the other hand, some carbon- ates of the metals which can be precipitated by ammonium car- bonate form acid carbonates which are slightly soluble. Ammo- nium carbonate as obtained either in the solid form or in solu- tion almost invariably contains acid ammonium carbonate. Free ammonia must therefore always be added to the solution to be precipitated. The precipitation is therefore usually said to be made by ammonia and ammonium carbonate, by the latter term being designated the commercial article. 61. Conversion of the Carbonate into the Oxide. — The ease of converting the carbonates of these metals into oxides varies greatly. Lead and bismuth carbonates lose all of their carbon dioxide below red heat; indeed at the ordinary temperature part of it passes off so that the precipitates formed by ammonia and ammonium carbonate are basic carbonates. In the case of the alkaline-earth metals the precipitate formed in each case is the normal carbonate, and it may be dried and weighed as such. Calcium carbonate may be completely converted by the heat of the blast-lamp into calcium oxide, so that this metal may be 64 DETERMINATION OF METALS AS OXIDE. 65 weighed either as the oxide or as the carbonate. It is impossible to completely convert strontium or barium carbonates into oxides. These metals are, therefore, weighed as carbonates.* As LEAD SALTS are very easily reduced to the metallic form, the bulk of the lead precipitate must be removed from the filter- paper before the latter is burned. The lead which remains on the paper is reduced to metal, and must be reconverted to oxide by treatment with concentrated nitric acid and ignition of the lead nitrate formed. The bulk of the lead precipitate is then placed in the crucible and by ignition converted into the monox- ide of lead. Bismuth precipitates are treated in the same manner as lead precipitates. If it is desired to weigh the calcium as oxide, the precipitate need not be dried or removed from the filter-paper, but is intro- duced into the crucible, dried by gentle heat, ignited with access of air to burn the paper, and finally heated with the blast-lamp to constant weight. This is, therefore, the simplest and most rapid method of weighing calcium. It has the disadvantage, however, that the calcium oxide is very hygroscopic, and when moist absorbs carbon dioxide from the air. The precipitate must, therefore, be weighed as quickly as possible. Calciuji carbonate is stable in the air and may, therefore, be weighed with the greatest accuracy. If the precipitate is to be weighed as carbonate, it must be dried and removed from the paper. This is not difficult, as it forms a loose powder on dry- ing. It is placed on a watch-crystal, while the paper is burned in the crucible or on a platinum wire. By this treatment the cal- cium remaining on the paper is converted into oxide, which must be changed to carbonate. This is accomplished by sprinkling some dry powdered ammonium carbonate on the calcium oxide, moistening with a drop or two of water, and heating very gently to drive off the excess of ammonium carbonate and water. This operation is repeated, then the bulk of the precipitate is added, which is also heated gently with the Bunsen burner. After weighing the precipitate, it must be treated with ammonium ♦Calcium carbonate decomposes at 825°, strontium carbonate at 1155°, and barium carbonate at 1450°. 66 DETERMINATION OF METALS. carbonate and a little water and again heated and weighed, to ascertain if the weight is constant. The precipitate of barium or strontium may be thrown moist into the platinum crucible and heated with the Bunsen burner until dry and the paper burned. The hot carbon reacts with the carbonates of these metals, forming carbon monoxide and the oxide of the metal. After the complete combustion of the filter-paper, the precipitate should be treated with ammonium carbonate and water and heated to expel the excess of the reagent, finally to dull redness with the Bunsen burner. EXERCISE 14. Determination of Strontium in Strontium Carbonate, SrCOj. Weigh out 1 gram of pure strontium oarbonate. Transfer to a 400-c.c. beaker. Add 50 c.c. distilled water and dilute hydrochloric acid drop by drop with stirring or agitation of the solution until the strontium is dissolved. Dilute to 150 c.c, add a few cubic centimeters dilute filtered ammonia, then ammonium carbonate solution with constant stirring until no further pre- cipitation occurs. Allow the beaker to stand in a \^•arm place for several hours. Filter and wash the precipitate with water containing a little ammonia. Transfer the moist precipitate to a weighed platinum crucible. With the lid on the crucible, heat gently until the precipitate is dry. Remove the lid, place the crucible on its side, and heat strongly until the paper is burned. Cool, sprinkle some powdered ammonium carbonate over the precipitate, moisten with a little water, heat gently with the lid on the crucible until the precipitate is dry, then more strongly, but not to red- ness. Cool in a desiccator and weigh. Repeat until constant weight is obtained. The theoretical percentage of strontium in strontium carbonate is 59.37. PRECIPITATION OF CALCIUM BY AMMONIUM OXALATE. 62. Calcium is almost invariably precipitated as oxalate, since this salt of calcium is more insoluble than the carbonate, while on ignition it can be converted into the carbonate or the oxide as desired. It forms a very finely divided powder when first pre- cipitated, which gives considerable trouble by passing through the pores of the filter-paper. By digesting the precipitate with the hot solution from which it has been separated, the crystals increase in size so that after two or three hours it can be filtered without any difficulty. The same object is more quickly accom- phshed if the solution of calcium is Ijrought to a boil before adding DETERMINATION OF METALS AS OXIDE. 67 the ammonium oxalate, which should also be hot. The solution should also be vigorously stirred during the precipitation. If the precipitate is heated very gently, it decomposes almost completely, according to the equation CaC204 = CaC03 + CO. Almost invariably, however, the calcium carbonate will be dark- colored because of the formation of some free carbon. On heat- ing the precipitate strongly enough to cause the combustion of this carbon and thus to give it a pure-white color, a large portion of the precipitate will be converted into oxide. That portion which remains with the filter-paper will also be converted into oxide. Instead of attempting to produce the carbonate in the first instance, it is simpler to heat the precipitate with the paper from the beginning quite strongly with the Bunsen burner or the blast-lamp until it is perfectly white. It may then be brought to constant weight as calcium oxide, or it may be converted into carbonate by the addition of at least an equal bulk of ammonium carbonate, moistening with water and heating gently. This is repeated until constant weight is obtained. By weighing the calcium as oxide, the result is undoubtedly obtained more quickly, and if the weighing is done rapidly it is quite as accurate as when the calcium is weighed as carbonate. The weight of the calcium oxide may also be verified by converting it into sulphate. A few drops of sulphuric acid are added cautiously to the oxide. The excess is evaporated off on the hot plate and the calcium sulphate finally brought to constant weight by heating to red- ness with the Bunsen burner. EXERCISE 15. Determination of Calcium in Calcium Carbonate. Weigh out 0.5 gram of pure calcium carbonate. Transfer to a 500-c.c. beaker, add a little water, and while covering the beaker with a watch- crystal add 5 c.c. dilute hydrochloric acid. When the calcium is dissolved, rinse the watch-crystal and the sides of the beaker, bringing the bulk of the solution to about 250 c.c. Neutralize with filtered ammonia, heat nearly to boiling on the hot plate, and add ammonium oxalate solution slowly with vigorous stirring of the solution until no further precipitation occurs. Keep the solution on the hot plate, so that it nearly boils, for several hours. Decant the clear solution through the funnel, wash with hot water by decantation several times, then transfer the precipitate to the paper 68 DETERMINATION OF METALS. and complete the washing with hot water, testing the wash-water for chlo- rides. Transfer the moist precipitate to the weighed platinum crucible, dry, and burn the paper in the usual manner, finally heating with the blast- lamp for about ten minutes. Cool in a desiccator and weigh. Heat with the blast-lamp again for ten minutes and weigh, first placing the weight found necessary by the first weighing on the pan and finally taking the crucible out of the desiccator and making the final adjustment as rapidly as possible. Continue heating and weighing until constant weight is obtained. CoNVBHT the OXIDE into the carbonate by adding at least an equal bulk of powdered ammo- nium carbonate * and moistening with water. Place the cover on the crucible and evaporate off the water by gentle heat, then place the crucible on a pipe-stem triangle which is supported about 1 inch above a wire gauze which is heated by a Bunsen burner. Repeat this treatment, cool in the desiccator, and weigh. Continue this operation until IIM-i Vy constant weight is obtained. This weight should (-J--J ^> ^\ equal that of the carbonate weighed out for the analysis. The theoretical percentage of calcium oxide in calcium carbonate is 56. Fig. 13. DETERMINATION OF ZINC, MANGANESE, AND CADMIUM BY PRECIPITATION WITH SODIUM CARBONATE. The salts of these metals readily form double salts with ammo nium compounds. They cannot, therefore, be completely pre- cipitated by sodium carbonate in the presence of considerable amounts of ammonium salts. The latter are readily removed by evaporating the solution to dryness and volatilizing the ammo- nium compound by gently heating the dry residue. 63. Cadmimn Carbonate is almost insoluble in ammonium salts and completely insoluble in the fixed alkali carbonates and in water. On burning the filter-paper, the cadmium carbonate left upon it is reduced to metallic cadmium, which is very volatile at the temperature of the Bunsen burner. The precipitate should therefore be dried and removed as completely as possible from the paper. The loss of cadmium may be almost completely pre- vented by saturating the paper with a strong solution of ammo- nium nitrate, drying and igniting. The oxygen from the nitrate serves to reoxidize any cadmium which may have been reduced. * If on volatilizing one or two grams of the ammonium carbonate a weighable residue is left, it must be purified by sublimation. DETERMINATION OF METALS AS OXIDE. 69 The paper during the ignition should also be held in the oxidizing zone of the Bunsen burner, for the same reason. The error may be still further reduced by first washing the paper with a little dilute nitric acid and water, and evaporating the solution to dryness in the weighed crucible. The paper may then be burned as directed previously. The precipitate is converted by ignition into oxide, which is not decomposed nor volatilized at white heat. If the precipitate is small it need not be dried, but may be imme- diately dissohed in nitric acid and the solution e^'aporated to dryness in the porcelain crucible. 64. Manganes3 Carbonate is quite soluble in ammonium salts. The recently precipitated carbonate is white, but on standing it becomes brown, being oxidized by the oxygen of the air. It is insoluble in solutions, of potassium or sodium carbonate. On ignition in the presence of air the carbonate is converted into protosesquioxide of manganese which approximates the formula Mn304. Pickering has shown that the percentage of manganese in the ignited oxide varies from 69.69 to 75, according to the temperature to which the oxide has been heated. The theoretical percentage of manganese in Mn304 is 72.05. Only small amounts of manganese should be weighed in this form if accurate results are desired. No manganese is lost on burning the filter-paper. If the precipitate is large and not removed from the paper some difficulty will be found in securing complete combustion of the paper. 63. Zinc is precipitated as the basic carbonate. Carbon diox- ide is, therefore, liberated during the reaction. The bicarbonate produced holds some of the zinc in solution, but this is completely precipitated on boiling. It is, therefore, best to precipitate the zinc from a boiling solution and keep it at this temperature for some time. If even a small amount of ammonium salts are pres- ent, the precipitation is not complete. As the boiling alkaline solution acts on glass quite energetically, it is best to use a porce- lain or platinum dish. The precipitate is washed with hot water. On ignition, the carbonate is converted into oxide. In the pres- ence of carbon or other reducing substances, both the oxide and carbonate are reduced on ignition to metallic zinc which volatilizes. The precipitate must, therefore, be treated in the same manner 70 DETERMINATION OF METALS. as the cadmium precipitate. The zinc oxide is stable when ignited in the air in the absence of reducing agents. EXERCISE i6. Determination of Zinc in Zinc-ammonium Sulphate, (NHi)2SOi.ZnSO,.6H20. Weigh out 2 grams of crystallized zinc-ammonium sulphate. Transfer to a platinum dish or crucible. Heat very gently with a small Bunsen burner flame or on the hot plate until the water is expelled. Keep the dish covered to prevent spattering. Increase the heat somewhat and volatihze the ammo- nium sulphate. When no more fumes come off, allow to cool and dissolve the zinc sulphate in water and a little hydrochloric acid if necessary. If a crucible has been used, transfer the solution to a platinum or porcelain dish, heat nearly to boiling and add with stirring a freshly made sodium carbonate solution until the zinc is entirely precipitated. Continue the heating for about fifteen minutes, allow to settle, wash several times by decantation, finally transfer to the funnel and complete the washing with hot water. Dry the precipitate, detach it from the paper and place on a watch-crystal. Replace the paper in the funnel and wash with 25 c.c. of water to which a little nitric acid has been added. Evaporate the wash-water to dryness in the weighed platinum crucible. In the meantime moisten the paper with a saturated solution of ammonium nitrate and dry in the steam- bath. Burn the paper on the platinum wire, being careful to touch it only with the oxidizing portion of a Bunsen-burner flame which is entirely free from a luminous zone. The precipitate is now transferred to the crucible and ignited with the Bunsen burner and finally with the blast-lamp. Theo- retical percentage of zinc in zinc-anlmonium sulphate is 16.28. IGNITION OF SALTS OF VOLATILE ACIDS. 66. Determination of the Metals by Ignition of the Salts of Volatile Acids. — As has already been observed, it is possible to ignite the salts of many metals and obtain oxides of definite com- position, from the weight of which the percentage of the metal present may be calculated. Calcium oxalate and carbonate, cadmium, manganese, zinc, and lead carbonates are examples of such salts which are obtained by precipitation. In many cases the nitrates and nitrites may be treated in this manner, but not the chlorides since they are quite volatile an.d the oxygen is not present to unite with the metal to form the oxide. In some cases the oxygen may be furnished and the chlorine driven off, as a volatile compound by igniting the salt with mer- DETERMINATION OF METALS AS OXIDE. 71 curie oxide, nitric acid, or ammonium carbonate or nitrate. Zinc chloride may be converted into oxide by treatment with excess of mercuric oxide or nitric acid. Magnesiima chloride may be converted into oxide by ignition with mercuric oxide, ammonium carbonate, or nitrate. In the absence of substances of this character, the acid with which the metal is combined must contain oxj^gen and be readily decomposed and volatilized by heat. The sulphates and phos- phates, though containing the necessary oxygen, are usually very stable, and the acid is with difficulty volatilized by heat. The metals whose salts of volatile oxygen acids may be decom- posed into oxides are lead, bismuth, copper, tin, iron, alumin- ium, . CHROMIUM, MANGANESE, NICKEL, ZINC, and MAGNESIUM. No other volatile substance except the metal may be present. 67. Ignition of Salts of Organic Acids. — Many of the metallic salts of ot'ganic acids may be ignited so as to leave the metal in a weighable form. The carbon, hydrogen, nitrogen, and some of the oxygen of the organic compound are either volatilized in the free condition or combined with each other, or they unite with the oxygen of the air to form simple oxygen compounds. Most metals retain at least some of the oxygen of the organic com- pound, or are oxidized by the oxygen of the air. Under these conditions barium, strontium, and sodium remain as carbonates with a partial reduction to oxide, which is converted to carbonate by treatment with ammonium carbonate, so that these metals are weighed as carbonates. Sodium compounds must be heated only to very low redness, because of the volatility of this metal. It is best not to attempt the complete combustion of the carbon by heat. The charred mass should be extracted with water and filtered, the solution of the sodium carbonate being evaporated to dryness in a weighed dish. Organic compounds of calcium, magnesium, aluminium, CHROMIUM, manganese, NICKEL, COPPER, LEAD, and BISMUTH may be ignited in the air until the organic matter is burned off and the oxide of the metal remains. Many organic salts froth on being heated. The crucible should therefore have a tight lid, and the heat should be applied gently at first. If this confinement is not sufficient, the crucible should be set in a larger covered cruci- 72 DETERMINATION OF METALS. ble, since the error due to loss of metal by the mechanical carry- ing away of fine particles by the products of the combustion may be considerable. The ignition of the salts containing lead or bismuth must be conducted at as low a heat as possible, as these metals are readily reduced and are volatile at red heat. When the organic material is charred and the carbon ceases to burn readily the flame is removed from the crucible and ammonium nitrate added. The crucible is covered and gently heated. The ammonium nitrate fuses and oxidizes the remaining carbon and converts the metallic lead or bismuth into nitrate which is con- verted into oxide by red heat. Concentrated nitric acid may be used for the same purpose. When copper salts are ignited, some of the copper is left as cuprous oxide, which it is most advanta- geous to oxidize by igniting with mercuric oxide. EXERCISE 17. Determination of Lead by Ignition of Lead Nitrate, Pb(N03)2. Heat a porcelain crucible to redness with the Bunsen burner for ten minutes. Cool in the desiccator and weigh. Weigh out one-half gram of lead nitrate and transfer to the crucible. If the salt is in large crystals, it should be reduced to a coarse powder by grinding in a clean and dry porcelain mortar. It should be weighed out as soon as powdered. Heat gently, at first with the lid on the crucible. When decrepitation ceases and copious nitrous fumes cease to be evolved the lid may be removed. Heat to dull redness until no more fumes are evolved. Cool in the desiccator and weigh as PbO. Repeat until the weight is constant. Calculate the percentage of lead in the lead nitrate. CHAPTER VI. DETERMINATION OF METALS AS SULPHATE AND AS SULPHIDE. A LARGE number of the metals form sulphates which are of definite composition and which can be heated high enough to expel all water and volatile matter without suffering decomposi- tion. Most of the sulphates, however, are soluble. Only three, those of BAEiUM, STRONTIUM, and LEAD, are insoluble enough for separation of the metal from water solution by precipitation. 68. Barium Siilphate is the most insoluble, requiring about 400,000 parts of water to dissolve one part of the salt. It is somewhat more soluble in even dilute solutions of hydrochloric acid and of ammonium chloride. It readily dissolves in concen- trated sulphuric acid, especially when hot, from which it is com- pletely precipitated by the addition of water. Barium sulphate has the very peculiar though by no means uncommon property of rendering insoluble some very soluble salts, notably the nitrates, chlorides, sulphates, and chlorates of the alkaline earths and alkali metals, ferric, aluminium, and chromic salts. Potassium salts give more trouble than sodium salts. Once carried down by the barium sulphate, these salts cannot be entirely washed out. Ammonium sulphate is also carried down and being volatile, is expelled when the precipitate is heated, thus producing low results. The solution must be freed from nitrates and chlorates by evaporation with hydrochloric acid. If ferric salts are reduced to ferrous salts, iron will not be carried down by the barium sul- phate. Powdered aluminium in a hydrochloric acid solution may be used for this purpose. A precipitate contaminated by any of these impurities may be dissolved in concentrated sulphuric acid and reprecipitated by diluting with much water. It may also be fused with sodium carbonate, which converts the sul- phate of barium into the carbonate, which may be filtered off, washed, redissolved, and reprecipitated with sulphuric acid. It has been shown that the contamination of the precipitate 73 74 DETERMINATION OF METALS. takes place at the moment of its formation. The most favorable conditions for avoiding the contamination have been found to be a mling-hot, slightly acid solution of large bulk, and the addition of the precipitant in very dilute solution drop by drop, and with the most vigorous stirring. If from failure to observe any of these conditions salts are carried down, washing will not remove them. On account of its slight solubility, barium sulphate tends to come down in such a finely divided condition that it passes through the pores of the filter-paper. By precipitating as directed above and keeping the solution near the boiling-point for some time the precipitate settles out, leaving a clear solution and does not pass through the filter-paper. On burning the paper some of the sulphate is reduced to sul- phide of barium. It is customary to add a drop or two of sul- phuric acid in order to reconvert the sulphide into sulphate, and then to ignite the precipitate and volatilize the excess of sulphuric acid. By this treatment barium pyrosulphate is produced, which is decomposed slowly and only at a high temperature. It has been shown that the addition of sulphuric acid is unnecessary since the barium sulphide is oxidized to sulphate by ignition in the air. Large crystals of barium sulphate which settle rapidly may also be obtained by adding a cold dilute solution of the barium to the cold dilute solution of the sulphate vnthout stirring.* This method has been developed and thoroughly tried out by Bishop and Allen. f They use a "precipitating cup" made by drawing out the tube of a Gooch funnel to a capillary tube of such size that the 5% barium chloride solution is delivered at a rate not exceeding 5 c.c. per minute. This capillary tube is bent so that it will deliver the reagent under a watch-crystal into the sulphate solution contained in a large beaker. The sulphate solution is diluted so that for one gram of barium sulphate produced, the volume is not less than 320 c.c. The solution is not stirred, while the barium chloride is being added, but after all is m, the supernatant liquor is well mixed by gentle atirring. The barium sulphate is allowed to settle, preferably for twelve hours. The * Otto Folin, Z. Anal. Chem., 45, 31 (1906). tAUen and Bishop, Eighth Inst. Con. of App. Chem., 1912, VI, p. 33. DETERMINATION OF METALS AS SULPHATE. 75 crystals of barium sulphate are large enough to be easily retained and washed on a Gooch asbestos filter. 69. Strontium Sulphate is much more soluble in water and acids than barium sulphate, 1 part of the former being dissolved by about 7000 parts of cold water, and 10,000 parts of boiling water, while if the water contains sulphuric acid, from 11,000 to 12,000 parts are required. One part of the precipitate is dis- solved by about 500 parts of dilute nitric or hydrochloric acid, while it is practically insoluble in 95% or absolute alcohol and in boiling ammonium sulphate solution (1 to 4). The precipitate should therefore be washed by one of the two latter solutions, or by dilute sulphuric acid. On ignition in presence of carbon or reducing gases it is converted into sulphide. This is volatile in the Bunsen burner flame, coloring it red, so that the precipitate must be detached as completely as possible from the paper. The subsequent treatment is similar to that of barium sulphate. 70. Lead Sulphate is quite insoluble in water, requiring for its solution about 23,000 parts of cold water, but if the water con- tains sulphuric acid 36,500 parts are required. The presence of alcohol reduces this solubility. Solutions of ammonium salts, especially the nitrate, tartrate, and acetate dissolve lead sulphate quite readily. The two latter should be rendered strongly alka- line with ammonia. It is also dissolved by hot solutions of caus- tic soda or potash. Dilute nitric and hydrochloric acids dissolve considerable quantities of lead sulphate which is reprecipitated on the addition of dilute sulphuric acid. These acids are readily removed by evaporating the solution after the addition of sul- phuric acid, until fumes of the latter appear. On diluting the solution with considerable water and adding alcohol, the lead is completely precipitated. The lead precipitate should be washed by alcohol or dilute sulphuric acid, preferably by the former as the paper would char on drying if sulphuric acid were left in it, and if washed out by water some lead would be lost. On burning the paper some of the lead sulphate is reduced to metallic lead. The precipitate should therefore be removed from the paper, the latter placed in a porcelain crucible, and ignited until all carbon is burned. A drop or two of nitric acid and the same amount of sulphuric acid 76 DETERMINATION OF METALS. is added and evaporated off, care being taken to avoid spattering. The remainder of the precipitate should tlien be added and heated with a Bunsen burner. Sulphate of lead may very advantage- ously be filtered off on a Gooch crucible and dried on the hot plate. EXERCISE i8. Determination of Barium in Barium Chloride, BaCl2.2H20. Weigh out about one-half gram of crystallized barium chloride. Transfei to a 500-c.c. beaker and dissolve in about 250 c.c. of distilled water. Heat nearly to boiling on the hot plate and add, while stirring- vigorously, dilute sulphuric acid drop by drop until no further precipitate is produced. The acid may be drawn up in a glass tube or pipette and the flow of acid regu- lated by pressing the index finger against the upper end of the tube After allowing the precipitate to settle a little, it can easily be seen if a drop of acid produces any precipitation. The heating of the solution is con- tinued until the precipitate has settled completely, leaving a clear solution. About an hour is usually sufficient. It is washed several times by decanta- tion, using about 75 c.c. of hot water each time and stirring up the precipi- tate thoroughly. The precipitate is transferred to the funnel and the washing with hot water continued until the wash-water is free-from chlorides. The moist precipitate with the paper may be transferred to a weighed platinum crucible or dish and ignited over the Bunsen burner with free access of air and finally over the blast-lamp. The theoretical percentage of barium in barium chloride is 56.25. 71. Determination of Metals by Evaporation with Sulphuric Acid. — Many metals form sulphates which are stable enough to be ignited and weighed, but which are soluble and therefore can- not be precipitated and washed. Potassium, sodium, calcium, MAGNESIUM, COBALT, NICKEL, and M.INGANESE form SUCh SUl- phates. A salt or compound of one of these metals which con- tains no other non-volatile metal and which contains only volatile acids may be analyzed by ignition with excess of sulphuric acid. The sulphate of the metal will remain, from the weight of which the percentage of the metal present may be calculated. In this manner the salts of nitric and nitrous acids, the halogen acids, organic acids, etc., may be analyzed. Organic salts of sodium and potassium are treated with a moderate excess of dilute sulphuric acid in a covered platinum crucible and gently heated to dull redness. The evolution of DETERMINATION OF METALS AS SULPHATE. 77 white fumes indicates an excess of sulphuric acid. The readily- fusible acid sulphate of the alkali remains. Small lumps of ammonium carbonate which leave no residue on ignition are thrown into the hot crucible and the heating continued. This is repeated until the alkali sulphate no longer fuses. The crucible is cooled and weighed, and the addition of ammonium carbonate and the heating repeated until constant weight is obtained. The alkali-acid sulphate may be converted into the neutral sulphate by the application of a white heat, but some loss of the metal results. Barium, strontium, and calcium may be determined in salts of volatile acids by the addition of an excess of sulphuric acid, evaporating to dryness and igniting, at first gently and finally to full redness. To remove the excess of sulphuric acid, a little ammonium carbonate is sprinkled on the precipitate, which is again heated to redness. Magnesium may be deter- mined in the same manner, but a large excess of sulphuric acid should be avoided and the residue may be exposed to a moderate red heat only or loss of sulphuric acid will result. Magnesium sulphate is very hygroscopic and must be weighed rapidly. Nickel, cobalt, and manganese ma}- be weighed as sul- phates, but care must be exercised not to overheat, as sulphuric acid may be lost. If this has occurred, as shown by blackening of the material, dilute sulphuric acid must be added, and after evaporating off the excess the ignition is repeated. The platinum dish must not be heated above dull redness. Lead may be determined in organic salts, and those of volatile acids by treatment in a weighed porcelain dish or crucible with excess of dilute sulphuric acid. Loss by spattering is avoided by keeping the vessel well covered and evaporating with very gentle heat. If the residue from the ignition of an organic compound is not perfectly white, the addition of sulphuric acid and the ignition must be repeated. Lead sulphate is stable when exposed to a red heat, but reducing gases must not be allowed to come in con- tact -with it. 78 DETERMINATION OF METALS. EXERCISE 19. Determination of Magnesium in Magnesium Carbonate. Weigh one-half gram of magnesium carbonate and transfer to a weighed platinum crucible. To prevent loss by spattering expel most of the car- bonic acid by heating the crucible at first gently, then strongly with the Bunsen burner, the Avhole operation taking about ten minutes. Allow the crucible to cool quite completely and then add about 3 c.c. dilute sul- phuric acid, at first cautiously, to prevent spattering in case much carbonic acid still remains. Evaporate the solution by first heating on the hot plate and then very gently with the Bunsen burner until acid fumes cease to be evolved. If white fumes do not escape enough acid has not been added. Heat for a few minutes to dull redness, cool in the desiccator, and weigh quickly, as magnesium sulphate is hygroscopic. Repeat the- ignition with the Bunsen burner until the weight is constant. The residue should dissolve completely in water and should form a neutral solution, otherwise it has been too highly heated. Calculate the percentage of magnesium carbonate in the sample tested. Ordinary magnesiimi carbonate approximates the formula 4MgC03.Mg(OH),.5H30. DETERMINATION OF METALS AS SULPHIDE. A large number of the metals form sulphides which are almost entirely insoluble in water, especialty when it contains an excess of the precipitant. The separation of metals as sulphides from their solution is therefore very complete, and as in some case.- the precipitates can be ignited as sulphides, giving very pure products, the determination of the metal as sulphide is in manj cases of the highest accuracy. This is true of copper, zinc, and manganese. Copper, mercury, lead, silver, bismuth, cadmium, arsenic, ANTIMONY, and TIN may be precipitated by hydrogen sulplr.de from acid solutions. Silver, cadmium, lead, and bisjiuth may also be precipitated by ammonium sulphide in alkaline solution, though the precipitation by hydrogen sulphide in moderately acid solution is generally used. Lead and silver may also be precipitated by hydrogen sulphide or ammonium sulphide in neutral solution. Cadmium may be precipitated from a potas- sium cyanide solution by hydrogen sulphide. DETERMINATION OF METALS AS SULPHIDE. 79 72. Copper may be precipitated from a neutral or slightly acid solution by ammonium sulpliocyanate after reduction by sulphurous acid, ammonium sulphite, or hypophosphorous acid. The copper is precipitated as cuprous sulphocyanate and may be dried at 100° and weighed as such, or it may be converted into cwprous sulphide by ignition with sulphur in a stream of hydrogen. Many copper compounds containing no non-volatile constituent except copper may be determined by heating a weighed portion with sulphur in a stream of hydrogen, thus converting the copper into cuprous sulphide. The oxides of copper, the sulphate, nitrate, etc., may be analyzed in this manner. If the copper solution is free from oxidizing substances, it is advisable to heat it nearly to boiling before passing hydrogen sulphide, as under these condi- tions the precipitate collects and settles better. 73. Washing. — Some of the sulphides form so-called colloidal solutions with water, while others are readily oxidized, forming sulphates which dissolve in the wash-water. For this reason the wash- water must contain hydrogen sulphide, and in the case of cadmium and arsenic hydrochloric acid should be present, more being necessary for arsenic sulphide than for cadmium sulphide. 74. Removal of Free Sulphur and Drying at 100°. — The sul- phides of SILVER, MERCURY, CADMIUM, and ARSENIC Cannot be ignited, but must be dried at 100° before weighing. A Gooch crucible is most advantageous for this purpose, though a weighed filter-paper may be used. As hydrogen sulphide is very readily oxidized even by the oxygen of the air, the precipitated sulphides wiU almost invariably contain sulphur. When the precipitate can be ignited, this contaminant does not interfere with the accuracy of the determination. When the sulphide is dried at 100° the removal of the free sulphur presents some difficulties. The solu- tion from which the metal is to be precipitated should be freed as far as possible from oxidizing agents. The precipitation should be carried out in a cold solution, which should be protected from the air as much as possible by placing it in an Erlenmeyer flask and filtering and washing immediately after precipitation. Even under these circumstances more or less sulphur is apt to be formed, so that the precipitate should always be treated for the removal of this element. Carbon disulphide and a moderately strong 80 DETERMINATION OF METALS. solution of sodium sulphite have been used for this purpose. If carbon disulphide is used, the washed precipitate is first dried, then digested with small quantities of carbon disulphide until no sulphur is obtained by evaporating a portion of the carbon disulphide. If a Gooch crucible is used, it may be placed in a small beaker and the carbon disulphide poured over the precipi- tate. The precipitate may also be freed from water by washing with alcohol, after which the carbon disulphide may be added. If the sodium sulphite solution is used, the precipitate need not be dried, but must be washed thoroughly after the sulphur has been removed. The sodium sulphite solution must be heated very nearly to the boiling-point. The solution may be tested for sulphur by acidifying and allowing to stand for a few moments. A milky precipitate indicates sulphur. The removal of the sul- phur by either of these solutions is incomplete. On drying at 100° bismuth sulphide first loses water and then i)egins to oxidize and gain in weight. This precipitate should Tnerefore be weighed after drying for half an hour. The drjdng for periods of half an hour and weighing is continuctl until the weight is constant or begins to increase. The lowest weight found is that of the dry sulphide. 75. The Sulphides of Tin cannot be weighed as such, but are con- verted into the dioxide by ignition in the air. The heat must be applied gently at first or some of the tin will be volatilized as sul- phide. The stannic sulphide is more volatile than the stannous sulphide. Wlien sulphur dioxide is no longer given off copiously the precipitate is heated more intensely. To completely expel the oxides of sulphur a little powdered ammonium carbonate is sprinkled over the precipitate, which is again ignited. Antimony sidpldde is ignited in the same manner and converted into the tetroxide ShjO^. EXERCISE 20. Determination of Copper in Copper Sulphate, CuSO^.sH^O. Weigh out 1 gram of pure crystallized copper sulphate. Transfer to a 300- «.c. beaker and dissolve in about 200 c.c. of water. Add a few cubic centimeters of hydrochloric acid, heat nearly to boiling, and pass hydrogen sulphide until precipitation is complete. Allow the precipitate to settle for a few minutes, decant the solution through the filter, and transfer the precipitate. Wash DETERMINATION OF METALS AS SULPHIDE. 81 with hydrogen-sulphide water. Before discarding any portion of the filtrate, test for copper Avith hydrogen sulphide. Copper sulphide is very readily oxidized by the oxygen of the air to copper sulphate, which passes through the filter-paper. The hydrogen sulphide of the wash-water is present to reprecipitate this dissolved copper. The precipitate should be exposed to the air as little as possible. For this purpose it should be washed down into the point of the funnel and kept well covered with the wash-water containing hydrogen sulphide. 76. Ignition of the Precipitate with Sulphur in a Stream of Hydrogen. — The precipitate is dried in the steam-oven, detached from the paper and placed in a weighed Rose crucible. The paper is then burned and the ash added to the precipitate. If the paper crumbles considerably, it may be placed in a porcelain dish or crucible and burned by heating the dish or crucible. Unburned particles of carbon sometimes remain unless the dish is heated strongly for some time. The residue is transferred to the crucible, the contents of which are mixed with an equal bulk of powdered sulphur which leaves no residue on ignition. A gentle stream of dry hydro- gen is now led in through the porcelain tube which passes through the lid of the crucible. The hydrogen is freed from arsine by passing through a solution of m.ercuric chloride. The arrangement of the apparatus is shown in Fig. 12, page 63. The hydrogen must be allowed to pass until the air has been displaced from the wash-bottles as well as from the crucible. About ten minutes is generally sufficient if the stream of hydrogen passes quite rapidly. The danger of explosion is greatly reduced if a plug of glass wool or cotton is loosely inserted in the glass tube nearest to the crucible. The crucible is at first heated gently with the Bunsen burner, and when the excess of sulphur is nearly all volatilized it is heated to redness vnth the full flame of the burner. After a few minutes' strong heating the burner is removed and the crucible is allowed to become very nearly cold in the stream of hydrogen. It is then placed in the desiccator and after a few minutes is weighed. The precipitate is again mixed with sulphur, heated in the stream of hydrogen and weighed. This is repeated until constant weight is ob- tained. The percentage of copper in crystallized copper sulphate is 25.47. The ignited precipitate is cuprous sulphide, CujS. EXERCISE 21. Determination of Arsenic in Arsenious Oxide, AsjO^. Weigh out about 0.3 gram of pure resublimed arsenous oxide. Transfer to a small beaker and dissolve in a little freshly made dilute solution of caustic soda. Transfer the solution to a 250-c.c. Erlenmeyer flask. Acidify with hydrochloric acid and then add about 30 c.c. of the dilute acid. Dilute to about 200 c.c. or until the flask is nearly full and pass hydrogen sulphide until precipitation is complete. Filter immediately through a Gooch crucible which has been dried at 100° to constant weight. Wash with about 250 c.c. 82 DETERMINATION OF METALS. of water containing a few cubic centimeters of dilute hydrochloric acid* and a considerable amount of hydrogen sulphide. Suck the water out as com^ pletely as possible with the filter-pump, wash two or three times with alco- hol, place the crucible in a small beaker and pour in carbon disulphide until the precipitate is covered. The carbon disulphide used for this purposs should leave no residue on evaporating a few cubic centimeters on a watch- crystal. If a residue is left, the carbon disulphide must be redistilled. Allow to stand about fifteen minutes. Pour off the carbon disulphide and extract again with a fresh portion. Continue this operation until a portiou of the carbon disulphide evaporated on a watch-crystal leaves no deposit of sulphur. Dry the precipitate in the steam-oven and weigh. Theoreti- cal percentage of arsenic in arsenious oxide is 75.75. 77. Precipitation of Zinc as Sulphide. — Zinc as well as man- ganese and iron are precipitated by ammonium sulphide in alka- line solution. Zinc may be precipitated by hydrogen sulphide in a solution which is only very slightly acid with a mineral acid or moderately acid with acetic acid. This sulphide is so finely divided that it passes through the filter-paper, making it very difficult to filter and weigh. If some mercuric chloride is added to the zinc solution, a precipitate is obtained which can be filtered and washed with ease. In the subsequent ignition the mercury is completely volatilized. The zinc sulphide obtained by adding ammonium sulphide to a solution of zinc made strongly alkaline with caustic soda is also easily washed, but it generally is con- taminated with alkali and sihca. The zinc sulphide is mixed with sulphur and ignited in a rose crucible in a stream of hydrogen. 78. The Two Sulphides of Manganese. — Manganese forms two sulphides, one of which is green, while the other is pink. The pink sulphide resembles the sulphide of zinc in being' so finely 'livided that filtration is difficult. The green sulphide on the other hand offers no difficulty of this kind in manipulation. It is, therefore, highly desirable to obtain this variety in determining manganese. This is accomplished by placing an excess of color- less ammonium sulphide together with some ammonium chloride in an Erlenmeyer flask, diluting to about 100 c.c. and heating nearly to boiling. The concentrated L^id hot manganese solution is then added all at once, and the flask vigorously shaken. The solution is kept hot and occasionally vigorously agitated, until * The acid prevents the formation of a " colloidal " solution. DETERMINATION OF METALS AS SULPHIDE. 83 the sulphide changes from pink to green and settles readily, leav- ing a clear supernatant liquid. Failure to obtain this result is generally due to the presence cf an insufficient amount of ammo- nium sulphide. The precipitate is filtered immediately, and washed with water containing ammonium sulphide and ammonium chloride. 79. Precipitation of Iron as Sulphide. — Iron is precipitated as sulphide from solutions containing organic matter, which prevents the complete precipitation of the metal in the usual manner as hydroxide. Ammonium chloride must be present, and either the yellow or colorless ammonium sulphide may be used for pre- cipitation. The solution must stand in a warm place until the precipitate has completely settled, leaving ■ a clear supernatant fluid. Ammonium chloride and sulphide must be present in the wash-water, as the sulphide oxidizes to sulphate, which passes through the filter-paper unless it is reprecipitated by the ammo- nium sulphide. The bulk of the dried precipitate should be removed from the paper before burning the latter, though there is no danger of losing iron. The oxide formed during the com- bustion of the paper is reconverted to sulphide by ignition in hydrogen after the addition of sulphur. EXERCISE 22. Determination of Manganese in Potassium Permanganate, KMnO^. Weigh out one-half gram of pure potassium permanganate. Transfer to a small beaker, add a few cubic centimeters of cone, hydrochloric acid, cover with a watch-crystal, and evaporate to dryness on the water-bath. Dissolve in a few c.c. of water and neutralize the solution with ammonia. Prepare some colorless ammonium sulphide bj' adding 100 c.c. of water to 50 c.c. concentrated ammonia (sp. gr. 0.90), saturating half of this solu- tion ynth. hydrogen sulphide and then adding the other half of the solution. In an Erlenmeyer flask of about 250 c.c. capacity is placed 25 c.c. of the ammonium sulphide solution, 10 c.c. of a five times normal solution of ammonium chloride or about 3 grams of the dry salt. The solution is diluted to about 100 c.c. and warmed on the hot plate or with a Bunsen burner. As soon as it comes to a boil the hot manganese solution is added, the beaker being rinsed with a little water. The flask is shaken vigorously and the solution kept nearly at the boiling-point. After alternate shaking and heating for a few minutes the manganese sulphide which came down pink turns green and settles readily, leaving a clear supernatant liquid. 84 DETERMINATION OF METALS. Failure to secure this result is due to the ab3enc,e of a sufficient amount of ammonium sulphide, the reagent being weak or the salt was boiled away before adding the manganese solution. If the addition of more ammonium sulphide and heating fails to produce the green precipitate, the operation should be repeated, using more ammonium sulphide. The green sulphide is allowed to settle for half an hour, the solution being kept warm. The precipitate is then filtered off and washed with water con- taining ammonium sulphide and ammonium chloride. It is dried and de- tached from the paper, which is burned, the ash being placed on top of the precipitate in a weighed porcelain crucible. The precipitate is mixed with two or three times its bulk of sulphur which is free from non-volatile matter. It is ignited in a stream of hydrogen which has been passed through a solu- tion of mercuric chloride to free it from arsine and then dried over calcium chloride or concentrated sulphuric acid. The crucible is heated, at first gently, until the sulphur begins to burn, and then with the full flame of a Bunsen burner, giving a flame at least 15 cm. long when burning free. The crucible is allov/ed to become nearly cold in the stream of hydrogen. If brown particles of oxide.5 of manganese are still visible, the precipitate should be again ignited after mixing with sulphur, and after cooling in the stream of hydrogen is ti'ansferred to the desiccator, completely cooled and weighed. No further change in weight should be observed on again heating with sulijhur in the stream of hydrogen. CHAPTER VII. DETERMINATION OF METALS AS PHOSPHATE, ARSE' NATE, CHROMATE, AND CHLORIDE. Manganese, zinc, and magnesium may be precipitated and weighed as phosphates. Although most of the metals form phos- phates which are quite insoluble, only those of the three metals mentioned are sufficiently insoluble or stable on ignition for quantitative determination. Phosphoric acid is one of the most non-volatile acids, but on account of its weak affinit}^ and ten- dency to form acid salts it is difficult to procure a precipitate of constant composition. It may also be reduced more or less com- pletely during the combustion of the filter-paper. When phos- phoric acid is present with iron, aluminium or chromium, these metals are precipitated with ammonia, which produces phosphates of these metals if the phosphoric acid is in ex- cess, otherwise a mixed precipitate of the hydroxides and phos- phates is produced. In either case the precipitate is ignited and weighed and the phosphoric acid determined and cal- culated as P2O5 and subtracted from the weight of the mixed ignited precipitate to obtain the weight of the oxides of the metals present. 80, Precipitation of Manganess and Zinc. — With zinc and manganese an amorphous precipitate is first formed. This is the neutral metallic phosphate Mg(P04)2,* and must be converted into the double ammonium phosphate MNH4PO4. This is ac- compUshed by heating the amorphous precipitate with the solu- tion containing a large amount of ammonium chloride. 20 grams of the dry salt should be present for 0.2 gram of the metal. 5 to 10 c.c. of a cold saturated solution of microcosmic salt is then * In this formula M may be either Mn or Zn. 85 86 DETERMINATION OF METALS. added and the solution diluted to 200 c.c. The solution is then cautiously neutralized with ammonia, and on the first appearance of a precipitate it is heated nearly to boiling and vigorously stirred until the amorphous phosphate has become crystalline. Ammo- nia is added drop by drop and the stirring continued after each addition of ammonia until the precipitate is crystalline. The addition of ammonia and the stirring is continued until no fur- ther precipitate forms. With manganese a large excess of ammo- nia should be avoided. The zinc solution should be made exactly neutral. The heating of the solution must be continued during the addition of ammonia until the phosphate is completely pre- cipitated and converted into the crystalline form. A platinum dish is most suitable for this purpose. A porcelain dish is prefer- able to a beaker, as the latter is attacked by the hot alkaline solu- tion, contaminating the precipitate with silica. The solution should contain no other metals than the alkalies. . The precipitate is washed with water containing ammonium nitrate. 8i. Magnesium forms a crystalline precipitate much more read- ily than zinc and manganese. The precipitation is carried out in a cold solution which must be allowed to stand for at least twelve hours in order to secure complete precipitation. If the solution is stirred or vigorously shaken, the precipitation is com- plete in an hour or two. If the amount of ammonium salts pres- ent is considerable, the magnesium is precipitated more slowly, requiring at least twenty-four hours' standing. In this case the composition of the precipitate is also different, containing rela- tively more ammonia. Such a precipitate is not suitable for weighing, but must be dissolved in the least amount of hydro- chloric acid and reprecipitated by the addition of a little alkali- phosphate solution and ammonia drop by drop with vigorous stirring. The volume of the solution should be kept small on account of the solubility of the precipitate. It should not exceed 100 c.c. 82. Removal of Ammonium Salts. — If one does not wish to make two precipitations in this manner, the ammonium salts must first be removed. A very convenient method of accom- pHshing this object is that suggested by J. L. Smith. The solu- tion is evaporated in a casserole or porcelain dish over the free DETERMINATION OF METALS AS PHOSPHATE. 87 flame to a sirupy consistency. Dilute nitric acid is then added in small portions, and the heating continued until nitrous fumes are no longer evolved. A little hydrochloric acid is then added in small portions and the heating continued until the excess of nitric acid is decomposed and chlorine is no longer evolved. The ammonium salts may also be removed by evaporating the solution to dryness in a porcelain dish and gently heating the residue. The magnesium remains as oxide and is dissolved in hydrochloric acid. The solution is then made alkaline with an-mionia. If a precipitate of magnesium hydroxide is formed, the solution should be acidified with hydrochloric acid and again made alkaline with ammonia. This operation should oe repeated until no precipitate occurs on making the solution alkaline. Finally the solution is acidified and a solution of di- sodium phosphate added in sufficient amount to precipitate all of the magnesium. It is then made alkaline by the addition of ammonia with constant stirring until no further precipitate is formed. The solution must be allowed to stand twelve hours before filtering off the precipitate. A few drops of the clear liquid should give a distinct test for phosphoric acid with mag- nesia mixture, otherwise enough of the precipitant has not been added. The precipitate is washed with dilute ammonia, one part of strong ammonia to ten parts of water. 83. Ignition of Phosphate Precipitates. — All of the phosphate precipitates lose anamonia and water on ignition and are con- verted into pyrophosphates according to the equation 2NH,MP0, =2NH3 +R,0 +M,P20,. If too much ammonium salts have been present in the solution, the precipitate will contain more ammonia than corresponds to the formula NH^MPO^. Some metaphosphate of the metal, M(P03)2, win be formed on the ignition of such a precipitate. The meta- phosphate loses phosphorous pentoxide, and is converted into pyrophosphate only after very prolonged and intense ignition. Such a precipitate could be considered constant only when it does not lose weight after a half-hour's ignition with the blast-lamp. For this reason the excess of ammonium salts is removed before 88 DETERMINATION OF METALS. precipitation, as already directed. The evolution of ammonia and water is quite rapid if the precipitate is heated strongly, resulting in appreciable loss of material due to spattering. The precipitate should therefore at first be heated very gently with the Bunsen burner with the lid on the crucible. When no more ammonia is evolved the crucible may be heated with the blast- lamp. As the phosphates fuse readily, the complete combustion pf the filter-paper is secured with some difficulty. It is best to dry the precipitate and remove it as completely as possible from the paper, and burn the latter on the platinum wire, using the oxid- izing flame of the Bunsen burner to assist the combustion. It the ash which falls into the crucible contains carbon which cannot be burned by simple ignition, it may be moistened with a drop oi- two of concentrated nitric acid which is volatilized by gentle heat and the ignition continued. The difficulty experienced in burn- ing the carbon is frequently due to insufficient washing of the precipitate, so that it still contains sodium acid phosphate. If the platinum crucible is used for the ignition of phosphates, it is sometimes damaged by the reduction of the phosphates and the union of the phosphorus with the platir.um. The likelihood of damage seems to be much reduced if the moist precipitate and paper are introduced into the crucible and the ignition conducted by gradually bringing the crucible to redness with the Bunsen burner. The reducing gases from the flame must be absolutely excluded. Ignition of the precipitate in a porcelain crucible is entirely satisfactory. EXERCISE 23. Determination of Magnesium in Magnesium Sulphate, MgSOj.yH^O. Weigh out 1 gram of the pure recrystallized salt. Transfer to a small beaker and dissolve in 50 c.c. of water. Add 5 o.c. dilute hydrochloric acid and 15 c.c. of disodium phosphate solution. Add dilute filtered ammo- nia slowly with vigorous stirring, being careful to avoid rubbing the sides of the beaker with the stirring-rod, as very firmly adhering crystals would form where the beaker is rubbed. Add a few c.c of ammonia in excess, allow it to stand for twelve hours, then transfer to a funnel containing a small paper and wash with dilute filtered ammonia, (1 part of strong ammonia DETERMINATION OF METALS AS PHOllPHATE 89 to 10 parts of water). The filtrate may be tested for chlorides by acidify- ing with nitric acid and adding a drop of silver nitrate solution. As the ammonia frequently contains chlorides, the washing should be continued until the filtrate gives no more turbidity with silver nitrate than the dilute ammonia. The precipitate is dried in the steam-oven, removed from the paper quite completely by means of a brush and transferred to a watch-crystal The paper is burned on the platinum wire, heating the charred portions from time to time in the oxidizing flame of the Bunsen burner. The ash is allowed to drop into a weighed porcelain crucible. If unburned carbon still remains, the open crucible should be placed on its side and heated with the Brmsen burner. If necessary, the combustion is completed by allowing the crucible to cool, adding a few drops of concentrated nitric acid and placing the Hd on the crucible and heating gently until the nitric acid is volatilized. When the carbon has been completely burned the main portion of the precipitate is added and the covered crucible very gently heated until no more ammonia fumes escape. It is then heated with the fuU flame of the Bunsen burner and finally for five or ten minutes with the blast-lamp. It is then weighed and again heated with the blast-lainp and weighed. This is repeated until constant weight is obtained. Theo- retical percentage of magnesium in crystallized magnesium sulphate is 9.88. 84. Determination of Arsenic as Magnesium Pyroarsenate. — When present in its higher state of oxidation arsenic may be precipitated as magnesium ammonium arsenate, which is ignited and weighed as magnesium pyroarsenate, MgjAsjOj. Solution of compounds of arsenic in hydrochloric acid and potassium chlorate, aqua regia, or strong nitric acid converts the arsenic into arsenic acid.* The solution is filtered if necessary, evapo- rated to a small bulk, neutralized with ammonia, and magnesia mixture added with stirring. Alcohol is added to the extent of one-fourth of the volume of the solution. After standing twelve or, still better, as long as forty-eight hours the precipitate is filtered off and washed with a mixture of 2 volumes of strong ammonia, 2 volumes of water, and 1 volume of alcohol. The precipitate is dried, detached from the paper and placed on a watch-crystal. The paper is returned to the funnel and the remaining portions of the precipitate dissolved in hot dilute nitric acid, and the paper washed with hot water. If it is not large, the * Note for student: ^^'rite the equations for these reactions. 90 DETERMINATION OF METALS. entire precipitate may be dissolved in nitric acid. The solution is evaporated to dryness in a weighed porcelain crucible and the remainder of the precipitate added. With the lid on, the crucible is heated very gently \mtil the ammonia is expelled. There is great danger of loss of arsenic before the ammonia is expelled because of its reducing action on the arsenic. The presence of nitric acid tends to prevent this, while by heating in a stream of oxygen the time required to expel the ammonia may be reduced to ten minutes. Moistening the precipitate with nitric acid is almost as efficient. When the ammonia is completely expelled the crucible is heated to redness with the Bunsen burner. The precipitate is then magnesium pyroarsena,te, MgjAsjO,. The pre- cipitate may also be collected on a Grooch crucible and ignited with the precautions given. 85. Determination of Lead, Barium, and Chromium as Chro- mates. — ^Lead and barium form very insoluble chromates which are of definite composition and may be completely dried without suffering decomposition. These precipitates may therefore serve to determine not only lead and barium, but also chromium. The precipitation is carried out in a solution acid with acetic acid. In determinations of lead and barium, potassium dichromate is used as the precipitant, while chromic acid is precipitated by lead acetate or barium chloride. Solutions acid with hydrochloric or nitric acid are rendered acid with acetic acid by addition of sodium acetate. The precipitate is washed with water, filtered on a Gooch crucible, and dried on the hot plate, or it may be gently ignited with the Bunsen burner, care being taken not to allow it to fuse. 86. Determination of Potassitxm as Platino-chloride. — Potas- sium may be precipitated as potassium platino-chloride, collected on a Gooch crucible which has been dried at 100° and weighed, or it may be collected on a filter-paper, washed with 80% alcohol, dissolved in hot water, and the solution evaporated to dryness on the water-bath in a dish and weighed. The precipitate may also be decomposed by ignition with a reducing agent, such as oxalic acid, or in a stream of hydrogen. The potassium then remains as chloride and is washed out of the reduced spongy platinum by hot water. The platinum mav then be gently ignited DETERMINATION OF METALS AS CHLORIDE. 01 and weighed. Potassium platino-chloride is soluble in 100 parts of cold water and 20 parts of hot water, in 40,000 parts of absolute alcohol and 25,000 parts of 80% alcohol. The presence of sodium platino-chloride and excess of the precipitant, platinum tetra- chloride, or, more correctly, chlorplatinic acid, decreases the solu- bility. As the precipitate is washed with alcohol, salts insoluble in this liquid must be absent. Practically the solution may contain only chlorides of sodium and potassium but sufficient platino- chloride solution must be added to convert both of these salts into the double platinum salt. 87. Lindo-Gladding Method. — If the Lindo-Gladding method of determining potassium is used, sulphates may be present, but the heavy metals should be absent. If chlorides are not present, sodium chloride should be added, and after the addition of an excess of platino-chloride solution the solution is evaporated to a sirupy consistency. The sodium platino-chloride and the excess of the precipitant is washed out by means of 80% alcohol. The sulphates are washed out with a solution of ammonium chloride made by dissolving 100 grams in 500 c.c. of water and shaking up the solution with 10 grams of potassium platino-chloride. The ammonium chloride is now washed out with 80% alcohol. While the Lindo-Gladding method has been extensively used by Ameri- can chemists, the German workers still prefer to remove sulphates by precipitation with barium chloride. 88. Determination of Silver as Chloride. — Silver as ■ well as mercury and bismuth may be precipitated as chloride and weighed as such. The neutral or slightly acid solution of silver is precipi- tated by the addition of hydrochloric acid or sodium chloride. It should be heated nearly to boiling before precipitation and the chloride added while stirring the solution vigorously. For this purpose it is convenient to place the solution in an Erlenmeyer flask. The addition of a considerable excess of chloride should be avoided, as silver chloride is appreciably soluble both in hydro- chloric acid and bi sodium chloride. After each addition of the reagent, and vigorous shaking or stirring of the solution, the pre- cipitate should be allowed to settle and a drop of chloride added to the clear Uquid. If a precipitate appears, more of the precipitant must be added. When the silver has been completely precipi- 92 DETERMINATION OF METALS. tated the solution should be heated on the hot plate or the water- bath with occasional stirring until the solution is clear. The pre- cipitate is filtered off on a Gooch crucible which has been dried on the hot plate or in an air-bath at a temperature above 120°. Silver chloride loses chlorine when exposed to light, acquiring a blue color. It should therefore be protected as much as possible during the manipulation from the action of strong light. It may also be filtered off on paper and washed, as already de- scribed. The precipitate is dried and removed from the paper as_ completely as possible and placed on a watch-crystal. The paper is now burned in a weighed porcelain crucible. The silver chloride left on the paper is reduced to metallic silver. It is reconverted to chloride by the addition of a drop of nitric acid and one of hydrochloric and evaporating to dryness. As silver chloride is quite volatile, the amount left on the paper should be very small. The main portion of the precipitate is now trans- ferred to the crucible and dried by gently heating with the Bunsen burner. The precipitate must not be allowed to fuse. The Bun- sen burner, having a small flame, should be held in the hand, and immediately removed when the precipitate begins to fuse around the edge. The results are very accurate if the determination is properly carried out. 89. Washing the Precipitate without Filtration. — The sih-er chloride may also be washed a,nd transferred to a crucible without Fig. 14. the use of filter-paper. The precipitation is carried out in an Erlenmeyer flask. When all of the silver has been precipitated the fl.ask is shaken vigorously, until the liquid is perfectly clear DETERMINATION OF METALS AS CHLORIDE. 93 and the precipitate has collected into a compact mass. The liquid is carefully decanted, and the precipitate washed by decan- tation. The flask is then filled with water, and a weighed porce- lain crucible inverted over the mouth of the flask. By quickly in^'erting the flask the precipitate falls into the crucible, which is placed on the desk while the flask is held upright with the mouth in the bottom of the crucible. By gently tapping the flask any adhering particles of the precipitate fall into the crucible. By tilting the flask slightly, bubbles of air enter and the water flows quietly into the crucible. When the latter is full the flask may be removed by quickly passing it to one side of the crucible. The water in the crucible is -decanted as much as possible, the portion remaining being evaporated on the water-bath. The precipitate is then heated in the usual manner and weighed. 90. Determination of Mercury as Mercurous Chloride. — For precipitation as chloride, mercury must be in the mercurous con- dition. The cold and highly dilute solution is mixed with sodium chloride solution until precipitation is complete. After allowing the precipitate to settle, filter on a Gooch crucible dried at 100°. Wash with cold water. Weigh after drying at 100°. If the solu- tion contains much nitric acid it should be nearly neutralized with sodium carbonate before precipitation of the mercury.. If the metal is present in the solution as mercuric mercury, it may be reduced by means of phosphorous acid. Hydrochloric acid together with excess of phosphorous acid is added to the cold solution, which is allowed to stand for twelve hours cold or heated not higher than 60°. The precipitate of mercurous chlo- ride is filtered off, washed, dried, and weighed as already directed. 91. Determination of Bismuth as Oxychloride. — Bismuth is precipitated and weighed as the oxychloride, BiOCl. If the solution is strongly acid, potassium, sodium, or ammonium hydrox- ides are added until not over J per cent of free acid remains. If chlorides are not present, ammonium chloride is added and the solution is then largely diluted by the addition of water. After standing some time water is added to a portion of the clear liquid. If a precipitate forms, more water must be added to the entire solution. When all of the bismuth has been precipitated the 93a, DETERMINATION OF METALS. oxychloride is filtered off on a Gooch crucible which has been dried at 100°. It is washed with water containing a few drops of hydrochloric acid, dried at 100°, and weighed. If sulphuric or phosphorous acids are present, small quantities of these acids are apt to contaminate the precipitate. The solution should contain only nitric and hydrochloric acids. Because of the ten- dency of the precipitate to lose chlorine, the results are not quite as reliable as when the bismuth is weighed as oxide. A little hydrochloric acid is added to the wash-water to prevent this loss. EXERCISE 24. Determination of Silver in Silver Nitrate, AgNOj. Prepare a Gooch crucible with a layer of asbestos about J cm. thick. Wash until no fine particles of asbestos come through with the wash-water. Dry on the hot plate or in an air-bath at 120° to 150° and weigh. Usually two hours' drying is sufficient. Weigh out J gram of silver nitrate and transfer to a 250-c.c. Erlen- meyer flask. Dissolve in 100 to 150 c.c. of water and heat nearly to boiling. Add dilute hydrochloric acid, a few drops at a time, shaking the flask vigorously after each addition. To avoid adding a large excess of the acid notice after each addition if a precipitate is produced. When sufficient hydrochloric acid has been added the flask should be warmed and vigorously shaken until the curdy precipitate is well collected and the solution is very nearly clear. Decant the clear hquid on the weighed Gooch crucible, wash two or three times by decantation, using hot water which contains a few drops of concentrated nitric acid. Finally transfer the precipitate to the crucible, using the " policeman " to clean the flask if necessary. Dry the Gooch crucible as before and weigh. The silver chlo- ride should be exposed to the light as little as possible. The theoretical percentage of silver in silver nitrate is 63.50. Review. — At this point the student is advised to review the deter- minations of the metals by tabulating the methods as follows : In the first column -write the name or symbol of each metal studied. In the next column headed "Precipitated as" write the formulas of the precipitates in which each element may be separated from solution. In the next column headed "Remarks^ sources of error, etc./' indicate briefly the DETERMINATION OF METALS AS CHLORIDE. 936 special precautions to be observed in each precipitation. In the next column headed "Weighed as" give the formulas of the dried or ignited precipitates. In the next column headed " Methods of drying" state whether the precipitate is dried at 100° on the hot- plate or by ignition. DETERMINATION OF ACIDS, CHAPTER VIII. DETERMINATION OF THE HALOGENS, SUL- PHUR, AND NITROGEN. The gravimetric determination of acids is more difficult than the determination of metals because the acid radicals are as a rule more volatile and unstable than the metallic compounds. In the determination of metals various acid radicals have formed a part of precipitates which can be washed and weighed. Among these are the carbonates of the alkaline-earth metals; the sul- phates of lead and barium, the sulphides of mercury, cadmium, silver, and arsenic, the phosphates of magnesium, zinc, and manga- nese, the chromates of barium and lead, and the chlorides of silver, mercury, and bismuth. In a good many cases it is found possible to use a given precipitate for the determination of either the metal or the acid radical. DETERMINATION OF THE HALOGENS. 92. Precipitation as the Silver Salt. ^ — Where a given acid forms a weighable precipitate with several metals, choice is made of the most insoluble and most stable precipitate. The chloride OF SILVER has been found to be the most advantageous form in which to precipitate and weigh either silver or the chlorine radical. In the determination of a chloride or of hydrochloric acid the process is very nearly identical with the determination of silver as given on page 90. The chloride solution must be slightly acid. For this purpose the neutral or alkaline solution is acidified with nitric acid, adding a drop or two in excess to the neutral solution. The solution is heated to about 60° and silver nitrate solution added with constant stirring until no more precipitate is formed. The solution is then heated nearly to boiling and stirred vigorously until the precipitate coagulates, leaving a clear superna- 94 DETERMINATION OF THE HALOGENS. 95 tant liquid. Failure to secure this result may be due to the absence of an excess of silver nitrate. The precipitate is filtered off on a weighed Gooch crucible and washed with hot water containing a little silver nitrate. The presence of sih-er nitrate reduces very much the solubility of silver chloride in water. When the other impurities have been washed out the silver nitrate is removed by washing with as little distilled water as possible. The pre- cipitate is then dried on the hot plate, or in an air-bath at about 150°. Silver chloride should be protected from the light as much as possible, as it turns dark, losing chlorine. The method gives very accurate results. Iodides and beomides as well as the free acids may be determined in the same manner. 93. Ignition of Paper. — If a Gooch crucible is not at hand, the silver halides may be filtered off on paper, washed in the manner already described, and dried. The precipitate is then separated from the paper as completely as possible and placed on a watch- crystal. The paper is burned on the platinum wire or in a weighed porcelain crucible. The silver chloride or bromide remaining on the paper is reduced to metallic silver. The ash of the paper in the crucible is treated with a drop or two of nitric acid and a drop of hydrochloric or hydrobromic acid to reconvert the silver into the proper halide salt. The excess of acid is evaporated ofl by gentle heat and the remainder of the precipitate added. The crucible is now heated very gently with the Bunsen burner, care being taken not to fuse the precipitate, which volatilizes slightly when fused. The silver iodide which remains on the filter-paper is not reduced when the latter is burned, but on account of its volatility it should be removed as completely as possible. The silver precipitate may also be washed without filtration as described in Chapter VII, page 91. 94. Determination of the Halogens in Metallic Salts. — The chlorine in mercuric chloride, stannous chloride, platinic CHLORIDE, the chloride of antimony, and the green chloride of chromium cannot be determined by precipitation with silver nitrate. The precipitate is contaminated in the case of the stannous salt with stannic oxide and silver oxide, in the case of the mer- curic and antimony salts with metallic mercury or antimony. All of the chlorine is not precipitated from the chromium solution. 96 DETERMINATION OF ACIDS. Platinous chloride contaminates the precipitate of silver chloride from platinic chloride. The tin in stannous chloride is therefore precipitated by boiling the concentrated neutral solution with ammonium nitrate. The mercury and antimony are precipitated with hydrogen sulphide, tartaric acid having been added to the anti- mony solution. The chlorine in platinic chloride may be determined by fusing the salt with sodium carbonate, filtering off the platinum from the solution of the fusion and determining the chlorine in the filtrate. The chromium in the green chloride is precipitated by means of ammonia, filtered off and washed, and the chlorine precipitated with silver nitrate in the filtrate. The corresponding salts of BROMINE and iodine are treated in a similar manner. The INSOLUBLE CHLORIDES, of LEAD, SILVER, and MERCURY must be decomposed by alkalies for the separation of the metal. Silver chloride may be fused in a porcelain crucible with sodium and potassium carbonates. Lead chloride may be treated in the same manner or digested with alkali bicarbonates and water, while mercurous chloride may be decomposed by digestion with a solution of sodium or potassium hydroxide. The insoluble BROMIDES and iodides may be treated in a similar manner. SEPARATION OF CHLORINE, BROMINE, AND IODINE. 95. Separation of Iodine as Palladous Iodide from Chlorine and Bromine. — ^Very excellent methods have recently been worked out for the separation of chlorine, bromine, and iodine. Iodine may be precipitated either as palladous or thallous iodide from solutions which contain chlorine or bromine. For the precipita- tion as palladous iodide the solution is made slightly acid with nitric acid and palladous nitrate added in slight excess. The solution is allowed to stand for twenty-four to forty-eight hours. The brownish-black precipitate of palladous iodide, Pdlj, is fil- tered off on a Gooch crucible or on a tared paper, washed with warm water, and dried at 100° to constant weight. The paper may also be burned, and the precipitate ignited in a Rose crucible in a stream of hydrogen. It is then reduced to metallic palla- dium, and may be weighed as such. If only hydrobromic or hydrochloric acid is present in the filtrate, it may be precipitated with silver nitrate, after removal of the excess of palladium with hydrogen sulphide. Tiie excess SEPARATION OF THE HALOGENS. ■ 97 of hydrogen sulphide must be removed by means of ferric sul- phate or hydrogen peroxide, and the precipitated sulphur filtered off after digesting some time on the water-bath. If sufficient material is at hand, the iodine may be determined by means of palladous nitrate or chloride in one portion, and the iodide together with the chloride or bromide precipitated and weighed as the silver salt in another equal portion. The weight of silver iodide is then calculated from the weight of the palladium pre- cipitate and subtracted from the weight of the combined silver precipitate. 96. Separation of Iodine from Chlorine as Thallous Iodide. — Iodine may also be separated from chlorine by precipitation as thallous iodide, according to Jannasch and Aschoff. Both thallous iodide and chloride are slightly soluble in cold water, the iodide less than the chloride, however. The presence of ammonium sulphate increases the solubility of the chloride and decreases the solubility of the iodide, which is almost absolutely insoluble if alcohol is present. To the solution of the chloride and iodide, which should have a volume of about 50 c.c, the same volume of 20% ammonium sulphate solution is added and 30 c.c. of alcohol. A 4% solution of thallous sulphate is now added until the iodide is precipitated. The solution is gently warmed, well stirred, and allowed to stand for twelve hours. The precipitate of thallous iodide. Til, is filtered off on a Gooch crucible, and washed with a mixture of 5 parts ammonium sulphate, 70 parts of water, and 30 parts of alcohol. It is dried at 100°. After expelling the alcohol from the filtrate the chlorine may be precipitated with silver nitrate after diluting considerably to prevent the precipitation of silver sulphate. For the same reason the solution should be digested hot for some time. 97. Separation of Chlorine, Bromine, and Iodine. — Jannasch and Aschoff have also worked out a method of separating the three halogens which depends on the fact that iodine is liberated from hydriodic acid by nitrous acid, while neither hydrochloric nor hydrobromic acid is oxidized under these conditions. The liberated iodine may be removed from the solution by means of carbon disulphide, or it may be volatilized with steam after dilut- ing the solution so that neither hydrobromic nor hydrochloric 98 DETERMINATION OF ACIDS. acid is expelled. Ferric sulphate and a number of other reagents have been found to liberate iodine and not chlorine or bromine. The hydrobromic acid may be oxidized by potassium permangan- ate and acetic acid, and the bromine driven out of the solution by means of steam. The chlorine may then be precipitated with sil- ver nitrate. A suitable arrangement of apparatus is shown in Fig. 15. The ground-glass joint of the tubes entering the flask K is very essen- Fig. 15. tial, as neither cork nor rubber is unacted on by the halogens. The most satisfactory substitute for a grotmd-glass joint is a cork stopper coated with parafiine. The stoppers in the absorp- tion-flask as well as the Peligot tube are of this character. The distilling-flask K should have a capacity of at least 1 liter, and the absorption-flask a capacity of 500 c.c. W is a tin or metallic, vessel for generating steam. The neutral solution of the halogens is placed in the flask K, diluted to about 750 c.c. with water, 5 c.c. of dilute sulphuric acid added, as well as 1 gram of sodium nitrite dissolved in a little water. The flask is closed immediately. The absorption-flask should contain a mixture of 50 c.c. of caustic soda * solution and * If caustic soda free from the halogens is not at hand, it may be made by dis- solving metallic sodium in distilled water. SEPARATION OF THE HALOGENS. 99 an equal volume of hydrogen peroxide solution. The bulbs of the Peligot tube should be nearly filled with a similar solution, and a little dilute caustic soda solution should be placed in the Erlenmeyer flask. 98. Distillation of the Iodine. — On heating the contents of the flask K to boiling and passing a current of steam, the iodine is driven over into the absorption-flask, ■'.vhich is kept cold by means of the cold water in the beaker. The distillation is continued until the solution in K is colorless, which requires about fifteen minutes. All of the iodine will then be driven out. On being absorbed by the caustic soda, part of the iodine is converted into iodate according to the equation 6NaOH +3I2 =NaI03 +5NaI +3Rfi. The hydrogen peroxide present converts the iodate into iodide: NaI03 +3HP3 =NaI +3H,0 +30,. 99. Determination of the Iodine as Silver Iodide. — While the steam is still passing, the deliver5'-tube is withdrawn from the absorption-flask and rinsed with a little water. The contents of the flask as well as the Peligot tube are transferred to a porcelain dish, 50 c.c. of hydrogen peroxide added, and the solution warmed for some time on the water-bath. Silver nitrate is added until a permanent precipitate of silver hydroxide is formed. On digesting the precipitate the brown silver hydroxide is converted into the yellow iodide. If the brown color disappears entirely, more silver nitrate must be added. Finally the solution is acidified with nitric acid, warmed, and the silver iodide filtered off on a Gooch crucible, washed with hot water, dried on the hot plate or in the air-oven at 150° and weighed. 100. Distillation of the Bromine. — The contents of the dis- tilling flask are made faintly alkaline with caustic soda and con- centrated to a volume of 500 c.c. The solution is cooled, 60 c.c. of 33% acetic acid is added, and 1 to 1^- grams of potassium per- manganate dissolved in a little water. The absorption flask and Pelisot tube are filled, and the distillation conducted exactlv as directed for the iodine. The bromine is expelled from the solution with more difficulty than in the case of the iodine, 45 to 75 minutes 100 DETERMINATION OF ACIDS. being generally required. The distillation must be continued foi some time after bromine vapors are no longer visible. The solution of the bromine is transferred to a porcelain dish, 50 c.c. of hydrogen peroxide solution added, and warmed on the water-bath. The bromine is then precipitated, washed, and weighed in the same manner as the iodine. loi. Determination of the Chlorine. — The potassium per- manganate in the distilling-flask is decomposed by neutralizing the solution with caustic soda, adding a little methyl alcohol and warming. The manganese dioxide is filtered off and washed with warm water. The chlorine in the filtrate is precipitated in the usual manner. 102. Determination of Chloric Acid. — Chloric acid may be determined by precipitation with silver nitrate after reduction with pure zinc and sulphuric acid. 103. Determination of Hydrocyanic Acid. — Hydrocyanic acid may be precipitated and weighed as silver cyanide. The solution should be dilute and slightly acid with nitric acid. Excess of silver nitrate is added to the cold solution with vigorous stirring. The precipitate may be collected on a weighed filter-paper or, still better, on a Gooch crucible and dried at 100°. It may also be col- lected on an unweighed filter-paper and converted into metallic silver by simple ignition \intil the weight is constant. This opera- tion should be carried out in a porcelain crucible. Heating to redness for one-quarter hour is generally sufficient to completely decompose the cyanide. 104. Separation of Hydrocyanic Acid from Chlorides, Bromides, and Iodides. — If chlorides, bromides, or iodides are present in the solution of the cyanide, the precipitated silver salt of the halogen must be separated from the silver cyanide. For this purpose the precipitate is digested with a solution of mercuric oxide in dilute acetic acid. The silver cyanide is converted into solublft mercuric cyanide and silver acetate: 2AgCN +Hg(C,H30,), =Hg(CN), +2AgC,H30,. A solution of mercuric oxide in 100 c.c. of water and 5 c.c. dilute acetic acid is sufficient to dissolve the silver cyanide from 0.1 gram of potassium cyanide. The halogen salts of silver are not decom- DETERMINATION OF SULPHUR. 101 posed by the mercury solution, which is filtered off and the dis- solved silver precipitated as chloride after the addition of nitric acid. The silver chloride must be ignited in a stream of hydrogen in a porcelain crucible. Some mercury is carried down with the silver and cannot be entirely removed by drying the silver chloride. The presence of the halogens do not interfere in the volumetric determination of cyanogen. DETERMINATION OF SULPHUR. los. Detennination of Sulphur by Fusion with Alkali Carbon- ates and Nitrates. — Sulphur in sulphides and in most of its com- pounds is usually converted into sulphuric acid and precipitated and weighed as barium sulphate. Many methods of oxidation have been proposed for converting the various sulphur compounds into sulphuric acid or soluble sulphates. Perhaps the most gen- erally applicable method is fusion with sodium carbonate and potassium nitrate. The finely powdered and weighed material is intimately mixed with six parts of dry sodium carbonate and four parts of potassium nitrate. The proportion of potassium nitrate should be reduced if the percentage of sulphur is small, as when much nitrate is present the action on the platinum crucible is considerable. This mixture is transferred to the crucible, and gradually heated until fused. The iUuminating-gas used for the Bimsen burner frequently contains sulphtu", which may be ab- sorbed by the fusion mixture. To prevent this it is advisable to place the crucible in a hole cut in a piece of asbestos board.* The crucible should not be heated higher than necessary to keep the contents fused. When cool, dissolve the residue in hot water, filter, and wash with water containing a little sodium carbonate. Acidify the filtrate with hydrochloric acid and evaporate to dry- ness in a porcelain dish. Add a little dilute hydrochloric acid and evaporate to dryness again to completely remove the nitric acid. Dissolve the chlorides in water and a drop or two of dilute hydrochloric acid. Filter and wash the paper free from chlorides. The filtrate should have a bulk of about 250 c.c. Heat to boiling, add barium chloride solution slowly with constant stirring, digest * As alcohol is free from sulphur the crucible is frequently heated with a spirit lamp. 102 DETERMINATION OF ACIDS. hot until the solution is clear; filter, wash free from chlorides, ignite and weigh. This method has been largely used for determining sulphur in sulphides, either minerals or laboratory products. It has the disadvantage that large amounts of alkali salts are introduced which contaminate the precipitate of barium sulphate. If the sul- phides lose sulphur on heating, the fusion mixture should consist of 4 parts of sodium carbonate, 8 parts of potassium nitrate, and 24 parts of pure and perfectly dry sodium chloride. As the reagents used may contain sulphur, a blank determination is nec- essary in careful work. For this purpose the fusion and precipi- tation must be carried out exactly as in the actual determination. The amount of barium sulphate obtained in this manner is deducted from that obtained in the analysis of the unknown. io6. Fusion of Sulphur Compounds with Sodium Peroxide. — Sulphur may also be oxidized to sulphuric acid by fusion with sodium peroxide. The finely ground material is mixed with five or six times its weight of powdered sodium peroxide, and the mixture fused in a nickel or copper crucible. The crucible must be covered with a closely fitting lid, and must be very cautiously heated, as the action is quite violent in the beginning. The flame of the Bunsen burner must be small and not touching the crucible. After a few minutes the heat is increased just sufficiently to keep the contents of the crucible in the state of fusion, which is main- tained for a few minutes. The crucible is then allowed to cool, placed on its side in a beaker, and water added from a wash-bottle while a watch-crystal is held over the beaker. When the fused mass is dissolved the crucible and lid are taken out with a clean pair of tongs and washed with water. The solution is filtered, and the residue washed with water containing a little sodium carbonate. If the residue is small^ and especially if it is free from iron, this filtration may be omitted. The filtrate is acidified with hydrochloric acid, and if silica was present in the original material the solution is evaporated to dryness finally on the water-bath, where it is heated for half an hour. The residue is treated with water and a few drops of hydrochloric acid and filtered. The sulphuric acid in the filtrate is precipitated and weighed as barium sulphate. As the sodium peroxide is not always free from sulphur, DETERMINATION OF SULPHUR. 103 a blank determination must be made as in the preceding method. 107. Oxidation of Sulphur Compounds with Fuming Nitric Acid. — Digestion with red fuming nitric acid is one of the best of the methods in which a liquid oxidizing substance is used. Most sulphides and sulphur compound? may be oxidized in an Erlen- meyer flask in the neck of which a small funnel is placed to pre- vent the too rapid escape of the fumes. With substances which are readily decomposed, giving off sulphuretted hydrogen, it is advisable to conduct the operation in a glass-stoppered bottle. A small, flat-bottomed glass tube or weighing-bottle, which may easily be inserted into the bottle, is provided. The material to be analyzed is weighed out and placed in this small tube, which is then lowered into the bottle so as not to spill the contents. The fuming nitric acid is now added from a graduate. 30 c.c. is sufficient for quantities of material which will yield not more than 1 gram of barium sulphate. The stopper is then inserted in the bottle and held firmly with the hand, while the bottle is tilted so that the acid comes in contact with the sulphide. When the first violent reaction has ceased shake the bottle a little, cool the bottle and contents by holding it under a tap of running water for a few minutes, and then cautiously remove the stopper so as to relieve the pressure without loss of material by spattering. The stopper is replaced in a slanting position to allow exit for the fumes, and the bottle is placed on the water-bath. If after half an hour's digestion particles of sulphur remain floating on the liquid, strong hydrochloric acid, potassium chlorate, or liquid bromine may be added in small portions, the potassium chlorate and liquid bromine being the most efficient. Care should be taken to prevent the fuming nitric acid or bromine from coming in contact with the hands, as serious burns are very quickly produced. The digestion on the water-bath is then continued until the sulphur is oxidized. 108. Decomposition of Insoluble Sulphates. — The complete disintegration of the materialis frequently indicated by a change in color or physical appearance. The nitrates of the metals are generally insoluble in the concentrated acid. Insoluble sulphates, of such metals as barium and lead, may also be produced. In the 104 DETERMINATION OF ACIDS. case of minerals, gangue, quartz, etc., will be left. The concentrated acid is diluted with water and digested for a few minutes to dis- solve soluble salts. If barium is present, the insoluble material must be filtered off, washed thoroughly, and fused with five parts of mixed sodium and potassium carbonates. The melt is dis- solved in water, and the insoluble carbonates filtered off and thor- oughly washed. The filtrate is acidified with hydrochloric acid and added to the nitric acid solution. If lead is present, the sul- phate of this metal may be decomposed by digesting with a solu- tion of acid sodium or potassium carbonate, filtering, washing, and treating the filtrate as before. 109. Removal of the Nitric Acid. — -The nitric acid solution, now containing all of the sulphur as sulphuric acid, is evaporated to dryness after the addition of a little sodium chloride, if alkalies have not already been introduced. The evaporation is repeated after the addition of dilute hydrochloric acid, until the nitric acid is entirely removed. The residue is dissolved in water, a few drops of hydrochloric acid are added, the solution is filtered if necessary, and the sulphuric acid determined as barium sulphate. no. Other Oxidizing Solutions Used for Sulphur Compounds. — The digestion with liquid bromine or aqua regia made by mixing one part of strong hydrochloric acid with three parts of strong nitric acid is carried out in a similar manner. The bromine may either be added gradually to the material mixed with a little nitric acid or to the dry material. A combination of carbon tetrachloride, liquid bromine, and nitric acid is also very satis- factory. These methods are largely used for the determination of sulphur ill pyrites or crude sulphur. (See Exercise 36, p. 172.) The dry material may also be mixed with powdered potassium CHLORATE in an Erlenmeyer flask, and concentrated hydrochloric acid added in small portions. Finally the flask is gently heated on the water-bath. The subsequent treatment is identical with that given after treatment with nitric acid. A very powerful oxidizing agent consists of a saturated solution of potassium chlorate in concentrated nitric acid. III. Oxidation of Sulphur Compoimds by Means of Chlorine. If lead is present, the oxidation of the sulphur by means .of chlo- rine in alkaline solution is advantageous, because the lead is pre- DETERMINATION OF NITROGEN. 105 cipitated as peroxide. Iron is also precipitated as ferric hydrox- ide. On passing the chlorine for a considerable time, however, the iron passes into solution as ferrate, producing a red tint. Heat- ing the liquid for a few minutes with powdered quartz serves to precipitate the iron. The finely powdered sulphide or crude SULPHUR is heated for some time with a dilute solution of caustic potash which is free from sulphuric acid. Free sulphur as well as the sulphides of arsenic and antimony dissolve in the caustic potash. Chlorine-gas is now led into the warm solution for some time. Most of the metals together with the gangue remain in the precipitate, which is filtered off and well washed. The filtrate is acidified with hydrochloric acid, again filtered if necessary to remove silica, and the sulphuric acid determined as barium sul- phate in the usual manner. 112. Determination of Telltirium and Selenium. — These ele- ments are precipitated from acid solutions by means of sulphur dioxide or sodium or potassium sulphite. If the solution contains nitric acid it must be evaporated to dryness after the addition of hydrochloric acid and sodium or potassium chloride. The solu- tion should be heated nearly to boiling, and kept at this temp- erature for a considerable time to insure complete precipitation. If the solution is dilute or a small amount of the element is present, it miist be allowed to stand for several days in a warm place. The precipitate is washed by decantation with water containing sulphur dioxide, and finally transferred to a weighed filter-paper, or to a Gooch crucible, and dried at 100°. DETERMINATION OF NITROGEN. Nitrogen unites with hydrogen to form a fairly strong base, and with oxygen and hydrogen to form one of the strongest acids. The methods used for determining the nitrogen when it exists in one of these forms are not generally applicable to the determina- tion of nitrogen in the other form. The methods most commonly used are volumetric. In this chapter the gravimetric methods, which are at times convenient, will be given. 113. Determination of Ammonia as Ammonium Chloride. — Ammonia may be weighed as ammonium chloride, which can be dried at 100°. A solution which contains nothing but free ammo- nia, ammonium chloride, or the ammonium salt of a very readily 106 DETERMINATION OF ACIDS. volatilized acid, such as carbonic acid or hydrogen sulphide, may be acidified with pure hydrochloric acid and evaporated to dryness on the water-bath in a weighed platinum dish. Many impurities, such as silica and the alkaline earth metals and the alkalies, do not interfere, since after weighing the residue in the dish the ammo- nium chloride hiay be volatilized by gentle heat, and the non- volatile residue weighed. Carried out in this manner the deter- mination of ammonia is quicKly and accurately made. It is a convenient method when so few determinations are required that the preparation of volumetric solutions is not warranted. 1 14. Determination of Ammonia as Ammonium Platino-chloride. — ^Ammonia may also be precipitated as ammonium platino- chloride which is washed with alcohol and dried at 100°, as de- scribed for potassium. The precipitate may also be decomposed by ignition into metallic platinum. The heat must be applied cautiously to the covered crucible, otherwise the escaping ammo- nium chloride will carry away platinum. 115. Determination of Nitric Acid by Means of Nitron. — No inorganic compound of nitric acid is known which is sufficiently insoluble to be used for the quantitative determination of this acid. Recently an organic base has been made which forms an insoluble nitrate.* This base is 1.4 diphenyl-3.5 endanilodihy- drotriazol, having the formula Cj^Hj^N^. For convenience in ordinary use and because it precipitates nitric acid, the name "nitron" has been given to the base. For use as a reagent a 10 per cent solution in 5 per cent acetic acid is made. At ordinary temperatures nitron precipitates nitric acid from solutions of 1 in 60,000, while at 0° C it win precipitate 1 part in 80,000. Unfortunately nitron also precipitates a number of other common acids, such as hydrobromic when present to the amount of 1 part in 800 of the solution, hydriodic acid (1 in 20,000), chromic acid (1 in 6000), chloric acid (1 in 4000), perchloric acid (1 in 50,000), sulphocyanic acid (1 in 15,000), and nitrous acid (1 in 4000). These acids must therefore be first removed from the solution. Hydrobromic and hydriodic acids may be removed by addirg chlorine water and boilirg out the liberated halogen, * M. Busch Ber., :«, 801 (1905). J. Chem. Sec, S8, II, 282, 418 flCOS). DETERMINATION OF NITROGEN. 107 or in the case of hydriodic acid it may be oxidized to the iodate. These acids may also be removed by means of silver sulphate. Nitrous and chromic acids may be reduced by means of hydrazine sulphate. To remove the nitrous acid the neutral concentrated solution (5-6 c.c.) is dropped slowly on one-fourth gram of finely powdered hydrazine sulphate. It is advisable to place the hydra- zine sulphate in a small flask in order to avoid loss by spattering and to facilitate cooling the mixture by means of cold water, as a small amount of nitric acid is formed if the mixture is allowed to become warm. The volume of the solution of the nitrate or nitric acid should be small because of the slight solubility of the precipitate. About 0.1 gram of nitric acid should be present in a volume of 80 to 100 c.c. which has been acidified with 10 drops of sulphuric acid. The solution is brought almost to the boiling-point and 10 to 12 c.c. of the nitron reagent is added. The hot solution gives a crystal- line precipitate. The beaker is then immersed in ice-water, where it is allowed to stand for one and one-half to two hours. The pre- cipitate is filtered off on a Gooch crucible which has been dried at 110°. It is transferred to the crucible by means of the mother- liquor and is washed with small portions of ice-water, using not more than 10 to 12 c.c, the precipitate being sucked dry after each addition of ice-water. The precipitate is then dried at 110° until constant, which requires about three-fourths of an hour. The precipitate has the formula CjjHijN^.HNOg, and contains 16.8% HNO3 or 14.4% N^Og. 1 16. Volatilization of NjOj by Fusion of Nitrates with Silica or Potassium Chromate. — Some very excellent methods of determin- ing nitric acid depend on its conversion into ammonia, nitric oxide, or nitrogen. Most of these methods will be described in the chapters on volumetric and gas analysis. It may also be deter- mined indirectly by heating the dry nitrate with a weighed amount of silica or potassium-acid chromate until all of the nitric acid is expelled. The loss in weight is due to expulsion of N2O5. The method can obviously not be applied to material which may lose anything but nitric acid. It has been extensively applied to sodium and potassium nitrates which can be dried at 130° and 100° respectively. The silica used must be finely powdered and ignited 108 DETERMINATION OF ACIDS. before being weighed. About seven times as much silica as nitrate must be taken. The material is intimately mixed and heated to redness in a platinum crucible for two to four hours or until constant weight is obtained. Instead of the silica a mixture of equal parts of neutral and acid potassium chromate may be used. This material is fused in a plat- inum crucible, allowed to cool, and finely powdered. About 3 grams of the chromate mixture are weighed out and placed in a platinum crucible. .8000 gram of the dried nitrate is added and thoroi' ghly mixed with the chromate by means of a platinum or glass rod. The crucible is heated gently so as to gradually bring the contents to a state of fusion. When no more acid fumes are evolved, the crucible is cooled in the desiccator and weighed. It is reheated until constant weight is obtained. CHAPTER IX. DETERMINATION OF CARBONIC, BORIC, PHOSPHORIC ACIDS. AND DETERMINATION OF CARBON DIOXIDE BY LOSS. Two general methods are in use for the determination of carbon dioxide. By the first method the carbonate is decom- posed by means of a strong mineral acid or other means, and the carbon dioxide expelled. The amomit present is ascertained by weighing the apparatus before and after the expulsion of the carbon dioxide. By the second method the gas evolved by the decomposition of the carbonate is led into a caustic-potash solu- tion, which is weighed before and after the absorption of the car- bon dioxide. The gain in weight represents the amount of carbon dioxide present. Various forms of apparatus have been devised for the deter- mination of carbon dioxide by loss when the carbonate is decomposed by means of acid. Two of these will be described, one of which may be purchased complete, while the other may be put together from the usual laboratory apparatus. 117. The Schrotter Apparatus shown in Fig. 16 may be purchased ready for use. A weighed amount of the carbonate is placed in the small bulb e and the stopper a a firmly inserted. The tube c is filled with dilute nitric or 10% hydrochloric acid, and h is about half filled with concen- trated sulphuric acid. The entire appa- ratus is then carefully weighed. By open- ing the stop-cock d, acid is allowed to flow into e at such a rate that the bubbles of gas can readily be coimted while passing through the concentrated sulphuric 109 Fig. 16. 110 DETERMINATION OF ACIDS. acid in b. When all of the acid has been allowed to flow out of c, and carbon dioxide is no longer evolved, the stop-cock d is closed and the solution in e is gently warmed on the water-bath or a hot plate to expel the carbon dioxide. The stop-cock d is now opened and air is sucked through the apparatus at a moderate rate until the carbon dioxide has been displaced. After being allowed to come to the atmospheric temperature, the apparatus is again weighed. The loss in weight represents the amount of carbon dioxide evolved. ii8. Sources of Error in the Determination of Carbon Dioxide by the Schrotter Apparatus. — Though the manipulation of this apparatus is very simple, it is impossible to obtain results accu- rate within more than J%. The error involved in weighing a glass apparatus of large surface is considerable. The gas which leaves the apparatus is dried over concentrated sulphuric acid, while that which enters is not dry. It is customary to ascertain when all of the carbon dioxide has been displaced by air by the taste of the gas which is sucked through by the mouth. The ordinary chemical tests for carbon dioxide cannot be used unless the air drawn through is freed from this gas. The accu- racy of the method would imdoubtedly be increased if the air drawn through were first passed through tubes containing soda-lime and concen- trated sulphuric acid. Care must also be taken to have the apparatus at the same temperature each time it is weighed. 119. Simple Laboratory Apparatus for Determining Carbon Dioxide. — Various forms of apparatus which can be made from the usual laboratory material have been suggested. A very simple device is shown in Fig. 17. The small flask A is closed with a two-holed rubber stopper, through which the thistle-tube passes as well as the bent-glass tube b, the latter ^^' ' also passing to the bottom of the short test- tube or small bottle B. The thistle-tube is closed with a glass rod, over the end of which a short piece of rubber tube has been DETERMINATION OF CARBON DIOXIDE. Ill slipped. The weighed amount of the carbonate is placed in the flask, and covered with a little water. The rubber stopper is firmly inserted, and a few cubic centimeters of dilute hydrochloric acid placed in the thistle-tube. If the carbonate is completely decomposed by sulphuric acid, it may be used instead of the dilute nitric or 10% hydrochloric acid. The tube B !s half filled with con- centrated sulphuric acid. The apparatus is tested for gas leaks by drawing out a little air from the bottle A, by applying suction to the tube a. The sulphuric acid rises in the tube h, and should remain at the same height for several minutes. The apparatus is now carefully weighed, and the carbonate decomposed by allow- ing the acid in the thistle-tube to flow into the flask in small por- tions. When all of the acid has been introduced, the flask A is gently warmed to expel the carbon dioxide. The carbon dioxide is displaced with air by applying suction at a. In careful work the air must be purified before entering the apparatus by passing through soda-lime and concentrated sulphuric acid, these absorp- tion-tubes being connected to the thistle-tube by means of a one- holed rubber stopper. 120. Determination of Carbon Dioxide by Fusion of Carbonates with Anhydrous Borax or Microcosmic Salt. — The carbon dioxide in all anhydrous carbonates may be determined by fusion with anhydrous borax. The carbon dioxide is completely expelled and its amount computed from the loss in weight. About 4 grams of vitrified borax are taken for 1 gram of the carbonate. The borax is placed in a platinum crucible, and heated with the Bunsen burner until the material is fused. It is allowed to cool in the desiccator and then weighed. As there is dan- ger of loss by the cracking of the borax if cooled suddenly, the flame of the Bunsen burner is turned down graduaUy before placing the crucible in the desiccator. As the borax retains water somewhat persistently, it is advisable to fuse and weigh it again. The anhydrous material may be kept in a state of fusion for one-quarter to one-half hour without loss in weight. The borax will be volatilized if heated with the blast-lamp. The weighed quantity of the dry carbonate is placed on top of the borax, and the latter again fused. As there is danger of loss by the cracking of the borax on being heated, the crucible should be covered until the borax is melted. It must be kept melted until the carbonate lis DETERMINATION OF ACIDS. is dissolved and no more bubbles escape. A few generally remaia in the fused mass, and cannot be expelled except by very pro- longed heating. After being again cooled in the desiccator, the crucible is weighed. The loss in weight is due to carbon dioxide. The borax glass may generally be removed by pressing the crucible between the thumb and fingers, thus loosening the fused mass, which will drop out on gently tapping the inverted crucible. Microcosmic salt may be substituted for borax in this deter- mination.* This salt melts very easily, needing only a small flame of the Bunsen burner. The heating is continued until all of the ammonia and water has been expelled. It is cooled and weighed and reheated until the weight is constant. The weighed carbonate is then added and the contents of the crucible heated very cautiously to avoid loss by the foaming caused by the escape of carbon dioxide. When no more carbon dioxide escapes, the crucible is cooled and weighed. It is again heated until the weight is constant. The action of the borax on the carbonate produces metaborates from the pyroborate, while the fused microcosmic salt is changed from the metaphosphate to the normal phosphate, according to the following equations: Na^B.O, + CaCOj = 2NaB02 + CaCBO^)^ + CO^ ; NaPOs +CaC03=NaCaPO,+C02. DIRECT WEIGHING OF CARBON DIOXIDE. The most reliable and accurate method of determining carbon dioxide is that in which the gas is absorbed and weighed. The carbonate is decomposed by sulphuric or hydrochloric acid. The gas evolved is dried, freed from hydrochloric acid, and absorbed in soda-lime or caustic-potash solution. The last por- tions of carbon dioxide are swept out of the apparatus by means of a stream of air which has been freed from carbon dioxide. 121. Apparatus. — Many forms of apparatus have been proposed and used for this purpose. The simplest and most serviceable form is that given by Fresenius. The characteristic part of this apparatus consists of a long glass tube which acts as a condenser; and also holds the drying and purifying material. In other forms * O. Lutz and A. Tschischikow, Jour. russ. phys.-chem. Gesellsch., 36, 1274. DETERMINATION OF CARBON DIOXIDE. 113 of apparatus the drying materials are placed in a series of U-tubes which are connected by rubber tubing, stoppers, etc. The difh- culties in making absolutely gas-tight joints are numerous, espe- cially when the permeability of rubber to carbon dioxide is consid- ered. The small number of joints in the apparatus of Fresenius gives it a decided advantage over other forms. In all forms of apparatus the carbonate is placed in a small flask closed with a two-holed stopper. Through one hole a dropping-funnel is passed by means of which the acid is introduced, and through the other the exit tube for the gas is passed. The end of the dropping- funnel is drawn out to a capillary, which is bent upwards to pre- vent loss of carbon dioxide. The arrangement of the apparatus of Fresenius is shown in Fig. 18. The Erlenmeyer flask should be of about 150 c.c. capacity. The glass tube should be about 70 cm. long and at least 1.2 cm. in diameter. It is inclined slightly so as to allow any water which condenses in its lower half to flow back into the flask. A Fig. 18. wad of glass wool or cotton is pushed down to the centre of the tube, the upper half of which is then filled with granulated cal- cium' chloride, large lumps being first introduced and then the finer material. If this calcium chloride has not already been 114 DETERMINATION OF ACIDS. saturated with carbon dioxide, a slow stream of the dry gas is passed through the tube for some time and then completely dis- placed with air. 122. Drying and Absorption of the Carbon Dioxide. — The Gbisslee caustic-potash bulbs are filled two-thirds full with a solution of 1 part of caustic potash and 2 parts of water. The straight calcium- chloride tube is filled with soda-lime and fused calcium chloride, so that the gas leaves the apparatus after pass- ing over the calcium chloride. Soda-lime is sometimes used for absorbing the carbon dioxide. It is placed in U-tubes, which are weighed before and after absorption of the carbon dioxide. Unless the soda-lime is freshly prepared and somewhat moist the absorption of large amounts of carbon dioxide is somewhat slow, so that two U-tubes are usually required. A calcium-chloride tube must also be attached and weighed. Whetlier caustic potash or soda-lime is used, an unweighed guard tube filled with calcium chloride must be attached. A small tube filled with soda-lime is also fitted to the dropping-funnel to remove carbon dioxide from the air which is drawn through the apparatus to sweep out the carbon dioxide. 123. Decomposition of the Carbonate. — Dilute sulphuric acid should be used for decomposing the carbonate, provided insoluble sulphates are not formed. Wlien sulphuric acid cannot be used, dilute hydrochloric or nitric acids may be employed. If sul- phides or sulphites are present, chromic acid or a chromate is added to prevent the evolution of sulphuretted hydrogen or sul- phur dioxide and sulphuric or nitric acid must be used to decom- pose the carbonate. When hydrochloric acid is used or chlorides are present, the carrying of the acid to the absorption bulbs may be prevented by substituting for part of the calcium chloride in the long drying-tube pieces of pumice-stone which have been saturated with copper-sulphate solution and dried at 125°. The method of setting up the apparatus and the details of manipulation are given in Chapter XVI, p. 183, in the determination of carbon diox- ide in dolomite. DETERMINATION OF BORIC ACID. 115 DETERMINATION OF BORIC ACID BY THE METHOD OF GOOCH. Considerable difficulty has been met with in determining boric acid, since no insoluble salt of this acid is known which can be washed and weighed accurately without the expenditure of a great deal of labor. The acid can be readily and completely separated from all other acids and bases by distillation with methyl alcohol. The distillate may then be evaporated to dryness and ignited with a known amount of calcium oxide. No boric acid is lost and the amount present is given by the increase in weight of the lime. The details of this method were first pub- lished by Gooch.* 124. Apparatus. — The apparatus recommended by Gooch con- sists of a retort made of a 150-c.c. or 200-c.c. pipette. One of the stems of the pipette, which should have an internal diameter of at least 0.7 cm., is bent into the form of a gooseneck, and is con- FiG. 19. nected by means of a rubber stopper with an upright condenser, which in turn is connected by means of a loosely fitting stopper * Am. Chem. Jour., vol. 9, p. 23. 116 DETERMINATION OF ACIDS. with an Erlenmeyer flask of about 250 c.c. capacity. The other stem of the pipette is bent at right angles and is connected by means of a short piece of rubber tubing with a small dropping- fimnel. An oil- or paraffine-bath is provided of such a size that the retort may be immersed in the liquid. 123. The Solution of the Borate, which should not contain more than 0.2 gram of Bfi^, is acidified with nitric or acetic acid.* Hydrochloric acid should be absent, as a considerable amount of this acid would pass over into the lime and be weighed with it unless the solution is neutralized and then acidified with acetic acid. It is better, however, to remove the chlorides with silver nitrate. The filtrate may be distilled directly or the excess of silver nitrate removed with sodium hydroxide or carbonate. Substances insoluble in nitric acid are fused with sodium carbon- ate. If fluorine is present, it is removed by precipitation as calcium fluoride in the water solution of the fused material. 126. Distillation of the Boric Acid with Methyl Alcohol. — ^From 1 to 2 grams of calcium oxide are placed in a platinum dish and heated over the blast-lamp until the weight is constant. It is then rinsed with a little water into the Erlenmeyer flask. A large excess of the acetic or nitric acid used to acidify the solution of the borate should be avoided, since these acids volatilize and neutralize the lime. Their salts also form viscous liquids on evaporation. A drop or two of phenolphthalein should therefore be added to the solution and a drop or two of the acid added after the alkaline color of the indicator has been removed. The smaller amounts of calcium oxide may be used when acetic acid is employed. The solution of the borate is transferred to the retort, which is lowered so that at first only the bottom touches the oil or paraflane, which is heated to 130°-140°. As the liquid distils off the retort is lowered until half of it is immersed and the distillation continued until the water is entirely removed. 10 c.c. of methyl alcohol is then introduced through the dropping- funnel and the solution distilled to dryness. This is repeated five times. If nitric acid has been used to acidify the solution, introduce into the retort 2 c.c. of water between the second and third and between the fourth and fifth additions of methyl alco- *The small amounts of these acids which distil over into the lime are ex- pelled during the subsequent ignition. DETERMINATION OF PHOSPHORIC ACID. 117 hol and distil to dryness. If acetic acid has been used, add a few drops of this acid with the fourth portion of methyl alcohol. 127. Weighing the Boric Acid. — After the sixth distillation with the methyl alcohol transfer the solution in the Erlenmeyer flask to the platinum dish in which the lime was ignited and weighed. Evaporate the solution to dryness, being careful to avoid spattering. Finally heat the dish with the blast-lamp imtil constant weight is obtained. The increase in weight is due to B^Oa. The residue in the retort may be tested for boric acid by means of turmeric-paper. If nitric acid has been used the nitrous acid should first be oxidized by means of bromine and the excess of the latter expelled. DETERMINATION OF PHOSPHORIC ACID. 128. Precipitation of Phosphoric Acid as Magnesiiun-ammon- ium Phosphate. — Phosphoric acid may be most accurately weighed as magnesium pyrophosphate. The properties of magnesium-ammonium phosphate as well as the magnesium pyrophosphate produced by ignition of this precipitate, have already been discussed under the Determination of Magnesium in Chapter VII, page 85. Only the alkali metals may be pres- ent in the solution, as the phosphates of the other metals are in- soluble in ammonia and woidd contaminate the precipitate. Only a moderate amount of ammonium salts should be present, and the solution should be slightly acid with hydrochloric acid. The phosphoric acid is precipitated by adding with constant stirring magnesia mixture in slight excess. This mixture is made by dissolving 55 grams of crystallized magnesium chloride and 70 grams of ammonium chloride in water, adding 300 c.c. of dilute ammonia or 88 c.c. of concentrated ammonia (sp. gr. 0.90) and diluting to 1 liter. After standing for several days, the solution is filtered as required for use. 10 c.c. of this solution is sufficient to precipitate 0.15 gram of P^Oj. If the phosphate solution is not distinctly alkaline after the addition of the magnesia mixture, it should be made so by the addition of ammonia. After stand- ing for at least six hours the precipitate is filtered off, washed with water containing one-tenth its volume of strong ammonia, and the precipitate ignited as directed for the determination of magnesium. 118 DETERMINATION OF ACIDS. 129. Separation of Phosphoric Acid as Phosphomolybdate. — ^If metals other than the alkahes are present, the phosphoric acid must be precipitated from a nitric-acid solution as phosphomolybdate. All of the phosphoric acid is precipitated while the metals are held in solution by the nitric acid, in which the phosphorous pre- cipitate is insoluble. The percentage of phosphoric acid in the precipitate is not constant, but varies within quite wide limits with the conditions of the precipitation. Although by rigidly fixing these conditions a precipitate of sufficiently constant com- position for direct weighing can be obtained, more reliable results are obtained by dissolving the precipitate in ammonia, repre- cipitating with magnesia mixture and weighing as magnesium pyrophosphate. 130. Removal of Silicic and Arsenic Acids. — Before precipi- tating the phosphoric acid as molybdate, silicic acid must be removed from the solution by evaporating to dr3niess and heating at 100° for some time. The dried residue is dissolved in water and nitric acid. Arsenic acid must also be absent except in traces, as this acid forms a precipitate with molybdic acid, which is very similar to the phosphomolybdate. It is also reprecipi- tated with the magnesia mixture. It may readily be separated from phosphoric acid by passing; hydrogen sulphide through the hot hydrochloric-acid solution. As hydrochloric acid inter- feres with the precipitation of the phosphoric acid as molybdate, it must be expelled by evaporation with nitric acid after precipita- tion of the arsenic, or the separation of the arsenic acid may be delayed untQ the phosphoric acid has been precipitated by means of magnesia mixture. This precipitate, which will generally con- tain only a part of the arsenic acid, is dissolved in a little hydro- chloric acid, the solution warmed to 70°, and hydrogen sulphide passed for a considerable time. The precipitate is filtered off and washed and the filtrate evaporated to a volume of 50 to 75 c.c. The solution is allowed to cool, a few drops of magnesia mixture are added, and then ammonia, with constant stirring until a slight excess is present. This precipitate is filtered off, washed, ignited, and weighed as usual. 131. Precautions. — ^The precautions* to be observed in pre- cipitating the phosphoric acid as molybdate are as follows: *Zeit. Anal. Ch., 28, 141. ~ DBTERMINATIOX OF mOSPIIOKIC ACID. J 19 Hydrochloric, sulphuric, and boric acids should be absent or present only in small amounts. Free nitric acid should be present only to the extent of 26 molecules to one molecule of the phosphoric acid. If more than 80 molecules of free nitric acid are present to one molecule of phosphoric acid, the precipitation is incomplete. The presence of ammonium nitrate hastens the precipitation. Twice as much molybdic acid must be present as is necessary to form the precipitate; that is, 24 molecules to 1 molecule of phosphoric acid. 132. Precipitation of the Phosphomolybdate. — The solution of the phosphate should be concentrated so that about 0.1 gram P2O5 is present in 25 c.c, 0.2 gram P^Oj being sufficient for a determmation. Enough of a concentrated ammonium-nitrate solution (750 gr. per liter) is added to constitute 15% of the total volume after sufficient molybdate solution has been added to have 80 c.c. present for each 0.1 gram of P2O5. The solution is then heated on the water-bath to 80°-90° for about ten minutes, and is then allowed to stand for about one hour, when the precipi- tate may be filtered off and washed. The addition of the ammo- nium nitrate may be omitted, but more of the molybdate solution must be added, and the solution heated on the water-bath to 60° for from four to six hours. In either case the filtrate should be tested for phosphoric acid by adding more molybdate solution and warming. A white precipitate of molybdic acid must not be mistaken for the yellow phosphate precipitate. 133. Washing the Precipitate. — Various solutions have been used for washing the precipitate, which is not absolutely insoluble, so that the minimum amount of wash-water should be used. Wagner recommends a solution made by dissolving 150 grams of ammonium nitrate in water, adding 10 c.c. concentrated nitric acid, and diluting the solution to one liter. Fresenius recom- mends the use of a solution made by adding to 100 c.c. molybdate solution 20 c.c. nitric acid (sp. gr. 1.2) and 80 c.c. water. A solu- tion made by diluting the molybdate solution with an equal vol- ume of water is also frequently used. 134. Direct Weighing of the Phosphomolybdate. — When it is desired to weigh the molybdate precipitate, the following slight variations in the method of procedure already given must be ob- 120 DETERMINATION OF ACIDS. served: Silica must first be completely removed from the solution of the phosphoric acid. After the molybdate has been added, the solution should be heated to about 40° to effect complete precipitation. Serious error results if it is heated to a tempera- ture near 100°. About four hours' digestion at 40° is generally sufficient. The precipitate is filtered off on a Gooch crucible and washed, first with a dilute solution of molybdate until the metals present are removed, and then with water containing 1% of nitric acid imtn the molybdic acid is removed. The precipitate is then dried in an air-bath heated to 120°. It contains 1.63% of phosphorus. 135. Reprecipitation of the Phosphoric Acid as Ammonium- magnesium Phosphate. — ^When it is desired to weigh the phos- phorus as magnesium pyrophosphate, the molybdate precipitate is washed with the solutions given in par. 133 imtil the metals present are removed. The beaker in which the precipitation was made, and from which the yellow precipitate need not be wholly removed, is placed under the fimnel, the paper is pierced, and the precipitate washed into the beaker with a fine jet of water. The paper is moistened with a little ammonia and washed with hot water, a few drops of ammonia being added if necessary to dis- solve the precipitate. Not more than 50 c.c. of wash-water should be used. If the precipitate does not dissolve on stirring the solu- tion, a little more ammonia should be added. Neutralize the solution with strong hydrochloric acid, and if the yellow phos- phomolybdate begins to precipitate, add ammonia until dis- solved. A white flocculent precipitate insoluble in ammonia is probably silica, and should be filtered off. The solution is again rendered slightly acid with hydrochloric acid and enough mag- nesia mixture added to precipitate the phosphoric acid present, using 10 c.c. for 0.1 gram P^Oj. The solution is neutralized by adding ammonia while stirring constantly. A moderate excess of ammonia is added and the solution is allowed to stand for at least four hours. The precipitate is filtered off, washed with dilute ammonia, ignited, and weighed as usual. Heating with the blast-lamp for ten minutes is advisable, in order to expel any molybdic acid which may have been carried down with the mag- nesia precipitate. ANALYSIS OF ALLOYS.* CHAPTER X. ALLOYS OF SILVER, LEAD, COPPER, BISMUTH, CADMIUM, AND TIN. 136. Reasons for Analysis of Pure Salts.— In the determina- tions given in the preceding chapters pure compounds were taken for analysis. This method of procedure serves two purposes. In the first place the exact composition of the compounds being known, the error in a given determination can easily be ascertained. In the second place no error can arise from the contamination of a precipitate with foreign material other than the reagents used. When the student has acquired proficiency in the comparatively simple processes of analyzing pure compounds, he must undertake the determination of elements when existing in combination with each other. It is evident that the determination of iron by precipitation with ammonia is impracticable if aluminium or chromic salts are present. 137. Difficulty of Complete Separation of Elements. — Although calcium is the only metal which can be quantitatively precipitated as oxalate, yet it is by no means true that this element can be separated as oxalate from a solution containing all other metals without carrying many of these down with it. This is due to the fact that most of the oxalates are soluble with difficulty. All of the sulphates except those of barium, strontium, and lead are quite soluble, and yet when such soluble sulphates as those of iron or aluminium are present the determination of sulphuric acid by precipitation with barium chloride presents difficulties * The alloys and minerals whose analyses are given in this and the following chapters have been selected so as to afford practice in the separations required in the analysis of commonly occurring substances. The elements mentioned as occurring in a given alloy or mineral are those commonly found. Some of these are at times absent, while others not listed may be present. 121 \-S2 ANALYSIS OF ALLOYS. which have been the subject of investigation by many ingenious worlcers for more than half a century. As the great majority of the substances with which the chemist is called upon to deal are complex rather than simple, a careful study of the separation of the elements is of the greatest importance. Only by the closest attention to details can success be attained in the analysis of complex substances. The importance of testing precipitates for impurities and the solution for unprecipitated portions of an element cannot be too strongly urged. Only in this manner can the accuracy of an analysis be assured. 13S. Limit of Accuracy in Analysis. — If a complete analysis is made the sum of all the constituents must be very close to 100%. A summation which is within .5% can generally be ob- tained if the analysis is conducted with care and reliable methods are used. In general the analysis of an unknown substance should be conducted in duplicate. If the duplicate results do not agree within .2 or at most .3%, a third analysis should be made. As the error of most determinations is at least .1%, it is unnecessary to calculate results to more than hundredths of per cent. As the error in each determination of the analysis of a given substance may be either plus or minus, the practice of dividing the difference between the summation and 100^ among tlie various determinations is not justifiable. It is in some cases possible to analyze a substance in such a manner that the results are accurate to the hundredth of a per cent. Such results may be computed to the .001 of a per cent. This practice is common in the analysis of metals. Large quan- tities of the metal are taken, so that considerable quantities of the impurities which are present in small amounts are obtained for determination. The results may then be accurate to the hun- dredth of a per cent. This does not imply a higher degree of accuracy in the determination of a given element than . 1 of a per cent. For example, if iron were present in copper to the extent of .5%, a determination of the iron which is accurate to .01% of the impure copper would represent an error of ^l of the amount of iron present in the copper. In giving the results of such analy- ses the percentage of the main constituent is obtained by ditiarence. so that the summation is exactly 100%. SEPARATION OF SILVER. 123 139. Separation of Silver from Other Metals.— Silver is readily separated from other metals by precipitation as chloride. Silver chloride being more insoluble in dilute nitric acid than in water the precipitate is washed with dilute nitric acid. This also serves to dissolve compounds of metals such as bismuth and antimony, Avhich tend to form insohible basic salts. As silver chloride is soluble in mercuric nitrate enough hydrochloric acid must be added to convert all of the mercury present into chloride. This metal must be present in the mercuric condition. The small amount of silver remaining in solution is subsequently precipitated with the mercury as sulphide and may be separated by volatilization of the mercury. If lead is present the solution must be dilute and hot, and no more hydrochloric acid added than just sufficient to precipitate the silver. The silver chloride must be thoroughly washed with hot water. If sodium acetate is added the lead is more easily kept in solution. EXERCISE 25. Analysis of a Silver Coin. Alloy of Copper and Silver. Prepare two Gooch crucibles for weighing silver chloride. They may be marked 1 and 2 with a blue pencil * and the marks made permanent by heating to redness with the blast-lamp. While the crucibles are drying on the hot plate the following operations should be carried out: 140. Dissolving the Alloy. — Clean a dime thoroughly by rubbing with sand or " Sapolio." Cut into pieces and weigh carefully a piece of about one- half gram. Place in a 200-c.c. beaker and add about 5 c.c. concentrated nitric acid and an equal bulk of water. Cover the beaker with a watch- crystal and warm on the water-bath until the metal is converted into nitrate. Dissolve the residue in water. A little gold is sometimes present and will remain undissolved at this point. The tclution of silver and copper is decanted into a 300-c.c. Erlenmeyer flask and the gold washed by decanta- tion five or six times, using about 25 c.c. of water each time. The gold is now transferred to a small fdter-paper, which is washed with hot water until free from silver. The moist paper is transferred to a weighed porcelain crucible, burned in the usual manner, and the gold weighed. As the amount of gold is always small, it is best to collect on the same paper the gold from all of the Aveighed portions of the dime, which are dissolved separately in nitric acid to obtain duplicate determinations of the silver and copper. The paper is washed free from silver after eacli precipitate has been transferred to it. *A pencil iikhIt' for ninrking on glass, porcelain, etc , is used for this purpose 124 ANALYSIS OF ALLOYS. 141. SUver. — The volumes of the solutions of silver and copper in the Erlenmeyer flasks should not much exceed 200 c.c. By proper manipula- tion this result can easily be secured. Heat nearly to boiling and precipi- tate the silver by adding hydrochloric acid, drop by drop, with vigorous shaking of the flask. Add only a slight excess of hydrochloric acid and digest the precipitate on the water-bath with occasional vigorous shaking until the solution is very nearly clear. Decant the clear liquid through the weighed Gooch crucible and wash the precipitate two or three times by decantation with about 50 c.c. of hot water to which a few drops of nitric acid have been added. The precipitate should be digested each time for a few minutes on the water-bath. Finally transfer the precipitate to the crucible and wash vfith hot water containing a little nitric acid until the wash-water gives no precipitate with hydrogen sulphide. Dry the pre- cipitate on the hot plate and weigh. 142. Copper. — Ignite and weigh a Rose crucible. Heat the copper solution, the volume of which should be about 400 c.c, nearly to boiling and saturate with hydrogen sulphide. Filter immediately and wash with water containing hydrogen sulphide. Test the filtrate and wash-water for copper by passing hydrogen sulphide. Dry the precipitate, detach from the paper, burn the latter, and ignite the precipitate in the Rose cruci- ble in a stream of dry hydrogen after the addition of sulphur. 143. Division of Coin Volumetrically. — If there is available a 250-c.c. flask and a 50-c.c. pipette which have been carefully calibrated against each other, the division and weighing of the coin may be carried out in the following manner. The coin is weighed after thorough cleansing. It is then placed in a 200-c.c. beaker and treated with 10 c.c. concentrated nitric acid and an equal bulk of water. The beaker is covered with a watch- crystal and warmed on the water-bath until the metal is converted into nitrate. The watch-crystal is rinsed off and the solution is evaporated to dr3nQess on the water-bath. The residue is dissolved in water, the gold, if present, filtered off and weighed, the filtrate and washings being allowed to flow into the 250-c.c. flask. Finally the solution in the flask is diluted to the mark with water, thoroughly shaken, and 50-c.c. portions taken out with the pipette. The analysis of these portions is carried out as already directed. The duplicate determinations of silver and copper should agree Avithin about .2%. American silver coins contain very nearly 90% of silver and 10% of copper. 144. Economizing Time. — In carrying out this and subse- quent exercises the work should be carefully planned to econo- mize time. Crucibles should be prepared and weighed during the intervals when precipitates are digesting, filtering, or drying, so that the latter may be transferred and weighed without delay. SEPARATION OF TIN. 125 Two precipitations and filtrations can be carried out at the same time just as readily as one. As experience is gained, a greater and greater number of analyses can be carried on simultaneously. Careful and systematic marking of beakers, funnels, etc., is abso- lutely necessary. All weights must be recorded in the note-book under systematic headings and numbering of duplicate analyses. 145. Separation of Tin as Stannic Oxide. — ^Tin is usually sepa- rated from other metals as the dioxide SnO^. On treating an alloy containing tin with nitric acid, most of the metals form nitrates, while the tin is converted into the dioxide, which remains undissolved on treating the residue with water. If antimony or ARSENIC is present more or less of these elements, depending on the amount present, will remain with the tin. Besides these ele- ments, which one would expect from the nature of their oxides to remain imdissolved by water or dilute acids, small portions of copper, lead, bismuth, iron, manganese, and zinxi remain with the stannic oxide and cannot be removed by digestion with dilute nitric acid. In the absence of more than traces of arsenic and antimony the stannic oxide is filtered from the nitric-acid solution, washed with dilute nitric acid, ignited and weighed. 146. Purification of the Stannic Oxide. — ^The weighed precip- itate is then fused in a porcelain crucible with six times its weight of equal parts of sodium carbonate and sulphur, or with the same weight of dry sodium thiosulphate. By this treatment the metals present are converted into sulphides. On treating the fused mass with hot water the sulphides of tin, arsenic, and antimony and a small amount of copper if this element is present dissolve as thio salts of sodium, while the sulphides of the other metals remain and may be filtered off. On treating the insoluble sulphides with warm dilute hydrochloric acid, cop- per sulphide remains undissolved and may be ignited and weighed as CuO. In the filtrate the lead may be precipitated with sul- phuric acid and weighed as sulphate. The iron is .precipitated with ammonia, while the zinc is separated from the manganese by passing hydrogen sulphide through the slightly acid solution. The manganese may be precipitated with hydrogen peroxide and weighed as MujOi. The weight of any or all of these metals computed as oxide is deducted from the weight of the impure stannic oxide. 126 ANALYSIS OF ALLOYS. If several of these metals are present in the alloy it is perhaps easier and fully as accurate to omit weighing the impure stannic oxide, but fuse it immediately as already described for its purifi- cation. The insoluble sulphides are dissolved in a little nitric acid and added to the nitric acid solution of the alloy. The solu- tion of sodium sulphostannate is nearly decolorized by boiling with the addition of pure caustic soda and hydrogen peroxide. The tin is precipitated as sulphide by acidifying with hydro- chloric acid. The precipitate is filtered off and washed with water containing ammonium acetate and a little acetic acid. It is then treated with a little concentrated ammonium carbonate solution to dissolve arsenic sulphide. After washing with water the sulphides of tin and antimony are dissolved in warm concentrated hydrochloric acid, j^ny antimony which is present may be pre- cipitated by digesting the diluted solution with metallic iron. The tin in the filtrate from the antimony is precipitated with hydrogen sulphide and washed. The precipitate is dried and detached from the paper, which is burned. It is then placed in a porcelain crucible and gently heated until sulphur dioxide ceases to be given off. It is then strongly heated after the addition of ammo- nium carbonate and weighed as stannic oxide, SnO^ 147. Separation of Lead as Sulphate. — Lead is most frequently separated from other metals by precipitation as sulphate. The nitric acid solution is evaporated with excess of sulphuric acid until the nitric acid is completely expelled, which is indicated by the evolution of dense white fumes of sulphuric acid. On dilu- tion with water the lead separates out almost completely. In the absence of much bismuth or iron, a more complete precipita- tion of the lead may be effected by the addition of an equal bulk of alcohol to the dilute sulphuric acid solution. Besides barium and STRONTIUM sulphates which may contaminate the precipitate, BISMUTH may be present as an insoluble basic sulphate. To keep the bismuth m solution considerable sulphuric acid must be present and the precipitate must be washed with dilute (10%j sulphuric acid, which is finally removed by washing with dilute alcohol 148. Separation of Lead as Chloride from Bismuth. — The precipitate of lead sulphate may be tested for bismuth by warming with concentrated hydrochloric acid. After allowmg to cool, SEPARATION OF LEAD. 127 50 c.c. absolute akohol is added to the solution of lead chloride, the volume of which should be about 5 c.c. The bismuth chlo- ride dissolves in the alcohol and may be filtered off from the pre» cipitated lead chloride. A small amount of the lead dissolves in the alcohol with the bismuth. The latter may be precipitated by largely diluting the alcoholic solution. This method may also be used for separating the lead and bismuth in the nitric- iicid solution of lead and bismuth alloys. Before the addition of the alcohol the nitric acid is expelled by evaporating twice with 20 c.c. concentrated hydrochloric acid. The lead chloride may be collected on a Gooch crucible, washed with alcohol, and dried for three hours at 150° and weighed. About 1 mg. of lead chloride remains in the alcoholic solution. The bismuth is precipitated as BiOCl by largely diluting the alcoholic solution and neutralizing most of the hydrochloric a3iJ with ammonia if a large excrss is present. It may be washed with water containing a few dro;;s of hydrochloric acid, dried at 110° and weighed. 149. Separation of Lead from Antimony and Barium. — If con- siderable antimony is present the alloy is dissolved in a rather small amount of nitric acid (4 c.c.) with the addition of tartaric acid (10 grams) and water (5 c.c). The lead is precipitated by the addition of 4 c.c. concentrated sulphuric acid and diluting the solution to about 250 c.c. If baritim is present, as when minerals are analyzed, so tl' it the sulphate of lead is contaminated with barium sulphate a separation may be effected by digesting the precipitate with ammonium carbonate solution. The sulphate of lead is converted into carbonate, while the barium sulphate is unaffected. The precipitate is filtered off and washed with ammonium carbonate solution, then with water until the wash-water no longer con- tains sulphates. The lead carbonate is then dissolved with dilute acetic or nitric acid. In this method of separation a little lead is lost on account of the slight solubility of lead carbonate in solu- tions of ammonium salts. It may be recovered by passing hydrogen sulphide through the ammonium carbonate solution. 128 ANALYSIS OF ALLOYS. EXERCISE 26. Analysis of Soft Solder. Alloy of Lead and Tin, Generally containing . Small Amounts of Arsenic, Antimony, Iron, and Zinc. 150. Solution of the Alloy. — One gram of the alloy is weighed out and transferred to a beaker of about 500 c.c. capacity. 10 c.c. of concentrated nitric acid and 5 c.c. of water are added. The beaker is covered with a watch-crystal and heated on the water-bath imtil the alloy is completely decomposed and the nitrous fumes are entirely expelled. 100 c.c. of water is added and the solution boiled for five minutes and allowed to settle for one hour. The stannic oxide is filtered off and washed with hot water. The moist precipitate may be introduced into a weighed porcelain crucible, the paper burned in the usual manner, and finally heated to redness for ten minutes. 151. Tin. — When the precipitate has been brought to constant weight, it is fused with six times its weight of a mixture of equal parts of sulphur and sodium carbonate. The fused mass is dissolved in hot water and the solution filtered. The insoluble sulphides are washed with hot water and treated with a little dilute hydrochloric acid and thepaper washed with water. If COPPER is present it will remain on the paper and the small amount present may be weighed as CuO after burning the paper in a porcelain cruci- ble and igniting the precipitate. The lead is precipitated by the addition of a few drops of sulphuric acid and 25 c.c. of alcohol to the solution, which should not exceed 50 c.c. After standing one hour, the precipitate is filtered off on a Gooch crucible, washed with alcohol, dried on the hot plate, and weighed. The filtrate is evaporated until the alcohol is completely expeUed. Any ikon present is precipitated with ammonia and weighed. Hydrogen sulphide is passed through the filtrate to precipitate any zinc present, which is filtered off. The filtrate from the insoluble sulphides will contain the tin as a thiostannate and part of the antimony present in the alloy as a thioantimonate. The solution is boiled after the addi- tion of caustic soda and hydrogen peroxide until it is nearly decolorized. On acidifying and passing hydrogen sulphide both metals are precipi- tated as sulphides. If antimony is present the metals should be separated by the method given in section 178, page 141. The weight of the im- purities found, computed as oxides, is deducted from the weight of the stannic oxide. 152. Lead. — The filtrate from the stannic oxide is transferred to a porce- lain dish, 5 c.c. concentrated sulphuric acid added, and evaporated until fumes of sulphuric acid are evolved. Cool the dish by floating it in cold water and add cautiously 75 c.c. of water. Stir thoroughly and add 25 c.c. of alcohol. Allow the solution to stand for at least one' hour, filter off the lead sulphate on a weighed Gooch crucible, wash with alcohol untU free from acid, dry on the hot plate, and weigh. ANALYSIS OF ROSE'S METAL. 129 153. Arsenic and Antimony. — ^The alcohol is completely expelled from the filtrate by evaporation and any arsenic present precipitated by passing hydrogen sulphide. If this precipitate is of an orange color instead of pure yellow, antimony is present. It should be filtered off and washed with water containing a little hydrochloric acid until free from iron and hydrogen sulphide. It is then washed with small portions of concentrated ammo- mum carbonate solution until the arsenic sulphide is entirely dissolved. The arsenic is reprecipitated by acidifying the solution with hydrochloric acid and passing hydrogen sulphide. It is filtered off on a Gooch crucible and washed with water containing hydrogen sulphide and a little hydro- chloric acid. The water is removed by alcohol and the precipitate digested with carbon disulphide untU sulphur is entirely removed. The arsenic sulphide is dried at 100° and weighed. If antimony is absent the treatment with ammonium carbonate is omitted, the precipitate being filtered off on a Gooch crucible, washed, dried, and weighed. If antimony is present it is ignited and weighed as directed in sectionl79, page 141. 154. Iron.— A few drops of bromine water are added and the solution is boiled to oxidize the iron and to expel the hydrogen sulphide. The iron is then precipitated by making the filtrate alkaline with filtered ammonia and warming for a few minutes. It is filtered off on a small paper and dissolved by adding a few drops of dilute hydrochloric acid. The paper is washed with about 75 c. c. of water in small portions. The iron is reprecipitated and filtered on the same paper after moistening with a few drops of ammonia. After washing free from chlorides, the moist paper is transferred to the weighed platinum crucible and ignited. 155. Zinc. — Hydrogen sulphide is passed into the combined filtrates to precipitate any zinc present, which is filtered off, washed, and weighed as sulohide. EXERCISE 27. Analysis of Rose's Metal. AUoy of Lead, Bismuth, and Tin, Generally containing Small Amounts of Copper, Arsenic, Antimony, Iron, and Zinc. One gram of the metal is weighed out and decomposed with nitric acid as directed in Exercise 26. The stannic oxide is weighed and the impurities determined as directed in the same exercise. 156. Lead. — To the filtrate from the stannic oxide containing the ni- trates of lead and bismuth, 5 c.c. concentrated sulphuric acid is added. The solution is evaporated in a porcelain dish until sulphuric-acid fumes are given off. The dish may be placed on the hot plate, sand-bath, or wire gauze and the liquid heated to just below the boiling-point to avoid spatter- mg. When the acid becomes concentrated, the heat may be somewhat increased. The hot concentrated solution is diluted by slowly pouring it with constant stirring into about 100 c.c. of water and digested hot for about half an hour with occasional stirring. The lead sulphate is then filtered off on a 130 ANALYSIS OF ALLOYS. Gooch crucible, washed with 10% sulphuric acid until the wash-water no longer gives a precipitate on making it alkaline with ammonia, adding ammonium carbonate, and warming. The sulphuric acid is then washed out with alcohol. The precipitate is dried and weighed. It is tested for a possible contamination with bismuth as follows : It is dissolved in 5 to 10 c.c. of warm concentrated hydrochloric acid and 50 c.c. of absolute alcohol are added to the solution. After standing for a few moments, the solution, containing the bismuth as chloride, is filtered off. By nearly neutralizing with ammonia, and largely diluting with water, the bismuth is precipitated as oxychloride and may be washed with water containing a few drops of hydrochloric acid, dried, and weighed. 157. Bismuth. — In the filtrate from the lead sulphate, the bismuth is precipitated by just neutralizing wdth filtered ammonia, adding a few drops of ammonium carbonate, and warming the solution gently for about fifteen minutes. The precipitate is filtered off and washed a few times with water. To free the precipitate from a small amount of basic sulphate it is dissolved in a small amount of dilute nitric acid and reprecipitated. The precipitate is washed with water containing a little ammonium nitrate and dried. It is removed from the paper as completely as possible and placed on a watch-crystal. The paper is replaced in the funnel, moistened with a few drops of dilute nitric acid, and washed with small amounts of warm water. The wash-water is evaporated to dryness in a fairly large weighed porcelain crucible, and the residue ignited until the nitric acid is com- pletely expelled. The main portion of the precipitate is now added, heated with the Bunsen burner and weighed as Bi.,0„. 158. Copper. — If copper is present in the alloy, it will be contained in the two filtrates from the bismuth precipitate. Combine these filtrates, acidify with hydrochloric acid, and concentrate to a convenient bulk. Pass hydrogen sulphide through the warm solution, filter, and wash with water containing hydrogen sulphide. Even if copper is absent, a small black precipitate of bismuth sulphide will be obtained at this point because of the slight solubiUty of the bismuth hydroxide or carbonate. The precipitate may be tested for bism.uth by treating with a little dilute hydrochloric acid and diluting the filtrate. A white precipitate indicates bismuth. The copper sulphide being insoluble in diiu'ce hydrochloric acid remains on the paper and may be ignited together with the paper and weighed as oxide. If AESENic, ANTIMONY, IRON, or ZINC are present they are separated and determined by the methods given in Exercises 26 or 28. 159. Separation of Copper and Cadmium. — Cadmium may be separated from copper by passing hydrogen sulphide through or adding ammonium sulphide to a solution of the two metals in l)otassium cyanide In such a solution copper sulphide is soluble, whil(! cadmium sulphide is not. Silver, bismuth, and lead must be absent, as these metals would be precipitated with the cad- ANALYSIS OF WOOD'S METAL. 131 mium. Cadmium may also be separated electrolytically from copper. The latter metal may be deposited from a solution con- taining 5% of acid by a current having a tension not exceeding 1.85 volts. Under these conditions cadmium is not precipitated. EXERCISE 28. Analysis of Wood's Metal. Alloy of Lead, Bismuth, Tin, and Cadmium, Generally containing Small Amounts of Copper, Arsenic, Antimony, Iron, and Zinc. One gram of the metal is weighed, dissolved in nitric acid, and the stannic OXIDE weighed and purified by the method given in Exercise 26. The filtrate from the tin is evaporated to dryness on a water-bath. The nitrates are converted into chlorides by evaporating twice on the water-bath to a small bulk after the addition of 20 c.c. of concentrated hydrocloloric acid. i6o. Lead. — After coding, 25 c.c. absolute alcohol are added. The mixture is stirred and after standing some time the chloride of lead is fil- tered off on a Gooch crucible, and washed with an ice-cold mixture of 4 parts of 95% alcohol and 1 part of concentrated hydrochloric acid. It is dried on the hot plate or at 150° for three hours and weighed. i6i. Bismuth. — ^The filtrate is diluted with about one-half liter of water and nearly neutralized with ammonia (about 40 c.c. of dilute ammonia will be required). After standing twenty-four hours the bismuth oxychloride is filtered off on a Gooch crucible, washed with water containing a few drops of dilute hydrochloric acid, dried at 110°, and weighed as BiOCl. The bismuth may also be precipitated as bismuth hydroxide by vola- tilizing most of the alcohol, neutralizing with ammonia, and warming gently. If IRON is present this precipitate will be reddish. In that case it is best to dissolve it in hydrochloric acid • and precipitate the bismuth as oxy- chloride. The bismuth hydroxide is ignited and weighed as oxide, Bi203. 162. Cadmium. — The filtrate from the bismuth oxychloride is evapo- rated to a bulk of 200 or 300 c.c. If the bismuth has been precipitated by means of ammonia, the filtrate is first acidified \vith hydrochloric acid and evaporated to a moderate bulk. The solution is saturated mth hydrogen sulphide and the precipitate filtered off and washed with water containing hydrogen sulphide. If the cadmium sulphide is dark colored or black, traces of lead or bismuth sulphides may be present because of incomplete separations, or copper may have been present in the alloy. Any arsenic v/liich may have been in the alloy or a trace of tin or antimony will also be present in this precipitate. 163. Arsenic, Antimony, and Tin.— It should be tested for these three elements by pouring over it a few drops of warm potassium or sodium sulphide and washing two or three times with warm water, being careful 132 ANALYSIS OF ALLOYS. to stir up the precipitate with the stream of water from the wash-bottle. A precipitate formed on acidifjang the filtrate indicates the presence of arsenic, antimony, or tin. If the characteristic orange color of antimony is absent, the supernatant liquid should be decanted and the precipitate warmed with a little concentrated hydrochloric acid. If it dissolves com- pletely, arsenic is absent and the tin may be reprecipitated by diluting and passing hydrogen sulphide. After washing, the moist precipitate with the paper may be burned and the sulphide of tin converted into oxide by ignition. If arsenic or antimony is present, it may be determined as directed in the following exercise. 164. Separation of Copper and Cadmium. — To dissolve out any copper which may be present with the cadmium sulphide, a few drops of potassium cyanide should be poured over the precipitate. It should be thoroughly stirred up with water and washed a few times. If a considerable amount of copper is present, the bulk of the precipitate should be transferred to a beaker by washing out the .paper while still in the funnel with a stream of water. The remainder of the precipitate on the paper is dissolved by washing with a little warm dilute nitric acid. The paper is then thor- oughly washed with small portions of hot water. The washings are allowed to flow into the beaker containing the main portion of the precipitate. The beaker is warmed and more nitric acid is added if necessary to dissolve the precipitate. The solution is neutralized with sodium carbonate and a slight excess of potassium cyanide added. A small white precipitate at this point may be lead or bismuth carbonates, which should be filtered off and determined. On passing hydrogen sulphide through the filtrate, the cadmium is precipitated as sulphide and may be filtered off on a Gooch crucible and washed with water containing a little hydrogen sulphide. It is finally washed with pure water and the free sulphur extracted by washing with alcohol and then with carbon disulphide. The precipitate is dried at 100° and weighed. 165. Copper. — The filtrate from the cadmium sulphide contains the copper and is acidified* with sulphuric acid and a little nitric acid and evaporated to fumes. The residue is dissolved in water, filtered if necessary, and the copper precipitated as sulphide. If it is small in amount it may be ignited and weighed as oxide. If considerable copper is present, it must be ignited with sulphur in a stream of hydrogen and weighed as cuprous sulphide, CujS. When much copper is present, it is better to determine it electro- lytically. 166. Separation of Iron and Zinc. — The filtrate from the first precipita- tion with hydrogen sulphide (§ 162) contains any zinc or iron which may have been present. These metals may be determined as described for the * This should be done vmder a hood with good draught to avoid any possi bility of inhaling the very poisonous hydrocyanic-acid fumes. ANALYSIS OF WOOD'S METAL. 133 similar filtrate in Exercise 26. Tliey may also be separated in the follow- ing manner : The solution is boiled to expel hydrogen sulphide, neutrahzed with ammo- nia, and acidified with acetic acid. Hydrogen sulphide is passed for some time and the solution allowed to stand for several hours. The clear liquid is carefully decanted through a filter-paper, and after replacing the beaker containing the clear filtrate with another beaker, the sulphide of zinc is brought on the paper and washed with water containing ammonium acetate and acetic acid. The precipitate is dissolved in a little dilute nitric acid and the paper washed with hot water. The solution of the zinc is evaporated to dryness in a weighed porcelain crucible, ignited finally over the blast-lamp to decompose any zinc sulphate which may have been formed, and weighed as oxide. The filtrate is boiled to expel the hydrogen sulphide. A little nitric acid is then added to oxidize the iron, which is precipitated with ammonia and weighed as oxide. A very convenient method of oxidizing the iron and removing the hydrogen sulphide is by the use of bromine water. The bromine should be added until the solution is colored, indicating com- plete oxidation of the iron and the presence of an excess of bromine. If a solution of bromine in concentrated hydrochloric acid is used a few drops will suffice and the solution will not be diluted to any extent. If manganese is to be removed together with iron, the presence of an excess of bromine is advantageous ; otherwise it must be boiled out. As the bromine oxidizes hydrogen sulphide in the cold, the excess of the latter need not be boiled out. CHAPTER XI. ANALYSIS OF ALLOYS CONTAINING ARSENIC, ANTIMONY, AND TIN. SEPARATION OF ARSENIC, ANTIMONY, AND TIN. 167. F. W. Clarke's Method. — A large number of methods have been proposed for the separation of arsenic from antimony and tin. One of the best is that of F. W. Clarke.* It is based on the fact that stannic sulphide is not precipitated by hydrogen sul- phide from a boiling solution of oxalic acid. The sulphides of arsenic dissolve slightly, and the sulphide of antimony still more so, in boiling oxalic-acid solutions, but both of these elements are completely reprecipitated by hydrogen sulphide. Free mineral acids must be absent from the solution. At least* 20 parts of crystallized oxalic acid must be present for 1 part of tin, and the solution should be diluted to 125 c.c. for each 0.1 gram of anti- mony present. Hydrogen sulphide is passed through the boiling solution for at least one-half hour. The precipitate must be filtered off immediately, as stannic sulphide separates out on standing. The precipitate generally contains a little stannic sulphide. After filtering off and washing two or three times it is dissolved m a little ammonium sulphide and the solution poured into excess of hot strong solution of oxalic acid. Hydrogen sul- phide is passed through the boiling solution for ten minutes. The precipitate is now free from tin, and is filtered off and washed with hot water. The oxalic-acid solution of the tin is evaporated down with the addition of sulphuric acid until the oxalic acid is decomposed. The tin may then be precipitated with hydrogen sulphide and weighed as oxide. 168. Separation of Arsenic and Antimony. — The sulphides of arsenic and antimony may be separated by treatment with a saturated ammonium-carbonate solution which dissolves the arsenic *Chem. News, 21, 124. 131 SEPARATION OF ARSENIC , ANTIMONY, AND 7 IN. 135 easily and the antimony only slightly. No more ammonium- carbonate solution than necessary to dissolve the arsenic sulphide should be used. The precipitate, which must first be washed entirely free from hydrogen sulphide, should be treated with small portions of the carbonate solution until no more arsenic is dis- solved, as indicated by the absence of a precipitate on acidifying the filtrate. The arsenic is completely precipitated by acidifying with hydrochloric acid, warming, and passing hydrogen sulphide. This method of separating arsenic and antimony is especially applicable if the amount of arsenic present is small. A better separation of the arsenic and antimony inay be effected by dissolving the sulphides in concentrated hydrochloric acid with the addition of small amounts of potassium chlorate. The solution is warmed to expel the chlorine, but not to boiling, so as not to melt the sulphur which is generally liberated. It is filtered through asbestos, which is washed with concentrated hydrochloric acid. It is cooled in ice-water and hydrogen sulphide passed for one hour. The arsenic is precipitated as As^Sj and is washed with strong hydrochloric acid and then with hot water. Antimony remains in solution as trichloride. This method also serves for the sepa- ration of ARSENIC and tin, the latter behaving like antimony. According to Neher,* for this separation the solution should contain 1 volume of water to 2 volumes of concentrated hydro- chloric acid (sp. gr. 1.20). If a stronger acid is used, the precipi- tation of the arsenic is hindered. The arsenic is best dissolved and reprecipitated as magnesium ammonium arsenate. If hydrogen peroxide is at hand which is free from phosphoric acid, aluminium, etc., the arsenic sulphide maybe dissolved in a warm mixture of ammonia and hydrogen peroxide. A very excellent solvent is a solution of sodium peroxide m water. The arsenic acid may then be precipitated by means of magnesia mixture. The arsenic sulphide may also be dissolved in strong hydrochloric acid and potassium chlorate or in fuming nitric acid. These oxidizing solutions should be warmed, but not high enough to melt the sulphur which usually separates out. The latter becomes very nearly white when all of the arsenic has been dissolved out. If the sulphur is large in amount or otherwise * Zeit. f. Anal. Chem., 1893, p. 45. 136 ANALYSIS OF ALLOYS. troublesome, liquid bromine may be added in small portions to dissolve it. This method of precipitation also serves to separate ARSENIC from ANTIMONY. Tartaric acid is added to prevent the precipitation of the antimony. 169. The Separation of the Antimony and Tin in the hydro- chloric-acid filtrate from the arsenic is effected as follows: The hydrogen sulphide is decomposed by the addition of a little potas- sium chlorate or bromine. The antimony is precipitated by digesting the somewhat diluted solution on the water-bath with the addition of pure iron. Iron produced by reduction in hydro- gen is best, and in any event it must be free from phosphorus. The precipitation of the antimony is complete in about one-half hour. It forms a black powder. The tin is reduced to the stan- nous condition. The solution is heated until the iron is nearly dissolved. The antimony is then filtered off on an asbestos filter over which a little iron has been sprinkled, and is washed with boiled water containing considerable hydrochloric acid. The TIN is precipitated by passing hydrogen sulphide through the filtrate, which for this purpose is diluted, and the bulk of the acid neutralized with ammonia. The stannous sulphide is washed with water containing hydrogen sulphide and a few grams of ammonium sulphate. It is dried, transferred to a weighed cruci- ble, ignited, and weighed as stannic oxide. The ANTIMONY is dissolved in hydrochloric acid with the addi- tion of a little potassium chlorate. The solution is diluted after the addition of tartaric acid and the antimony precipitated as sulphide and weighed as oxide. 170. H. Rose's Ssparation of Antimony from Arsenic and Tin. — This method of separation is based on the insolubility of sodium metantimonate. If the substance is metallic, it may be con- veniently oxidized by treatment with concentrated nitric acid in a porcelain evaporating-dish or crucible. The material is dried on the water-bath and transferred to a silver crucible, the porcelain dish being rinsed with caustic-soda solution. The con- tents of the crucible are evaporated to dryness, eight times the bulk of the mass of solid caustic soda added, heated cautiously to avoid spattering until all water is expelled and then fused for some time. The mass is allowed to cool, and is treated with hot water until the fused material is entirely disintegrated. One-third the SEPARATION OF ARSENIC, ANTIMONY, AND TIN. 137 volume of alcohol is added and tlie solution allowed to stand for twenty-four hours with frequent stirring. If the arsenic, antimony, and tin are obtained by separation as sulphides, they may be dissolved in solution of sodium sulphide and the cold-saturated solution treated with small portions of sodium peroxide until the solution is colorless and oxygen is copiously evolved on the further addition of the peroxide. The solution is boiled and after cooling one-third the volume of alco- hol is added. The solution is then allowed to stand for twenty- four hours. The precipitate obtained by either method is washed, first with dilute alcohol composed of equal volumes of alcohol and water and then with a mixture of one volume of water and three volumes of alcohol. A few drops of sodium carbonate solution are added to each, of the washing fluids. If much tin is present, it is advisable to fuse the metantimonate a second time with soda. The antimony is dissolved in hydrochloric acid containing a little tartaric acid, precipitated with hydrogen sulphide, and weighed as oxide. The filtrate containing the arsenic and tin- is acidified with hydrochloric acid and evaporated until all of the alcohol is ex- pelled. The solution is warmed and hydrogen sulphide passed untU precipitation is complete. The arsenic and tin may then be separated by one of the methods already given. 171. Electrolytic Separation of Antimony from Arsenic and Tin. — ^Antimony may also be separated electrolyti-vally from arsenic and tin, according to Classen.* The arsensc must be present as arsenic acid, otherwise a portion of it will be precipi- tated with the antimony, while the remainder will be oxidized to arsenic acid. A saturated solution of sodium sulphide must first be prepared. If the commercial sodium sulphide is used it must be dissolved in water and the solution saturated with hydrogen sulphide which has been washed with water and then passed through several tubes filled with cotton wool. While passing hydrogen sulphide the air should be excluded. The solution is filtered and rapidly evaporated in a large platinum or porcelain * Ber., 17, 2245; 18, 1110; 2S, 2160. 138 ANALYSIS OF ALLOYS. dish over the free flanij until a thin skin begins to form on top. The hot solution is poured into glass-stoppered bottles, which must be completely filled and immediately well stoppered. The sodium sulphide may also be prepared by completely satura^ting a solution of pure caustic soda with hydrogen sulphide, filtering, and evaporating down, as already directed. The sulphides of arsenic, antimony, and tin are transferred to a weighed platinum dish of about 150 c.c. capacity and dissolved in 80 c.c. of the sodium-sulphide solution. Add enough concen- trated solution of pure caustic soda to furnish 1 to 2 grams of NaOH. The caustic-soda solution must be clear and give no precipitate with the sodium-sulphide solution. The dish is gently warmed to aid solution. It is electrolyzed at a temperature of 50° to 60° with a current of 0.5 ampere. Precipitation of the antimony is complete in two hours. The watch-crystal is washed with a little water, which is allowed to run down the positive electrode. The solution is removed without interrupting the current, the antimony washed with water and then with alcohol, dried at 80° to 90°, and weighed. The TIN may be determined by acidifying the filtrate with hydrochloric acid, concentrating to a bulk of 200 or 300 c.c, filtering, if necessary, and passing hydrogen sulphide. The pre- cipitate is washed with water containing hydrogen sulphide and a few grams of ammonium sulphate. It is weighed as stannic oxide. 172. Separation of Arsenic. — Arsenic is very readily separated from other metals as arsenious chloride, AsClg. This compound is very volatile. When pute it boils at 130°. It is rapidly volatilized when its solution in hydrochloric acid is boiled. Even at tem- peratures considerably below the boiling-point the arsenic slowly volatilizes. When it is present in its higher state of oxidation, the arsenic is not volatilized on boiling its hydrochloric-acid solution. Various reducing agents have been used to convert the arsenic into the arsenious chloride. The most common among these are ferrous and cuprous salts and hydrogen sulphide. If small amounts of arsenic are present the solution may be distilled after the addition of concentrated hydrochloric acid. On repeat- ing the distillation once or twice after the addition of concentrated hydrochloric acid all of the arsenic will be carried over. If much SEPARATION OF ARSENIC, ANTIMONY, AND TIN. 139 arsenic is present, the solution must be saturated with hydrochloric- acid gas and the distillation carried out in a current of this gas. 173. Dissolving Arsenic Compounds. — ^Various methods of de- composing the ore or other material have been proposed, the objects sought being to prevent the loss of arsenic as trichloride and to leave the hydrochloric-acid solution as free as possible from oxidizing material. If the material is soluble in hydrochloric acid, it may be digested cold with concentrated hydrochloric acid for some time. Liquid bromine may then be added and the solu- tion warmed, at first gently and then more strongly, until the excess of bromine is expelled. The material may also be brought into solution by treatment with concentrated nitric acid, the excess being expelled by evaporation after the addition of a little sulphuric acid. The residue may then be dis- solved in concentrated hydrochloric acid. 174. Distillation of the Arsenic. — If the amount of arsenic present is small, the solution may be decanted into the distilling-flask of the apparatus shown in Fig. 20 and the insoluble residue washed with concentrated hydrochloric acid. About 5 grams of ferrous sulphate, or, still better, ferrous chloride or cuprous chloride, are added and 50 to 75 c.c. concentrated hydrochloric acid. The solution is heated to boiling, and when the liquid has boiled down to a volume of about 40 c.c, 50 c.c. concentrated hydrochloric acid is poured through the funnel and the solution again ^^ boiled down to 40 c.c. The distillate is received in a beaker or flask con- taining about 250 c.c. of hydrogen- Fig. 20. sulphide water. After the second dis- tillation this beaker is removed and replaced by another contain- ing hydrogen-sulphide water. The distillation is repeated after the addition of 50 c.c. of concentrated hydrochloric ac'd to the 14L) ANALYSIS OF ALLOYS. flask. A trace of arsenic is sometimes obtained by a fourth distillation. When LARGE AMOUNTS of ARSENIC must be distilled, the drop- ping-funnel is replaced by a glass tube leading to a flask in which hydrochloric-acid gas is generated. This gas is very readily obtained by aUowing concentrated sulphuric acid to drop into concentrated hydrochloric acid. After transferring the arsenic solution to the distillation flask it is saturated with hydrochloric acid by passing the gas into the solution, which has been cooled to zero by immersing the flask in a vessel filled with cracked ice. When the gas is no longer absorbed, but bubbles through the solution, the ice is removed and the solution brought to a boil. The stream of hydrochloric-acid gas is allowed to flow and the distillation is continued until the volume of the solution is reduced to about 40 c.c. The arsenic may be absorbed in hydrogen-sul- phide water and the precipitate of arsenious sulphide filtered off and weighed as such. EXERCISE 29. Analysis of Type Metal. Alloy of Copper, Lead, Antimony, Tin, with Small Amounts of Iron and Arsenic. 175. Solution of the Alloy. — To 1 gram of the alloy, which has been cut into small shavings with a clean knife, or sampled by means of a clean hack saw producing fine "sawings," is added 15 c.c. concentrated hydrochloric acid. The solution is gently warmed on the water-bath and a drop or two of concentrated nitric acid is added occasionally until solution is affected. All of the metals will be converted into chlorides which will remain in solution with the possible exception of lead chloride. An excess of nitric acid is to be avoided as it tends to form insoluble metas- tannic acid, which can be readily distinguished from the crystalline lead chloride. If metastannic acid forms, the operation must be repeated, using less nitric acid or adding it less frequently. After a few trials the correct method of adding the nitric acid is soon acquired. 176. Lead. — The solution is allowed to cool and then stand at least J hour or better over night to allow the lead chloride to crystallize out. Ten times the volume of absolute alcohol is then added in several por- tions. After standing for about half an hour, the lead chloride is filtered off on a Gooch crucible, washed with a mixture of 4 parts of 95% alcohol and 1 part of concentrated hydrochloric acid, and finally with pure alcohol. It is dried for three hours at 150° and weighed. The great advantage of this method of separating the lend is that the very trouble- some treatment of the sulphides of the metals present with sodium or potassium svilphide is avoided. The most difficult part of the operation is the solution of the alloy. ANALYSIS OF TYPE METAL. 141 177- Copper and Iron — Tlie filtrate from the lead chloride is heated until the alcohol is expelled. Two grams of tartaric acid and an excess of ammonia are added and the solution warmed until the precipitate dissolves. By the addition of five c.c. of saturated hydrogen-sulphide water, the copper and the small amount of lead still unprecipitated as well as a trace of iron which may he present may be precipitated with- out bringing down any of the tin and antimony. The solution is warmed and when the dark colored precipitate has settled, one e.e. of the hydrogen-sulphide water is added to the clear supernatant liquid. If no further precipitate is produced, the solution is filtered and the precipitate washed with water containing hydrogen sulphide. The precipitate is dissolved in a httle warm dilute nitric acid and the lead separated as sulphate, the nitric acid being expelled by evaporation after the addition of. sulphuric acid. The copper is precipitated from the filtrate as sulphide and if small in amount may be ignited and weighed as oxide. If considerable copper is present it must be weighed as sulphide or without precipitation as sulphide may be separated electro- lytically from the iron. One or two c.c. concentrated nitric acid are added and a current of one-half ampere passed until all the copper is precipitated. The iron may then be precipitated with ammonia and weighed as oxide. 178. Separation of Antimony and Tin. — The solution of antimony and tin is acidified with hydrochloric acid, hydrogen sulphide passed, and the precipitate filtered off and washed two or three times. A liole is made in the point of the filter-paper by means of a glass rod and the bulk of the precipitate washed into a beaker with a little water. Warm dilute hydro- chloric acid is poured over the paper to dissolve the portion of the precipi- tate still adhering to the paper. The precipitate in the beaker is dissolved by warming and adding concentrated hydrochloric acid. The hydrogen sulphide is decomposed by the addition of a crystal of potassium chlorate and warming. Some pure metallic iron is added and the solution heated on the water-bath for about one-half hour or until the iron is nearly dissolved. The precipitated antimony is filtered off on a Gooch crucible, a little iron having been sprinkled on the asbestos. The precipitate is washed with boUed water to wliich considerable hydrochloric acid has been added. 179. The antimony is dissolved in hydrochloric acid to which a little potas- sium chlorate has been added. The solution is warmed to expel chlorine and, after the addition of tartaric acid and water, hydrogen sulphide is passed. The antimony sulphide is filtered off and washed with water containing a little hydrogen sulphide. The moist precipitate is rinsed into a capacious porce- lain crucible with water. The small portion still adhering to the paper is dissolved in a little warm ammonium sulphide and the solution allowed to flow into the crucible. The solution is evaporated on the water-bath after the addition of a few cubic centimeters of concentrated nitric acid. If sulphur separates, a little liquid bromine is added when the solution has become quite concentrated. When the globule of sulphur has disappeared, expel the excess of nitric acid by heating on the hot plate or with the Bunsen 142 ANALYSIS OF ALLOYS. burner, finally heating to full redness. Cool a little, sprinkle some ammo- nium carbonate over the precipitate, and ignite again to completely expel sulphuric acid and weigh as antimony tetroxide, SbaO,. i8o. Tin. — To precipitate the tin in the filtrate from the antimony the excess of hydrochloric acid is neutrahzed with ammonia, the solution diluted somewhat, warmed, and hydrogen sulphide passed until the tin is entirely precipitated. The stannous sulphide is washed with water containing hydrogen sulphide and a few grams of ammonium sulphate. It is dried and detached from the paper which is burned. The precipitate and the ash are placed in a weighed porcelain crucible and heated very gently with free access of air until sulphur dioxide ceases to be given off. The oxidation may be assisted by the addition of a few drops of nitric acid. Finally the precipitate is strongly heated to expel sulphuric acid, which is completely removed by the addition of a little ammonium carbonate and again igniting. It is weighed as stannic oxide SnO^. i8i. Arsenic. — As only a trace of arsenic is present, a 5- or 10-gram portion of the alloy should be taken for its determination. Dissolve in hydrochloric acid and potassium chlorate and warm to expel the chlorine. Filter off the lead chloride on asbestos and wash a few times with dilute hydrochloric acid. Add one-third the volume of concentrated hydrochloric acid and pass hydrogen sulphide. Filter off the precipitate consisting of the sulphides of copper and arsenic on asbestos, wash with hot water containing hydrogen sulphide and a little hydrochloric acid. Dissolve the arsenic sulphide by washing the pre- cipitate with a little warm dilute ammonia. Evaporate the solution nearly to dryness in a porcelain dish. Oxidize the arsenic by warming mth concen- trated nitric acid, dilute the solution somewhat, neutralize with filtered ammonia, and add magnesia mixture. After standing twenty-four hours filter, wash, ignite, and weigh as magnesium pyroarsenate according to the directions given in Chapter VII, page 88. EXERCISE 30. Analysis of Britannia Metal. Alloy of Tin, Antimony, and Copper, with Small Amounts of Bismuth, Lead, and Iron. 182. Decomposition of the Alloy by Means of Chlorine. — Alloys containing a large percentage of tin are best decomposed by a stream of chlorine. The method is apphcable to alloys containing less than 15% of lead and copper. A hard-glass combustion-tube 70 cm. long is taken and one end drawn out, making a small tube 20 cm. long, which is bent at right angles. This small tube is connected by means of a cork stopper with a PeUgot tube the Dulbs of which are nearly filled with dilute hydrochloric acid (1 : 3) ANALYSIS OF BRITANNIA METAL. 143 containing about 1 gram of tartaric acid. A second Peligot tube is con- nected with the first and contains a solution of caustic soda (1:3). The chlorine is evolved in a 2-liter flasli containing pieces of pyrolusite, over which concentrated hydrochloric acid is poured. The flask is heated on a water-bath. The chlorine is passed through a wash-bottle containing water and then tlarough two wash-bottles containing sulphuric acid. It is then passed into the combustion-tube, connection being made by means of a cork stopper. Wherever rubber is used for making connections, it must be well coated with paraffine. This is also advisable for the cork stoppers. The chlorine is not allowed to pass into the combustion-tube until all of the air has been displaced from the flask and the wash-bottles. All escaping clilorine should be absorbed in caustic-soda solution. One gram of the alloy in fine turnings is weighed out and placed in a porcelain boat which is placed in the middle of the combustion-tube. The chlorine is first allowed to act on the alloy in the cold. When no further action is observed, the part of the tube in which the boat is situated is heated gently with the Bunsen burner, and then more strongly until the contents of the boat fuse. The chlorides of mercury, bismuth, arsenic, ANTIMONY, and TIN VOLATILIZE and are driven out of the tube by heating it gently from the boat to the end which is drawn out. These chlorides are absorbed in the hydrochloric acid contained in the first Peligot tube, whUe the excess of chlorine is absorbed in the caustic-soda solution con- tained in the second Peligot tube. The chlorine in the apparatus is then displaced by means of a stream of dry air or carbon dioxide, the chlorine generator having been removed. The apparatus is disconnected, the boat containing the chlorides of COPPER, LEAD, and IRON is placed in a porcelain dish, and the tube washed out with hot water which is aUowed to flow into the dish containing the boat. Hydrochloric acid is added and the dish warmed until the contents of the boat are dissolved. The latter is removed and washed. 183. Lead, Copper, and Iron. — The lead is precipitated by evaporation with sulphuric acid and diluting and is filtered off and weighed as sulphate. The copper is precipitated by means of hydrogen sulphide and weighed as sulphide or determined electrolytically from a nitric acid solution. The iron is precipitated by means of ammonia and weighed as oxide. The contents of the first Peligot tube are poured into a beaker and the Peligot tube well washed out with water to which hydrochloric acid is added if necessary. The solution is warmed and hydrogen sulphide passed until precipitation is complete. The filtrate should be heated to boihng, strong hydrochloric acid added, and hydrogen sulphide passed again to insure complete precipitation of the arsenic. 184. Bismuth. — If the sulphide precipitate is dark colored, bismuth is present. The precipitate is washed into a beaker, ammonium sulphide added, and the solution warmed. The solution is filtered through the same paper and the precipitate washed with warm water containing a little 144 ANALYSIS OP ALLOYS. ammoniuHi sulphide. The bismuth sulphide is dissolved in a little warm dilute nitric acid and the paper washed. The bismuth is precipitated with ammonia and ammonium carbonate, ignited, and weighed as oxide, Bi fi,. 185. Separation of Tin from Arsenic and Antimony. — The ammonium- sulphide solution of arsenic, antimony, and tin is poured with vigorous stirring into, a hot solution of 25 grams of oxahc acid in 200 c.c. of water. The solution is heated to boiling and hydrogen sulphide passed for about fifteen minutes. The precipitate is filtered off immediately and washed with hot water containing hydrogen sulphide. It is dissolved in ammo- nium sulphide and the treatment with hot oxalic acid and hydrogen sulphide repeated. 186. Tin. — The oxalic-acid solution of tin is evaporated down, with the addition of 5 c.c. concentrated sulphuric acid, to fumes. The solution is cooled, cautiously diluted with water, and hydrogen sulphide passed to insure complete precipitation of the tin. Wash the precipitate with water containing ammonium acetate and a little acetic acid, dry, ignite, and weigh as stannic oxide SnO„. 187. Arsenic and Antimony.— ^The precipitate of arsenic and antimony sulphides is treated with a little concentrated ammonium-carbonate solu- tion and washed to remove arsenic. The antimony is then weighed as oxide according to the directions given in Exercise 29, page 141. The arsenic is determined according to the directions given in the same exercise. CHAPTER XII. ANALYSIS OF ALLOYS CONTAINING IRON, NICKEL, AND ZINC. i88. Separation of Zinc as Sulphide. — Zinc may be separated very completely from most metals as sulphide. Hydrogen sul- phide passed through the moderately acid solution separates silver, lead, mercury, bismuth, copper, cadmium, arsenic, antimony, and tin from zinc. On neutralizing the solution almost com- pletely, or making it acid with acetic or citric acid, or a drop or two of a mineral acid, and passing hydrogen sulphide, zinc is pre- cipitated as sulphide, while cobalt, nickel, manganese, aluminium, and iron remain in solution as well as the alkaline-earth metals and the alkalies. Zinc may be separated from chromium and aluminium by adding sodium sulphide to a solution of these metals containing tartaric acid made alkaline with ammonia. Although the separation of zinc as sulphide from other metals is very perfect and the metal can be weighed with the greatest accuracy as sulphide, the method is not very largely used. This arises from the fact that zinc sulphide is very difficult to filter and wash because of its tendency to pass through the filter-paper. Ignition of a precipitate with sulphur in a stream of hydrogen is also somewhat troublesome. The latter difficulty may be obviated by dissolving the sulphide in hydrochloric or nitric acid and pre- cipitating the zinc as carbonate and weighing as oxide or precipi- tating as zinc-ammonium phosphate and weighing as zinc pyrophos- phate. The zinc sulphide is washed more readily if hydrogen sulphide and ammonium sulphate or acetate are present in the wash-water. 189. Separation of Zinc as Zincate. — Zinc may also be sepa- rated from cobalt, nickel, iron, and manganese by pouring the acid solution of these metals into excess of caustic-soda solution. The zinc hydroxide first precipitated redissolves in the excess of alkali, 145 146 ANALYSIS OF ALLOYS. forming sodium zincate, while the hydroxides of the other metals are insoluble in the caustic alkali and may be filtered off. Before filtering, the solution should be diluted to precipitate a small amount of cobalt, which dissolves in concentrated caustic-soda solution. The concentrated alkali is also liable to ruin the filter-paper. The zinc may be precipitated as sulphide by passing hydrogen sulphide through the filtrate. The sulphide obtained in this manner from an alkaline solution coagulates and settles readily and does not tend to pass through the filter-paper. If the hydro- gen sulphide is passed too long the zinc sulphide redissolves.* The presence of sodium or ammonium chloride tends to prevent this action. The danger may be avoided by adding hydrogen sulphide water until no more precipitate forms. The alkali is washed out of the precipitate with considerable difficulty, how- ever, and the precipitate also carries down silica. After igniting with sulphur in a stream of hydrogen the alkali may be extracted with hot water and the zinc sulphide again ignited and weighed as sulphide. For determining silica, the zinc sulphide is dissolved in hydrochloric acid and the solution evaporated to dryness. The zinc chloride is dissolved in water an.d the silica filtered off, ignited, and weighed. EXERCISE 31. Analysis of Brass or Bronze. Alloy of Lead, Copper, Tin, and Zinc, Small Amounts of Arsenic and Iron. igo. Solution of the AUoy.^ — ^Weigh out 1 gram of the alloy and place in a 300-c.c. beaker, add 10 c.c. concentrated nitric acid and 5 c.c. water. Cover the beaker with a watch-crystal and place in a dish of cold water. After one-half hour place the beaker on the water-bath and evaporate the solu- tion to dryness. 100 c.c. of boiling water and a few drops of nitric acid are added and the solution boiled for five minutes. 1 91. Tin. — The staimic oxide is filtered off and washed with hot water. The moist precipitate is introduced into a weighed porcelain crucible and the paper burned in the usual manner. If the amount of tin is small (less than l%)it is weighed at this point, otherwise it is fused with six times its weight of a mixture of equal parts of sulphur and sodium carbonate. Tlie fused mass is dissolved in hot water and the solution filtered. The copper, lead, and iron whicli were carried down A\ith the stannic oxide will remain on the paper as sulpliides, while the filtrate will contain all of the tin and *L. W. McOay, Jour. Am. Chem. Soc, 30. 376. ANAL) SIS OF BRASS OR BRONZE. 147 any arsenic or antimony which may have been present. The insoluble sulphides are dissolved in a little nitric acid, the paper washed, and the solution added to the filtrate from the stannic oxide. If arsenic and antimony are absent, the tin may be precipitated out of the sodium sulphide solution and weighed. The excess of sulphur should first be removed from the solution by heating to boiling after the addition of caustic soda and then adding hydrogen peroxide in small quantities until the solution is nearly decolorized. It is then acidified with hydrochloric acid while stirring constantly, heated, and hydrogen sulphide passed. The stannic sulphide is washed with hot water containing ammonium acetate and a little acetic acid. It is ignited and weighed as stannic oxide in the usual manner. 192. Arsenic and Antimony. — If arsenic is present in the alloy, a small amount of this element will be present in the sodium sulphide solution of the tin and will be precipitated with the stannic sulphide. It may be removed by treating the precipitate -udth a little concentrated solution of ammonium carbonate and washing. The solution of arsenic should be added to the nitric acid solution of the alloy. If antimony is also present in the alloy, the sulphides of arsenic, anti- mony, and tin must be separated by one of the methods given in Chapter XI, page 141. 193. Lead. — To the filtrate from the stannic oxide 5 c.c. concentrated sulphuric acid are added and the solution evaporated in a porcelain dish until the nitric acid is entirely expelled and white fumes of sulphuric acid are given off. The solution is cooled by floating the dish on cold water and diluted with 75 c.c. of water. After stirring thoroughly, 25 c.c. alcohol are added and the solution allowed to stand for at least one hour. The lead sulphate is filtered off on a Gooch crucible, washed with water containing about 1% of sulphuric acid and 25% of alcohol and then mth pure alcohol until free from acid. It is dried on the hot plate and weighed. 194. The copper is best determined electrolytically. The filtrate from the lead sulphate is heated on the hot plate until most of the alcohol is expelled. Two c.c. concentrated nitric acid are added and the warm solu- tion (about eO") electrolyzed with a current of one-half to one ampere for about six hours. Hydrogen sulphide is passed through the acid filtrate from the copper to precipitate traces of arsenic, antimony, or unseparated tin which may be present. If more than traces are found, the metals must be separated and determined by the methods given in the preceding exercises. When the amount of copper is large, as is generally the case, it is advisable to divide the solution into two portions for the electrolysis, as about 300 mg. of copper is generally sufficient for a good determination. The solution may be divided by weighing it and then pouring out about half of it and again weighing or the solution may be diluted to a loiown volume as 250 or 500 c.c. and a portion measured out. The copper may be determined in each portion and the filtrates combined for the zinc determination. For the 148 ANALYSIS OF ALLOYS. duplicate zinc determination the copper may be precipitated as sulphide, which is filtered off, well washed, and discarded. 195. Iron.— The filtrate from the copper is boiled to expel hydrogen sulphide and a little nitric acid added to oxidize the iron, which is precipi- tated with ammonia and weighed as oxide. If more than a small amount of iron is present, the precipitate must be redissolved and reprecipitated to completely separate it from the zinc. 196. Zinc. — The filtrate from the iron is evaporated to small bulk and the zinc precipitated and weighed as pyrophosphate. The zinc may also be precipitated and weighed as sulphide. EXERCISE 32. Analysis of German Silver. Alloy of Copper, Zinc, and Nickel, with Small Amounts of Lead, Iron, and Tin One gram of the alloy is weighed out and dissolved in nitric acid as directed in the preceding exercise. The tin, lead, and copper are deter- mined as directed in the same exercise. Hydrogen sulphide is passed through the acid filtrate from the copper to precipitate traces of arsenic, antimony, tin, or unseparated copper which m.ay be present. If more than traces are found, the metals must be sepa- rated and determined by the methods given in the preceding exercises. 197. Zinc. — The filtrate is boiled until the hydrogen sulphide is expel-led and the solution concentrated to a small bulk and the acid nearly neutral- ized vAth caustic soda. Five to ten grams of caustic soda are dissolved in about 50 c.c. of water and the solution of zinc and nickel added slowly with constant stirring. The solution is diluted with an equal bulk of water and the precipitate filtered off and washed. The zinc in the filtrate is pre- cipitated with hydrogen sulphide, filtered off, and washed free from alkali. The zinc sulphide is dried and detached from the paper as completely as possible. The portion still adhering to the paper is dissolved in nitric acid and the solution evaporated to dryness in a porcelain crucible. The remainder of the precipitate is added and the whole ignited with sulphur in a stream of hydrogen. If the precipitate is small it need not be dried, but is imme- diately dissolved in nitric acid and after evaporation converted into sul- phide. The sulphide is tested for alkali by digestion with hot water. If alkali is found it must be completely extracted and the sulphide again weighed after ignition with sulphur in hydrogen. The precipitate is then dissolved in nitric acid aud the solution evaporated to dryness. The zinc nitrate is dissolved in water and the silica filtered off, washed, ignited, and weighed. The zinc sulphide may also be dissolved in hydrochloric acid, the zinc precipitated as zinc ammonium phosphate and weighed as pyrophosphate. 198. Iron and Nickel. — If iron is absent, the nickel hydroxide may be washed and after transferring the precipitate to a weighed porcelain crucible ANALYSIS OP GERMAN SILVER. 149 and burning the paper it may be reduced to metallic nickel by heating in a stream of hydrogen and weighed. If iron is present, the precipitate is dissolved in hydrochloric acid and the iron precipitated with ammonia. Unless a very small amount is present it must be redissolved and repre- cipitated, and after washing is ignited and weighed as oxide. The nickel is then reprecipitated as hydroxide by means of an excess of caustic soda, reduced in a stream of hydrogen and weighed as the metal. EXERCISE 33. Analysis of Manganese-Phosphorus-Bronze. Alloy of Copper, Lead, Tin, Zinc, Manganese, Phosphorus (less than 1%), Traces of Iron. 199. Solution. — One gram of the alloy is weighed out and dissolved in nitric acid as directed in Exercise 31 , page 146. Nearly all of the phosphorus remains with the stannic oxide as a phosphate. After fusing the impure precipitate and separating the impurities as given in Exercise 31, and pre- cipitating the tin as sulphide, the solution containing only the phosphorus as phosphoric acid is discarded, as this element is determined in a separate portion of the alloy. 200. Lead, Copper, and Zinc are determined as given in Exercise 31. The phosphoric acid which did not remain with the stannic oxide will be present in the alkaline solution of the zinc. This elenient should therefore be precipitated and weighed as pyrophosphate. 201. Iron. — In order to separate manganese and iron from zinc, bromine or hydrogen peroxide is added to the filtrate from the copper. The solu- tion is boiled and excess of ammonium added to redissolve any zinc phos- phate which may be precipitated. The precipitate consisting of ferric hydroxide and manganese dioxide is filtered off and washed. It is dis- solved in a little hydrochloric acid and the paper well washed. The solu- tion is boiled until the chlorine is completely expelled, then neutralized with ammonia, warmed, and the trace of iron filtered off immediately. Unless the precipitate is very small it is redissolved in hydrochloric acid and again precipitated with ammonia and quickly filtered off and washed. It is ignited and weighed as oxide. 202. Manganese. — The combined filtrates from the iron contain aU of the manganese unless the amount of iron present is considerable. The solution should be evaporated to dryness in a porcelain dish and the ammo- nium chloride volatilized by gently heating with the Bunsen burner. The residue is dissolved in a few c.c. water and a few drops of hydrochloric acid ai}d the manganese precipitated and weighed as sulphide as directed in Exercise 22. 203. Volumetric Determination of Iron and Manganese. — If considerable iron is present, the method of separation given is not applicable. In this case the simplest methods of determining the two metals are volumetric. The ammonium precipitate should be dissolved in sulphuric acid with 150 AXALYSIS OF ALLOYS. expelled by boiling. The solution must be made up to a definite vol- ume and divided into two equal portions. For this purpose a 100-c.e. flask should be used which has been calibrated with a 50-c.c. pipette liy emptying the pipette twice into the dry flask and making a mark on the stem. The solution of iron and manganese is evaporated to small bulk, transferred to the flask, made up to the mark and thoroughly mixed. 50 c.c. are withdrawn with the dry pipette. The solution adhering to the walls of the pipette is rinsed out with distilled water and added to the portion remaining in the flask. One of these portions is re- duced with zinc and the iron titrated with standard permanganate. (See p. 316.) The other portion is shaken up with zinc oxide until the free acid is neutralized. One gram of zinc sulphate and a drop or two of dilute nitric acid are added and the solution diluted to several hundred cubic centimeters. The manganese is titrated with standard potassium permanganate according to Volhard. (See p. 327.) 204. Phosphorus. — For the determination of phosphorus a 5-gram por- tion of the alloy is taken, as the percentage of this element is usually small (seldom more than 0.2%). The material is placed in a 200-c.c. beaker and 20 to 30 c.c. concentrated nitric acid added. The beaker is covered with a watch-crystal and after the first violent action of the acid has ceased it is placed on the water-bath and heated until the alloy is completely decom- posed and the residue is pure white. All of the phosphoric acid will remain with the tin provided a sufficient amount of the latter is present in the alloy. From six to eight times as much tin as P^Oj must be present. Unless at least 5% of tin has been found, a preliminary test should be made by dissolving about a gram of the alloy in concentrated nitric acid, filtering, and testing the filtrate for phosphoric acid wdth molybdate mixture. If phosphoric acid is found in the filtrate, metallic tin must be added before dissolving the alloy in nitric acid. From one-half to one gram will usually be found sufficient. The nitric acid solution of the alloy is diluted and the stannic oxide containing the phosphoric acid is filtered off and washed a few times. After drying, the precipitate is transferred to a porcelain crucible, the jjaper is burned, and the ash added. After adding three times its weight of potassium cyanide, cover the crucible and fuse for a few minutes at a red heat. The stannic oxide is reduced to metallic tin and the phosphoric acid forms potas- sium phosphate. After cooling, extract the fused mass ^vith hot water, filter, and wash the paper with hot water. Expel the hydrocyanic and cyanic acids by boiling ivith concentrated hydrochloric acid. This operation must be conducted under a hood with good draught. Evaporate to dryness to dehydrate the silicic acid which has been dissolved from the porcelain by the action of the potassium cyanide. Dissolve the dry residue in a little hydrochloric acid and pass hydrogen sulphide to precipitate a small amount of tin and copper which is present. Filter, wash the precipitate, and destroy the hydrogen sulphide in the fil- ANALYSIS OF MANGA.'iESE-PHOSPHORUS-BIiUNZE. 151 trate by adding bromine water and boiling. If the volume of the solution exceeds 50 c.c, concentrate by boiling. Cool and precipitate the phos- phoric acid by adding about one-half gram of crystalhzed magnesium chloride or sulphate dissolved in a little water and then neutralizing the solution wth' filtered ammonia while stirring vigorously. Add a small excess of ammonia. Assure yourself that the phosphoric acid is aU pre- cipitated by adding a little magnesia mixture to the clear supernatant liquid. After standing several hours, filter, wash with dilute ammonia Ignite in a porcelain crucible, and weigh as magnesium pyrophosphate. The precipitation of the metals present ivith hydrogen sulphide may be omitted and the separation effected by precipitating the phosphoric acid as molybdate. The dry residue should then be dissolved in nitric acid, and flfter filtering off the silica, the phosphoric acid is precipitated as directed in Chapter IX, page 118. ANALYSIS OF MINERALS. CHAPTER XIII. MINERALS CONTAINING IRON, ALUMINIUM, AND CHROMIUM. SELECTION AND PREPARATION OF SAMPLE. 205. Taking the Sample for Analysis. — One of the mOst impor- tant operations preliminary to many chemical analyses is the selection of the sample to be analyzed. When the material to be analyzed is a chemical, such as the various salts made for commercial purposes, the task is comparatively simple since the product is usually quite imiform and small grained. Samples are taken from a number of the barrels or other receptacles. In many cases it is advisable to take samples from various parts of the barrel, such as the end, middle, etc. This is most readily accomplished by means of a " butter-tryer " or "thief" or some similar apparatus. These various samples are brought together, thoroughly mixed, and a smaller sample taken out. This may be done by taking small portions from various parts of the pile or by dividing it into quarters. One quarter or two diagonal quarters are taken, well mixed, and the sampling repeated in the same manner until a convenient quantity for analysis is obtained. Care in sampling dry and apparently uniform products is necessary for a good many reasons. Many substances when exposed to the air, as in the outer portions of a barrel, absorb or lose moisture, oxidize, lose volatile constituents, or in other ways change in composition. Even in the case of a mixture of two solid substances of different specific gravity the heavier constituent tends to settle to the bottom, so that after the most thorough mixing the material will not be uniform after standing for some time. If LIQUIDS are to be sampled they must be thoroughly shaken or parts of the first, middle, and last runnings of a cask or drum 152 SELECTION AND PREPARATION OF SAMPLE. 15.'^ must be taken. These portions should be thoroughly mixed before taking the sample for analysis. Metals and alloys are by no means uniform, since, on solidi- fying from the molten state, the heavier constituents are apt to settle to the bottom, as do also the constituents which crystallize out or solidify first. In the case of a bar the outer portions are not always of the same composition as the interior of the metal. Drilling through a piece of metal with a clean tool is therefore apt to give the best sample. If the metal is in the form of a sheet, portions should be punched or cut from the outer edges as well as from the centre. If car-loads of metals a,s" pigs" are to be sam- pled, one or more "pigs" in each car should be punched or drilled and the samples combined by melting or mixing before the labora- tory sample is taken. The turnings of the more brittle metals may be pulverized by grinding in a mortar or a coffee-mill. If a ROCK is to be analyzed, the sample is usually selected because of its apparent homogeneity. The physical properties usually give sufficient evidence of the presence of foreign material. If the rock or mineral is to be analyzed with a view to its use as the source of a given metal or other constituent, the sample must include all foreign material which would be taken out in mining operations. The cars or shiploads of orb are usually sampled by taking out shovelfuls of the material from various parts of the load. These portions are thrown together on a floor or other convenient place. The larger pieces are broken with a hammer and the material well mixed. The pile is quartered and the quarter taken is still further powdered and quartered. This process is continued until a sample of convenient size is obtained. About 25 grams will usually be found sufficient, though in some cases as much as a kilogram will be required for the determination of elements present in very small amount. 206. Pulverizing. — The tools used in breaking off and pow- dering ores or rocks include, in the first place, hammers with hardened surfaces. These are made in various shapes and sizes to suit the geologist prospecting in the field or the laboratory assistant in breaking up the large specimens of rock. A chilled IRON PLATE with a hardened steel muller to which a rocking 154 ANALYSIS OF MINERALS. motion may be given by means of a long handle may be used to reduce many ores and rocks to powder fine enough for analysis. Fig. 21. Hardened steel mortars are also useful in crushing moderately small pieces of ore or rock. The most convenient method of Fig. 22. crushing large samples of rock or ore is by the use of ore-crushers. The jaws between which the ore is placed are actuated by means Fig. 23. of a long handle so connected with levers that a very great crush- ing pressure may be brought to bear on the material. It is cus- tomary to remove by means of a magnet particles of iron which SELECTION AND PREPARATION OF SAMPLE. 155 have been detached from the tools during the process of pulverizing. Very few rocks or ores are entirely free from magnetic particles, so that this method of procedure is open to objection. Hard rocks and ores, as well as those known to contain iron, should be pulverized as far as possible in agate mortars. These, as well as the iron mortars, are frequently arranged to be oper= ated by mechanical means. Many chemists, however, still prefer the hand grinding. 207. Sifting. — ^The requisite degree of fineness is frequently secured by passing the powdered material through sieves of the requisite size of mesh. Eighty-mesh brass sieves are suitable for most ores. In special cases where the sample is very irregular, sieves of 100 to 120 mesh to the linear inch should be used. Even when kept scrupulously clean a small amount of the metal of the sieve is apt to contaminate the sifted material. This error may be entirely disregarded if a constituent present in very small amount is to be determined. When a complete analysis is to be made, many chemists prefer to use bolting-cloth or tine linen. Where the state of oxidation of the iron is to be determined, the shreds of the cloth, which almost necessarily contaminate the sifted material, undoubtedly influence the result to an appreciable degree. For this reason some chemists advocate grinding the material without sifting at all. In sifting through cloth large particles are readily forced through if the powder is rubbed over the cloth. A wide-mouthed bottle should be selected, thoroughly cleaned and dried. The cloth should be stretched over the mouth of the bottle and fastened securely by winding a string around the neck. The material to be sifted, after having been ground in small portions in an agate mortar, is placed on the cloth and a piece of sheet rubber or leather stretched over the mouth of the bottle and held securely with the hand or by a rubber band placed around the neck. The rubber or leather is now lightly tapped with a spatula until no more fine material sifts into the bottle. The coarser material is again ground in the mortar and sifted. In no case should the coarse material be discarded. Frequently the more friable portions of a rock or ore are different in compo- sition from the particles which resist grinding longest. In pul- 156 ANALYSIS OF MINERALS. verizing a sample the rule must be strictly adhered to that the whole of a portion taken for powdering must be ground and passed through the sieve. In sampling, the large pieces must be broken completely so as to mingle with the finer material and constitute a fair portion of the sample finally taken. 208. Drying. — ^After being reduced to a sufficient state of fineness, the material to be analyzed is frequently dried at a tem- perature of about 100° to remove hygroscopic moisture before being weighed out. Other workers weigh out the air-dried mate- rial and determine the percentage of water lost at 100°. It is held that the amount of hygroscopic water present is a very char- acteristic property of some rocks and minerals, and serves to distinguish minerals which are in other respects identical. Hygro- scopic material can also be weighed out much more accurately when air-dried than when entirely desiccated. In the latter case it gradually absorbs moisture, as it is exposed to the air from time to time in the course of the analysis, so that the material weighed out the first day is not identical in composition with that weighed out later. SEPARATION OF IRON, ALUMINIUM, AND CHROMIUM. Iron and aluminium occur very frequently together, and are generally precipitated together from solutions containing man- ganese, cobalt, nickel, and zinc. The methods suitable for pre- cipitating these two metals are not generally applicable to the complete precipitation of chromium. The separation of iron and aluminium will therefore be first considered, and then that of chromium. 209. Separation as Hydroxide. — In the presence of excess of ammonium chloride and ammonia, nickel, cobalt, and zinc are not at all, while manganese is only slowly precipitated by being oxidized to the hydrated sesquioxide, while ferric iron, aluminium and chromium are very nearly completely precipitated. If all of these metals are present in solution, the iron, aluminium, and chromium may be completely precipitated by means of ammonia and ammonium chloride, but the precipitate will be contaminated with considerable amounts of manganese and varying amounts of zinc, nickel, and cobalt. In the absence of chromium, bv SEPARATION OF IRON, ALUMINIUM, AJSID CHROMIUM. 157 redissolving and reprecipitating the iron and aluminium once or twice, the separation from niclcel and cobalt may be made com- plete and very nearly so from zinc. Only a small part of the other metals present in the solution will be carried down each time with the iron and aluminium pre- cipitate. On dissolving the first precipitate only this small part of the contaminating metals will be present in the solution; the bulk of these metals will be in the filtrate. On making a second precipitation the fraction carried down will be very small indeed, so that on a third precipitation the amount carried down may be wholly neglected. This process will not free the precipitate from manganese, because this metal is oxidized very readily in alkaline solution to the insoluble dioxide. The solution should be heated, excess of ammonia added, and if aluminium is present ammonium carbonate added. After digesting hot for some time the precipitate is filtered off, washed a few times, redissolved in hydrochloric acid, and the . operation repeated. If the amount of nickel and cobalt is considerable, a third precipitation is necessary. Manganese must be absent. The precipitate may be ignited and weighed, and one of the con- stituents determined, the other one being obtained by differ- ence. 210. Separation as Basic Acetates. — If manganese and zinc are present, the aluminium and iron must be precipitated as basic acetates. This method of separation is based on the fact that iron and aluminium may be precipitated as basic acetates by boiling a solution of these metals which contains ammonium or sodium acetate and free acetic acid. The acid present prevents the precipitation of manganese, zinc, cobalt, and nickel. The separation is therefore much sharper than when the iron and aluminiimi are precipitated from alkaline solution. Oxidizing agents must be absent or manganese will be precipitated, but the iron must be in the ferric condition. The proportion of free acetic acid to ammonium acetate must be carefidly regulated. They should be present in gram-molecular proportion. As the commercial ammonium acetate is not always neutral it should be tested before use and neutralized. The sim- plest way of securing the proper proportion of free acid and neu- 158 ANALYSES OF MINERALS. tral salt is to make the ammonium acetate by neutralization of acetic acid. If the same volume of acetic acid is neutralized as has already been added to the solution the correct amoimt of ammonium acetate will be secured. The volume of the solution must also be carefully regulated with reference to the amount of iron present. The perfectly cold hydrochloric acid solution must be made very nearly neutral by the addition of concentrated and then dilute ammonium carbonate solution until the slightest turbidity is noticed. If a precipitate forms it must be dissolved in hydro- chloric acid, and the solution again cautiously neutralized. One c.c. of 5 N acetic acid is added for each 0.1 gram of iron. Water is added until at least 150 c.c. is present for each 0.1 gram of iron. Heat to boiling in a porcelain dish, and add 1 c.c. of 5 N acetic acid which has been exactly neutralized with ammonia for each 0.1 gram of iron present. Continue the boiling for two or three minutes. . After allowing the precipitate to settle, decant the liquid and wash by decantation a few times, bringing the wash- water to a boil each time before decantation. The precipitate is free from zinc and all but a trace of manganese, but if nickel and cobalt are present it is apt to contain a little of these metals^ more of the former than of the latter. It may be freed from these metals by redissolving and reprecipitating. If the filtrate has a slight yellow color the iron has not been completely precipi- tated because of the presence of too much acetic acid. A few drops of ammonia may be added and the solution warmed until the precipitate collects. It is v/ashed with hot water containing a little acetic acid and ammonium acetate. This method gives better results with iron than with alu- minium, unless iron is also present in considerable amount. Small amounts of chromium may also be precipitated by this method if considerable iron is present. The precipitate cannot be ignited and weighed if alkalies were present in the solution because it carries down alkalies which cannot be washed out. It must be dissolved in hydrochloric acid and reprecipitated with ammonia. The method has the disadvantag-j that tJie precips^j^s is very bulky, and is frequently slany and difhcult to wash. For thy. reason it is not used when a large amount of iron must be handled. SEPARATION OF IWN, ALUMINIUM, AND CHROMIUM. 159 211. Separation of Iron and Aluminium as Basic Carbonates.— Iron and aluminium may also be separated from manganese, cobalt, nickel, and zinc, by precipitation of the former as basic carbonates. The hydrochloric acid solution is carefully neutral- ized with ammonium carbonate, until a slight turbidity appears. A drop or two of acetic acid is added. Ammonium chloride must be present to the extent of about 1 gram for each 0.1 gram of iron. The solution is gradually brought to a boil and the heating con- tinued until the carbon dioxide is completely expelled. The pre- cipitate is washed by decantation, a few drops of ammonia being added to the wash-water. A slight yellow color in the filtrate indicates incomplete precipitation of iron. A few drops of ammo- nia may be added, but care must be taken not to make the solution alkaline or manganese will be precipitated. 212. Separation of Chromium as Chromic Acid. — Chromium is best separated from other metals by methods which are based on the acid properties of chromium trioxide. It is' converted into the acid oxide by treatment with an alkaline oxidizing agent. A mixture of sodium carbonate and potassium nitrate has been largely used for this purpose. On extracting the fused mass with water the oxides of zinc, cobalt, nickel, and iron remain undissolved. Part of the manganese remains undissolved while part goes into solution as manganate. On boiling the solution after the addi- tion of a little alcohol the manganese, is completely precipitated as the dioxide. Part of the aluminium remains with the insoluble portion, while part passes into solution with the sodium chromate as sodium aluminate.. The aluminium may be separated from the chromium by precipitation as aluminium hydroxide. The solution must be acidified with hydrochloric acid to decompose the sodium aluminate. On adding ammonium chloride and ammonia until the solution is faintly alkaline and warming, the aluminium is completely precipitated as hydroxide. The chro- mium in the condition of chromic acid is not precipitated by the ammonia. • The potassium nitrate used to oxidize the chromium is reduced during the fusion to nitrite which is present in the water solution of the melt. On acidifying this solution, nitrous acid is liberated which tends to reduce the chromic acid. To destroy the nitrous 160 ANALYSIS OF MINERALS. acid potassium chlorate or still better liquid bromine is added to the alkaline solution. On acidifying and warming gently, the nitrous acid is oxidized and the excess of bromine or chlorine expelled. The chromium in the filtrate from the aluminium may be precipitated and weighed as lead or barium chromate. It may also be reduced by sulphurous acid, precipitated with ammo- nia and weighed as oxide, or it may be determined volumetrically by means of ferrous sulphate and a standard chromate solution. Recently fusion with sodium peroxide has come into exten- sive use. This flux seems to be especially adapted to decompose the refractory chrome iron ore. A silver, nickel, or copper cruci- ble must be used, as platinum is strongly attacked by the alkaline flux. A mixture of 5 to 6 parts of caustic soda with 3 to 4 parts of sodium peroxide has also been considerably used. If the chromium is in solution with other metals the separation is not so simple. The acid solution is nearly neutralized with sodium carbonate, sodium acetate added, the solution heated, and chlorine passed or bromine added. The chromium is oxidized to chromic acid and remains in solution. Iron and aluminium are precipitated during the boiling by the sodium acetate. Any manganese present is precipitated as the peroxide together with part of the cobalt, nickel, and zinc. 213. Rothe's Ether Separation of Iron. — ^The method pro- posed by J. W. Rothe for the separation of iron from chromium, aluminium, manganese, cobalt, nickel, and co-piper has been much used, and overcomes many difficulties met with in analyzing compounds containing much iron. It depends on the fact that ferric chloride may be almost completely extracted by means of ether from a hydrochloric acid solution of sp. gr. 1.100 to 1.105 containing the metals mentioned. Nitric acid, chlorine, and more than small amounts of sulphuric acid must be absent, and the solution must be kept absolutely cold, as in a warm solution the ferric chloride is reduced by the ether. All suspended matter such as silica, carbon, etc., must be removed by filtration The apparatus designed for this separation is shown in Fig. 24. The two bulbs A and B are connected by a three-way stop-cock at E^ so that liquid may flow from one bulb to the other or from either bulb through the exit-tube to a receptacle placed below. It the SEPARATION OF IRON, ALUMINIUM, AND CHROMIUM. 161 hydrochloric acid solution of the iron is placed in A, about an equal volume of ether is placed in i?. A slight pressure is pro- duced over the ether by means of a rubber bulb. By cautiously turning the stop-cock E into the position 2, the ether is allowed to flow into the bulb A. The liquids are well mixed by vigorous Fig. 24. shaking. During this operation it is well to keep the bulb A well cooled by holding it in a stream of cold water. After allowing a few minutes for the liquids to separate, the stop-cock E is turned so that the lower heavy liquid may flow into B. Leave just enough of 162 ANALYSIS OF MINERALS. this liquid to fill the capillary tube C, shake down any heavy liquid adhering to the sides of ^1, and allow to stand until the ethereal layer is clear. Again connect A and B by turning the stop-cock C, and transfer the remainder of the heavy liquid to B together with a little of the ether solution. Turn the stop-cock E into the position 3, and allow the ether solution to flow out into a beaker. Rinse the walls of the bulb A with a little ether, and after stand- ing a few minutes observe if any of the heavy solution collects in C If so, transfer as before to B. Repeat the extraction with 50 c.c. of ether. (See Exercise 35, page 170.) 214. Volumetric Separation of Iron and Alviminiiun. — The iron and aluminium precipitate obtained in separations from other metals is best* ignited and weighed as AljOj+FejOj and one of the metals determined, the other being obtained by differ- ence. For this purpose it is simplest to determine the iron volu- metrically. The precipitate is fused with acid potassium sul- phate until dissolved, the fused mass dissolved in water, and the solution filtered to separate a small amount of silica which may be present. This is ignited and weighed, and the amount deducted from the weight of the oxides. The solution of the iron and alu- minium is treated with zinc to reduce the iron, which is titrated with standard potassium permanganate solution as directed in Chapter XXIV, page 316. If the precipitate is large it may be dissolved in hydrochloric acid before weighing and divided into two portions. The iron is determined in one portion while in the other both metals are precipitated as hydroxides, ignited, and weighed as oxides. 215. Separation of Iron from Aluminium as Aluminate. — ^The two metals may also be separated by means of boiling caustic potash or soda solution. The hydrochloric acid or potassium bisulphate solution of the two metals is nearly neutralized, and poured slowly with constant stirring into excess of caustic potash or soda solution, heated nearly to boiling in a platinum or silver dish. The iron is precipitated as hydroxide, while the aluminium remains in solution. The iron precipitate is washed more readily if part of it is reduced by adding a little sodium sulphite and heating before pouring into the boiling alkaline solution. Before weighing, the iron precipitate must be dissolved in hydrochloric acid and re- ANALY,SIS OF CHROMITE. 163 precipitated with ammonia to free it from the alkali which is always present. The aluminium in the filtrate is precipitated by acidifying with hydrochloric acid and just neutralizing with ammonia and warming. The caustic potash or soda must be free from impuri- ties which are precipitated by ammonia and ammonium chloride. If pure caustic cannot be obtained by solution of the metal in water in a platinum or silver dish or other method, a blank deter- mination of the amount of impurities present must be made. EXERCISE 34. Analysis of Chromite (FeO,MgO)(Cr203,Alj03). Silica and Small Amounts of Manganese are Generally Present. This ore is chiefly of value for its content of chromium. The deter- mination of the amount of this element is the most important and frequently the onl}' dotermination required. For this reason, and also for convenience in determining the amount of the other elements present, a separate portion is taken for the determination of chromium. 216. Solution of Ore. — Weigh one-half gram of the finely pulverized ore and mix in a small copper or nickel crucible with 2 to 5 grams of powdered sodium peroxide. Brush off carefully the platinum or glass rod used for mixing the material. If the sodium peroxide is not pure yellow and fresh, the larger amount should be taken, as it decomposes rapidly, forming sodium carbonate. Heat the crucible with a small flame of the Bunsen burner, so that the contents are completely fused only after several minutes. Keep the material fused for about ten minutes. When cold, place the crucible in a porcelain dish and dissolve the con- tents in 50 to 100 CO. of hot water. Boil for fifteen to twenty minutes to decompose the sodium peroxide. 217. Chromium. — Filter off and wash the precipitate consisting of h'on, magnesium, manganese and some of the silica, as well as a little nickel or copper from the crucible. The filtrate vnW contain all of the chromium and aluminium and part of the silica. Acidify the filtrate with acetic acid, warm and filter off any alumina and silica which separates out. Precipi- tate the chromic acid by adding lead acetate or barium chloride and weigh as lead or barium chromate. The chromium may also be determined volu- metrically as directed in Exercise 63, page 333. This volumetric method will usually be found superior to the gravimetric method. 218. Iron. — The insoluble material filtered off from the solution of the sodium peroxide fusion may be used for the determination of iron. It is dissolved in a little sulphuric acid, the material adhering to the crucible- 164 ANALYSIS OF MINERALS. being dissolved in the same manner. Particles of urukcomposed chromite may be distinguished from the gelatinous silicic acid as a fine heavy powder which quickly settles to the bottom of the beaker on stirring the solution. If material of this kind is present it must be filtered off and again fused with sodium peroxide. The solution of the melt must be filtered and the filtrate added to the bulk of the chromium solution. The insoluble residue is dissolved in hydrochloric acid and added to the solution of the insoluble material from the first fusion. The iron may then be titrated as directed in Exercise 57, page 320. 219. Silica. — For the determination of the elements present other than chromium, a separate portion of one-half to one gram is fused with sodium peroxide, in the same manner as directed for the determination of chromium. The fused mass is dissolved in water, boiled to decompose the sodium per- oxide and the insoluble matter filtered off and washed free from chromium. The filtrate is acidified with hydrochloric acid, evaporated to dryness and heated one-half hour on the water-bath to dehydrate the silicic acid, which is filtrated off and washed. 220. Aluminium. — The filtrate will contain the aluminium and the chromium. It is neutralized with ammonium carbonate and a little hydro- gen peroxide or bromine added, and the solution boiled to reoxidize the chromium which has been reduced during the evaporation with hydro- chloric acid. The solution should be boiled in a porcelain or platinum dish to avoid solution of the glass of the beaker in the alkaline liquid. The alimiinium hydroxide is filtered off, washed free from chromium, ignited, and weighed as oxide. The chromium and aluminium may also be precipitated together and weighed as oxides, the amount of aluminium being found by deducting the weight of chromium oxide as obtained by the previous determination of this element. The filtrate from the sihca is warmed after the addition of alcohol or sulphurous acid to completely reduce the chromium. The aluminium and chromium are then precipitated with ammonia and weighed as oxides. 221. Silica. — The material insoluble in the sodium peroxide solution, which contains the manganese, magnesium, and part of the sUica, is dis- solved in warm hydrochloric acid and any undecomposed chromite filtered off and again fused with sodium peroxide. The acid solution of the insoluble matter is evaporated to dryness on the water-bath to dehydrate the silica which is filtered off, washed, and weighed, the weight being added to that of the silica already found, unless the same filter-paper was used for both precipitates. 222. Manganese. — If a copper crucible was used, hydrogen sulphide is passed through the acid filtrate from the sihca to precipitate the copper dis- solved, which is filtered off and washed. Bromine water or hquid bromine dissolved in hydrochloric acid is added to the filtrate to oxidize the iron and to decompose the excess of hydrogen sulphide. Sufficient bromine is added ANALYSIS OF CHBOMITE. 165 to color the solution distinctly. Ammonia is added until the reaction is distinctly alkaUne. The solution is warmed and the precipitate consisting of the iron and any manganese which may be present is filtered off and washed. It may be ignited and weighed, the manganese being obtained by difference, the amount of iron being known. The manganese will be present as MujO,. If the amount of manganese is small it may be determined more accurately by Ford's method. The precipitate is dissolved in a little nitric acid and the determination carried out as directed in Chapter XXVIII, page 385. 223. The magnesium will be found in the filtrate from the iron precipi- tate. It is evaporated to small bulk, precipitated as magnesium ammo- nium phosphate, and weighed as pyrophosphate. CHAPTER XIV. ANALYSIS OF SULPHIDES CONTAINING MAN- GANESE, NICKEL, COBALT, AND MERCURY. SEPARATION OF MANGANESE, NICKEL, AND COBALT. 224. Separation of Nickel and Cobalt as Sulphides. — One of the best methods of separating manganese from nickel and cobalt depends on the properties of the sulphides of these metals. The sulphide of manganese is readily soluble in dilute acids, including acetic acid. The sulphides of nickel and cobalt after precipitation are almost insoluble in all dilute acids. These two metals are not precipitated, however, from solutions acid with mineral acids, but do come down from warm acetic-acid solution. The acid solution of manganese, cobalt, and nickel from which zinc, iron, aluminium, etc., have been separated by methods already given, is concentrated to a small bulk (30 to 50 c.c). The solution is made alkaline with sodium carbonate, acidified with acetic acid and 3 to 4 grams of sodium acetate dissolved in a little water are added. The solution is heated to 70°, and hydro- gen sulphide passed until saturated. The sulphides of nickel and cobalt are filtered off and washed. The filtrate and washings are concentrated, a little ammonium sulphide and then acetic acid are added, and hydrogen sulphide again passed. A little more of the cobalt and nickel sulphides is frequently obtained. A second test of the filtrate in this manner sometimes yields a small precipitate. Another method of carrying out this separation which requires but one precipitation is as follows: The sulphate solution of man- ganese, cobalt, and nickel is transferred to a pressure-bottle of about one-half liter capacity. The solution is neutralized with ammonia, 30 c.c. ammonium-acetate solution (1 to 10), and 20 c.c. 50% acetic acid added. The solution is diluted to 300 to 400 c.c, and hydrogen sulphide passed for one to two hours. The bottle is closed, placed in a water-bath containing cold water which is 166 SEPARATION OF MANGANESE, NICKEL, AND COBAU 167 brought to a boil in one hour. The bottle is allowed to cool in the water-bath to 50°. The precipitated sulphides are filtered off and washed with water containing acetic acid and hydrogen sulphide. 225. Electrolytic Separation of Manganese from Cobalt and Nickel. — Small amounts of manganese can readily be separated from cobalt and nickel electrolytically. The sulphate solution of the three metals should be treated with 30 c.c. of a cold-saturated ammonium-sulphate solution, and 30 to 50 c.c. of ammonia (sp. gr. 0.96). The solution is diluted to about 150 c.c, and elec- trolyzed with a current of about 0.7 anipere at the ordinary tem- perature. The manganese separates as dioxide on the anode. This precipitate cariies down small amounts of nickel and cobalt unless these metals are deposited before the manganese dioxide forms. Small amounts of nitrates and chlorides hinder the pre- cipitation of nickel and cobalt. The deposited metal is washed with water without interrupting the current. Finally it is washed with alcohol, dried in the air-bath, and weighed. SEPARATION OF NICKEL AND COBALT. 226. Precipitation of Cobalt as Tripotassium Cobaltic Nitrite, Co(N02)3-3KN02, separates it from nickel as well as from man- ganese and zinc. The alkaline earth-metals and iron must be absent. The method is satisfactory for mixtures of nickel and cobalt in all proportions, but is especially suited for the separation of a little cobalt from much nickel. The solution of the metals is evaporated to small bulk, and neutralized with potassium hydrox- ide. About 5 grams of potassium nitrite dissolved in water are added. Acetic acid is added until the solution is acid and nitrous fumes are evolved. It is allowed to stand for at least twenty- four hours. A portion of the supernatant liquid is taken out and tested for cobalt by adding more potassium nitrite. If a precipi- tate forms after long standing, more potassium nitrite must be added to the main solution. The precipitate is filtered off and washed with a solution of potassium acetate (1 to 10) containing a little potassium nitrite. The precipitate may be dissolved in hydrochloric acid and the cobalt reprecipitated with caustic soda 168 Analysis of minerals. and weighed as oxide. It may also be dissolved in dilute sul- phuric acid and the solution evaporated to dryness in a weighed platinum dish or crucible. The residue consisting of the sulphates of cobalt and potassium is heated to dull redness and weighed. If weighed in this manner the potassium acetate and nitrite must finally be washed out with a little distilled water. The precipi- tate may also be dissolved in hot dilute sulphuric acid, the solution concentrated, treated with excess of ammonia, ammonium sulphate added, and the cobalt determined electrolytically. If the NICKEL has previously been weighed with the cobalt, it may be obtained by difference. Otherwise the nickel may be determined in the filtrate from the cobalt. For this purpose the solution is treated with hydrochloric acid and the nickel precipi- tated with caustic soda and bromine water. The nickelic hydrox- ide may be filtered off, washed, and reduced to metallic nickel by heating in a stream of hydrogen, or it may be dissolved in sulphuric acid and a little sulphurous acid and determined electrolytically. 227. Separation of Cobalt and Nickel by Means of Nitroso-/?- Naphthol. — The cobalt compound Co(CioH50(NO))3 is insoluble in hydrochloric acid, while the corresponding nickel compound is soluble. Nitric acid is expelled by evaporation with sulphuric acid. The solution is diluted, 5 c.c. hydrochloric acid and a freshly prepared hot solution of nitroso-,9-naphthol in 50% acetic acid added, until further addition produces no precipitate. The solution is digested in a warm place for a few hours. The precipitate, consisting of cobalt nitroso-/?-naphthol as well as considerable of the reagent, is filtered off and washed, first with cold and then with warm 12% hydrochloric acid until all the nickel has been washed out. It is then washed with hot water until free from acid. The precipitate is dried and placed in a weighed Rose crucible. A little pure recrystallized oxalic acid is added and the paper burned, at first with low heat and then with the fuU flame of the Bunsen burner. When the carbon is com- pletely burned the precipitate is reduced to metallic cobalt by heating in a stream of hydrogen. The filtrate is evaporated after the addition of sulphuric acid. The NICKEL may then be precipitated as hydroxide and weighed as oxide, or it may be determined electrolytically. ANALYSIS OF SMALTITE. iij^., EXERCISE 35. Analysis of Smaltite (Co,Ni,Fe)(S,As),. Silica and Small Amounts of Manganese Are Usually Present. 228. Solution. — One gram of the finely powdered material is weighed out and transferred to an Erlenmeyer flask of about 250 c.c. capacity. This is best done by means of a small tube sealed at one end. The tube containing about one gram of the smaltite is carefully weighed. It is then inserted into the neck of the flask which is tilted. On raising the flask up- right and tapping the tube gently the smaltite is deposited on the bottom of the flask. The tube is carefully withdrawn and weighed. Twenty c.c. of aqua regia, made by adding to 3 volumes of concentrated nitric acid 1 volume of concentrated hydrochloric acid, are added. A small funnel is placed in the neck of the flask and, after being shaken and standing until the violent action has ceased, the flask is warmed gently on the water-bath until the mineral is entirely decomposed. If sulphur floats on top of the solution, liquid bromine should be added, a few drops at a time, and the solution warmed until the sulphur is entirely oxidized. 229. Silica. — The solution is poured into a porcelain- dish, the flask and funnel rinsed with water and after the addition of concentrated sulphuric acid, evaporated on the hot plate until fumes of sulphuric acid are copiouslj evolved. The solution is cooled and diluted with water. The siliceous residue is immediately filtered off, washed, and weighed. 230. Arsenic, Antimony, and Tin. — A little sulphur dioxide is added to the filtrate, which is warmed to reduce arsenic to arsenous acid. The excess of sulphur dioxide is expelled by boihng. Hydrogen sulphide is passed through the warm solution until no more precipitate forms. The arsenious sulphide is filtered off on a Gooch crucible and the filtrate again treated with sulpimr dioxide and hydrogen sulphide to insure com- plete precipitation of the arsenic. After washing the arsenious sulphide the sulphur is extracted by treatment with alcohol and carbon disulphide. The precipitate is then dried and weighed. The color of the precipitate indicates the absence of lead, bismuth, and copper and all but a trace of antimony. The absence of the latter element as well as tin may be shown by treating the weighed precipitate with a concentrated solution of ammo- nium carbonate. The sulphides of antimony and tin remain on the filter and may be dissolved in hydrochloric acid and separated by the methods given in Chapter XI, page 134. The arsenic must also be repre- cipitated and weighed. 231. Iron. — The filtrate from the hydrogen &uiphide precipitate is evaporated with the addition of nitric acid to oxidize the iron. The solution is made alkaline with ammonia, warmed, and the precipitate filtered off 170 Analysis of minerals. B and washed two or three times. The filtrate will contain most of the cobalt and nickel, while the precipitate will contain all of the iron and a small amount of the nickel and cobalt. It is dissolved in dilute hydrochloric acid and the paper well washed. The solu- tion is evaporated to a syrupy consistency and transferred with hydrochloric acid of sp. gr. 1.124 to the bulb A of the apparatus shown in Fig. 24, on p. 161, and the extraction with ether carried out as there described. The apparatus shown in Fig. 25 may also be used. Before transferring the solution to the bulb D, a few cubic centimeters of ether are poured into the bulb and allowed to flow into B. By warming the bulb B with th« hand and opening and closing the stop-cock C, a partial vacuum is created in B, A and C are closed and the solution transferred to D, rinsing the beaker with hydrochloric acid of sp. gr. 1.124 until the bulb is filled to the mark on D, making the volume 60 c.c. Place the bulb B in cold water, open the stop-cock A, and allow the liquid in D to flow into B. Close A and fill D with ether (100 c.c.) and allow it to flow slowly into B. Mix the liquids gradually, placing the bulb in cold water from time to time to prevent a large rise in temperature. Finally close A. Shake well, opening A occasionally to relieve the pressure, then place the apparatus in cold water for about five minutes to allow the liquids to separate. Allow the lower liquid, which is very nearly free from iron, to flow into a beaker. Introduce about 10 c.c. of hydrochloric acid (sp. gr. 1.1) into the bulb D. Allow it to flow into B, shake, and after standing a few minutes allow the lower liquid to flow into the beaker. This liquid, together vnth the filtrate from the ferric hydroxide, will contain all of the cobalt, nickel, manganese, and aluminium and a little of the iron. Draw off the ether solution of the iron into another beaker and repeat the ether extraction of the iron, the nickel-cobalt solution being poured back again into the bulb D. Rinsing the beaker will be unnecessary. The ether solution of the ferric chloride should be poured into a Florence or distilling flask connected with a condenser by means of cork stoppers and the beaker rinsed with water. The ether is distilled off on the water- bath, care being taken that the ether is not placed near a Bunsen burner or other flame. The ferric cliloride is transferred to a beaker and the iron precipitated with ammonia and weighed as oxide, or reduced with zinc and determined volumetricaUy. 232. Aluminium and Chromium. — The combined cobalt nickel solu- tion is warmed to expel the ether and a little bromine or hydrogen peroxide added. The solution is neutralized 'Vi'ith ammonia, a little ammonium carbonate added, and boiled. A small precipitate of iron is usually obtained and any aluminium or chromium Dresent is precipitated, Fig. 25. AA'ALYSIS OF SMALTITE. 171 The precipitate, wliicli is usually quite small, is filtered off, redissolved, and reprecipitated. In the absence of aluminium and chromium, it is added to the main portion of the iron and weighed. If aluminium and chromium are present, the metals must be separated by methods already given. 233. Manganese, Cobalt, and Nickel. — About 10 c.c. of concentrated sul- phuric acid are added to the filtrate, which is evaporated on the hot plate until the hydrochloric acid is entirely expelled, as indicated by the evolution of dense fumes of sulphuric acid. After cooling, water is added and the solution is neutralized with ammonia. 30 to 50 c.c. of ammonia (sp. gr. 0.96) is then added and the nickel and cobalt determined electro- * lytically. The trace of manganese which may be present will be con- verted into the dioxide and deposited on the anode. It may be ignited and weighed as MujOj. The deposit of nickel and cobalt after being weighed is dissolved in warm dilute nitric acid and the solution evaporated to about 50 c.c. It is neutralized with caustic potash and 5 grams of potassium nitrite dis- solved in water are added. Acetic acid is added until the solution is acid and the slight precipitate produced by the caustic potash dissolved. After standing for twenty-four hours, the clear liquid is tested for cobalt by adding more potassium nitrite. If cobalt is present, the entire solution must be allowed to stand for some time after the addition of potassium nitrite. The precipitate is filtered off and washed with a 10% solution of potassium acetate containing a little potassium nitrite. It is then dis- solved in hot dilute sulphuric acid, excess of ammonia added, and the cobalt determined electrolytically. 334. Sulphur is determined in a separate portion of 0.5 gram, which is dissolved in aqua regia as already directed, or in fuming nitric acid. The reagents must be absolutely free from sulphur. If necessary, a blank determination should be made side by side with the determination of the mineral. Instead of the smaltite, a httle pure sodium or potassium chloride should be added to the blank. After the expulsion of the nitric acid with hydrochloric acid and filter- ing off the silica, the solution is ready for the precipitation of the sulphuric acid by means of barium chloride. If much sulphur is present, it is well to add a httle sodium or potassium chloride to the nitric acid solution before evaporation to prevent loss of sulphuric acid. The evaporation should be conducted on a hot plate or water-bath which is wide enough to keep the products of combustion of the illuminating-gas away from the solution, otherwise sulphur dioxide may be absorbed. Sulphuretted hydrogen or ammonium sulphide must also be absent from the atmosphere. The solution is diluted to 300 or 400 c.c, heated to boiling, and barium chloride solution added slowly with constant and vigorous stirring until no more precipitate forms on adding another drop of the reagent. After digesting the precipitate on the hot plate until the solution is clear, it is 172 ANALYSIS OF MINERALS. filtered off, washed with hot water, ignited, and weighed. If the color of the ignited precipitate is not pure white, but reddish, iron has been carried down. Several grams of sodium carbonate are added to the precipitate and after fusion the melt is dissolved in water and the barium carbonate filtered off and washed. The sulphuric acid in the filtrate is precipitated by barium chloride after acidifying with hydrochloric acid. The error due to a slight reddish color of the precipitate may generally be neglected. EXERCISE 36. Analysis of Pyrite, FeSj, Arsenopyrite, Fe(S,As)2, or Chalcopyrite, CuFeS^. Small Amounts of Silica, Arsenic, and Antimony May be Present. 235. Sulphur Available for Sulphuric-acid Manufacture. — As the natural sulphides are frequently used as a source of sulphur, especially in the manufacture of sulphuric acid, the determination of this element is frequently carried out for the purpose of ascertaining the amount of sulphur available under the conditions of manufacture of sulphuric acid. For this purpose the determination is carried out as follows: One-half gram of the powdered sulphide is treated in an Erlenmeyer flask of about 250 c.c. capacity with 10 c.c. of a mixture of 3 volumes of concentrated nitric acid and 1 volume of concentrated hydrochloric acid. The mouth of the flask is closed with a small funnel and the solution is gently heated on the water-bath until the material is decomposed. If unoxidized sulphur separates, a few crystals of potassium chlorate are added from time to time and the heating continued until the globules of sulphur have disappeared. Liquid bromine may be added for the same purpose. The solution is transferred to a porcelain dish and evaporated to dry- ness after the addition of one-half gram of pure sodium or potassium chloride. If the evaporation is conducted on the water-bath only, the addition of the alkali chlorides may be omitted. Five c.c. of concentrated hydro- chloric acid are added and the solution again evaporated to dryness. The residue is dissolved in a few cubic centimeters of dilute hydrochloric acid and 100 c.c. of hot water. The insoluble residue is filtered off on a small paper and well washed with hot water. This residue consists of silica, undecomposed silicates, and the sulphates of lead, barium, and calcium, if these elements are present. This material is reserved for analysis to obtain the percentage of total sulphur present. The filtrate is neutralized with ammonia and a considerable excess added. It is warmed for ten to fifteen minutes at 60°-70°, but not to boiUng. li the solution does not smell strongly of ammonia more must be added. The FERRIC HYDROXIDE is filtered off and washed twice by decantation with very hot water. The precipitate is transferred to the filter-paper, which should not be completely filled, so that it may be well stirred up with the stream of hot water. The precipitate must be washed until the wash- ANALYSIS OF PYRITE. 173 water no longer gives a precipitate with barium chloride. As basic sulphates of iron may be present in the ferric hydroxide, the precipitate is dried and fused with sodium carbonate. The melt is dissolved in water and the solu- tion tested for sulphuric acid. The filtrate from the iron is acidified with hydrochloric acid, evaporated to a bulk of about 300 c.c, heated to boiling, and barium chloride solution added slowly with constant stirring until a considerable excess is present. Twenty c.c. of a 10% solution is usually more than sufficient for 0.5 gram of pyrite. After digesting hot until the solution is clear, the barium sul- phate is filtered off, ignited, and weighed. The sulphur may also be determined by the method given in Ex. 35. It may also be determined by the method of Allen and Bishop * as fol- lows: 1.3736 grams of the sulphide ground to pass the 80-mesh sieve are put in a dry 300-c.c. Jena beaker, 10 c.c. of a mixture of 2 parts by volume liquid bromine and 3 parts carbon tetrachloride (which is free from sulphur compounds) are added and the beaker covered with a watch-glass. After standing fifteen minutes at room temperature with occasional shaking, 15 c.c. nitric acid (sp.gr. 1.4) are added and the mixture is allowed to stand fifteen minutes longer at room temperature with occasional shakiug. The beaker is then placed on an asbestos board on top of the steam-bath and allowed to remain there until all action has ceased and most of the bromine has been volatiHzed. It is then placed within the ring of the bath and the solution evaporated to dryness. Ten c.c. hydrochloric acid (sp.gr. 1.2) are next added and, after shaking to mix thoroughly, the solution is again evaporated to dryness; when completely dry the SiOj is dehydrated by heating in an air-bath at 100° C. for several hours. The iron is then reduced as follows: Four c.c. of hydrochloric acid (sp.gr. 1.2) are now added to the dehydrated mass, followed after five minutes by 100 c.c. of hot water. The cover and sides of the beaker are [carefully washed down with hot water and the beaker again covered. The mixture is gently boiled for five minutes to insure complete solution of all sulphates. After the solution has partially cooled by standing for about five minutes, 0.2 to 0.3 gram of powdered aluminium is added and the beaker gently shaken until the iron color has disappeared, showing complete reduction. Cool the solution and filter off the excess of aluminium powder, wash the beaker carefully and wash the residue on the paper at least nine times with hot water. To the filtrate add 6 c.c. more hydrochloric acid (sp.gr. 1.2) and enough cold water to make the volume 1600 c.c. After stirring to mix well, the beaker is covered with a large watch-glass and the sulphuric acid precipitated by adding 125 c.c. of 5% barium chloride solution. The reagent must be added at a rate not *EighthInt. Cong, of App.Chem., 1912, vol. 1, p. 33. 174 ANALYSIS OF MINERALS. exceeding 5 c.c. per minute. This is accomplished by attaching a capil- lary tube to a dropping funnel, the tube beiag bent so that it passes under the watch-crystal to the center of the beaker. The stop cock of the funnel is regulated so that 5 c.c. of the barium chloride will drop per minute into the sulphuric acid solution. The solution is not stirred while the barium chloride is being added but after all is in, the supernatant liquid is well mixed by gentle stirring. After the barium sulphate has settled, preferably by standing overr night, it is filtered off on paper or on a Gooch crucible. The weight of the barium sulphate multiplied by 10 equals the per cent of sulphur in the ore. 236. Total Sulphur. — To obtain the total percentage of sulphur the material insoluble in aqua regia is fused with sodium carbonate. If lead is absent, the fusion may be conducted in a platinum crucible, otherwise a porcelain crucible must be used. The melt is dissolved in hot water, fil- tered, the residue well washed, and the sulphuric acid in the filtrate pre- cipitated with barium chloride and weighed. The total percentage of sulphur may also be determined as follows: Fuse 0.5 gram of the sulphide with 10 grams of a mixture of two parts sodium carbonate and one part potassium nitrate. The material should be well mixed by stirring with a platinum or glass rod, which is then cleaned with a camel's-hair brush. The material is gradually brought to fusion and maintained in the fluid state for some time. The fused mass is extracted with ho't water. If lead is present it is precipitated by passing carbon dioxide through the solution. The insoluble material is filtered off and washed with hot water containing a little sodium carbonate, first by decanta- tion and then on the filter-paper. The filtrate is acidified with hydrochloric acid and evaporated to dry- ness to remove nitric acid. The residue is dissolved in hot water to which a little hydrochloric acid has been added. Any insoluble material present is filtered off and well washed. The sulphuric acid in the filtrate is deter- mined by precipitation with barium chloride. 237. Arsenic. — For the determination of arsenic, antimony, and copper, a portion weighing from 0.5 to 2 grams is taken. If the percentage of these elements is small, the .larger amount is taken. The material is treated cold in a beaker with 20 to 30 c.c. concentrated hydrochloric acid. The solution is agitated from time to time. One or two c.c. of liquid bromine or concentrated nitric acid are then added and the solution is heated, gently at first and then more strongly, until the bromine or nitrous fumes are expelled. The arsenic is then distilled and determined as directed in Chapter XI, p. 138. Ferrous sulphate or chloride is used as the reducing- agent. 238. Antimony. — The contents of the flask are rinsed into a beaker, the solution diluted to 200 to 300 c.c, warmed, and hydrogen sulphide passed. The precipitate, consisting of the sulphides of copper and antimony. ANALYSIS OF CINNABAR. 175 is filtered off and washed with water contaiaiag hydrogen sulphide. The antimony sulphide is dissolved in a little sodium-sulphide solution and the residue washed with water containing a Uttle sodium sulphide. The sodium sulphide may be made by saturating with hydrogen sulphide one-half of a caustic soda solution and then adding the other half. The antimony is precipitated by neutrahzing the sodium sulphide solution with hydrochloric acid and passing hydrogen sulphide. It may be ignited and weighed as the tetroxide after treatment with nitric acid. 239- Copper. — The copper sulphide is dissolved in a little nitric acid and determined electrolytically. 240. Iron. — Iron is usually determined volumetrically. The ammonia precipitate from the aqua-regia solution made for determining sulphur contains nearly all of the iron. A small amount is frequently left in the insoluble residue as ferrous silicate. On fusing this residue with sodium carbonate the ferrous silicate is decomposed and the insoluble ferric hj-drate may be filtered off. These two precipitates may be dissolved in hydrochloric acid and the iron titrated, as given in Chapter XXIV, page 320. Both iron and sihca may be determined in the portion fused with sodium carbonate and potassium nitrate for the determination of total sulphur. All of the iron will be present in the precipitate filtered off from the water solution of the fused material. It may be dissolved in hydrochloric acid, precipitated with ammonia and weighed as oxide, or it may be determined volumetrically. 241. Silica. — The silica constitutes the insoluble material filtered off after acidifying the sodmm carbonate solution, evaporating to dryness and dissolving in hydrochloric acid and water. It is ignited and weighed. EXERCISE 37. Analysis of Cinnabar, HgS. Iran, Manganese, Copper, Aluminium, and [Calcium may also be present. Sulphur is determined by the method given on p. 167. 242. Mercury. — One gram of the mineral is treated with 20 to 30 c.c. concentrated hydrochloric acid to which .40 c.c. water is added. One to two grams of potassium chlorate are added in small portions and the solu- tion warmed until the cinnabar is entirely decomposed. The solution is evaporated to dryness on the water-bath and the residue dissolved in water and hydrochloric acid. Any insoluble material is filtered off, ignited, and weighed. A little sulphurous acid is added to reduce any iron present to the ferrous condition. The solution is warmed to expel the excess of sulphurous acid and hydrogen sulphide passed until no more precipitate is formed. If copper is absent, the mercuric sulphide is filtered off on a Gooch crucible which has been dried at 100". Sulphur is extracted by 175a ANALYSIS OF MINERALS. washing the precipitate with alcohol and carbon disulphide. The precipi- tate is then dried at 100° and weighed. If copper is present the precipitate, after being washed, is dissolved in aqua regia and the chlorine expelled by boiling after the addition of dilute hydrochloric acid. When entirely free from chlorine or nitric acid, phos- phorous acid is ad led and the solution is allowed to stand in a warm place for at least twelve hours. The prefipitated mercurous chloride is filtered off on a Gooch crucible, washed with hot water, dried at 100°, and weighed. 243. Copper. — The filtrate is evaporated to small bulk after the addition of a few cubic centimeters of dilute sulphuric acid to expel the hydrochloric acid, diluted with water, and the copper determined electrolytically. The copper may also be precipitated as sulphide which is dissolved in 2 or 3 c.c. of concentrated nitric acid, the solution diluted and electrolyzed. 244. Mercury may also be determined very accurately in a separate portion as follows: A combustion-tube 30 to 45 cm. long and 10 to 15 mm. wide is taken. The end is sealed, then a layer of magnesite in large pieces, chalk, or sodium bicarbonate is introduced until the tube is filled for about 10 cm. Then a layer of recently ignited lime is introduced, occupying about 5 cm., then an intimate mixture of the ore with lime for about 6 cm. This mixture should be made by grinding the constituents together in a mortar. The mortar should be rinsed with lime, which is next introduced, occupying about 5 cm. Finally a layer of about 5 cm. of lime is introduced, and a loose plug of -asbestos inserted. The tube is then drawn out and bent at right angles. It is then tapped gently so as to produce a free passage for the gases the entire length of the tube, and is then placed in a combustion-furnace with the drawn-out end dipping into a flask. The burners of the furnace are lighted so that the lime near the drawn-out end of the tube is first heated to redness. The burners are then lighted in order until the magnesite in the closed end of the tube evolves carbon dioxide, which sweeps out the last traces of mercury vapor. While the tube is still hot it is cut off at the point where it was drawn out and bent and the globules of mercury washed out of the portion cut off by means of a jet of water. By shaking the flask the mercury is brought together into a single drop. The water is decanted and the mercury transferred to a weighed porcelain crucible. The water is absorbed in filter-paper and the mercury dried in a desiccator and weighed. 245. The Iron, Aluminium, Manganese, and Calcium will be present in the filtrate from the mercuric sulphide. If the amount of iron and alaminium is small, these metals may be precipitated as hydroxides and weighed as oxides. The separation from manganese is effected in the manner described in Exercise 33, p. 149, except that an excess of ammonia must be avoided or the aluminium will be redissolved. The manganese is determined as directed in the same exercise. The filtrate from the manganese is acidified with hydrochloric acid and evaporated to a bulk of 200 or 300 c.c. It is neutral' 'led with ammonia, warmed nearly to boiling, and a little ammonium ANALYSIS OF CINNABAR. 1756 oxalate added. After digesting for some time, the calcium oxalate is filtered off and washed with hot water. The moist precipitate is placed in the platinum crucible, the paper burned in the usual manner, and the precipitate ignited, finally, over the blast-lamp and weighed as oxide. CHAPTER XV. ANALYSIS OF CARBONATES CONTAINING CAL- CIUM, BARIUM, STRONTIUM, AND MAGNESIUM. SEPARATION OF CALCIUM, BARIUM, AND STRONTIUM FROM THE ALKALI METAI,S AND FROM EACH OTHER. 246. Calcium, Barium, and Strontium may be separated from magnesium and the alkali metals by precipitation as carbonates, by ammonia and ammonium carbonate. A sufficient amount of ammonium salts must be present to prevent the precipitation of magnesium. It is very seldom that this method of separation is used because the occurrence of these three elements together is very rare, although recent work has shown that traces of barium and strontium occur in rocks far more frequently than was for- merly supposed. Excellent methods of separation of each of the alkaline-earth metals from magnesium and the alkalies are avail- able, by which the metal is left in the best condition for weighing. SEPARATION OF CALCIUM. 247. Separation of Calcium as Oxalate. ^Calcium is usually separated from magnesium and the alkali metals by means of ammonium oxalate. The precipitation is carried out in a hot solution. Sufficient ammonium oxalate must be present to con- vert both calcium and magnesium into oxalates, since calcium oxalate is slightly soluble in magnesium chloride. Enough ammonium chloride must be present to prevent the precipitation of magnesium hydroxide when the solution is made alkaline. A sufficient amount is usually present from the precipitation of iron and aluminium. If an excessive amount has been introduced, it must be removed by evaporating the solution to dryness and vola- tilizing the ammonium salts. The solution may also be boiled down to small bulk and the ammonium salts decomposed by the addition of nitric acid. It is generally sufficient to treat in this 176 SEPARATION OF CALCIUM, BARIUM, AND STRONTIUM. 177 manner the filtrate from the second, third, and subsequent precipi- tations of the iron and aluminium. If the ammonium salts from the entire solution have been expelled, the residue is dissolved in hydrochloric acid and the solution neutralized with ammonia. If a precipitate forms it is dissolved in acid, and the solution again neutralized. This is repeated until a precipitate of magnesium hydroxide is no longer formed on making the solution alkaline. The first precipitate of calcium oxalate will invariably contain some magnesium, and also some sodium, which cannot be washed out. It must therefore be redissolved in hydrochloric acid, and reprecipitated by neutralizing with ammonia and adding a little ammonium oxalate. The second precipitate will be free from magnesium and sodium. If strontium is present it will be almost completely precipitated with the calcium. The two metals may be weighed together as oxides after ignition of the oxalate, and separated by methods given later. If barium is present to the extent of 3 or 4 mg., it will not contaminate the calcium after the second precipitation. If more than 3 or 4 mg. of barium are present, the calcium cannot be separated as oxalate. 248. Separation of Barium and Strontium as Sulphates. — Barium and strontium may be separated from magnesium and the alkali metals by precipitation as sulphate. If strontium is present an equal volume of alcohol must be added to insure com- plete precipitation. The precipitate must also be washed with alcohol. Barium sulphate frequently carries down a small amount of the alkalies which cannot be washed out. It must, therefore, in very exact analyses, be redissolved in a little warm concen- trated sulphuric acid and reprecipitated by diluting with water. 249. Separation of Calcium as Sulphate. — Calcium may also be separated from magnesium and the alkalies by precipitation as the sulphate. The solution must be evaporated until free hydro- chloric acid and water are expelled. The residue is dissolved in alcohol, and a slight excess of concentrated sulphuric acid added. After digesting for some time the solution is filtered and the pre- cipitate is washed with absolute alcohol until free from acid. A small amoimt of magnesium sulphate is then washed out by means of 35 to 40% alcohol. The alcohol may be expelled from the filtrate and the magnesium precipitated as phosphate. If barium 178 ANALYSIS OF MINERALS. and strontium are present they will be precipitated with the calcium. SEPARATION OF BARIUM, STRONTIUM, AND CALCIUM. 250. Precipitation of Barium as Chromate. — ^Two absolutely reliable methods for the separation of barium, strontium, and cal- cium have been devised by Fresenius. The metals are precipi- tated as carbonates with ammonia and ammonium carbonate. If the separation from the magnesium has been effected by sulphuric acid the precipitate is fused with sodium carbonate, the melt treated with hot water, and the carbonates of the alkaline-earth metals filtered off and well washed. The precipitate is dissolved in the least amount of dilute hydrochloric acid, and the excess removed by evaporation. The chlorides are dissolved in 300 c.c. of water, 6 drops of acetic acid (sp. gr. 1.065) and a few drops of ammonium acetate added. The solution is heated, and 10 c.c. of a neutral 10% ammonium chromate solution added. After allowing the precipitate to settle and the solution to cool for one hour, decant on a small filter and wash by decantation two to three times with water containing ammonium chromate. Finally transfer the precipitate to the paper and wash until the filtrate gives no precipitate with ammonia and ammonium carbonate. About 100 c.c. of wash-water should be used. The washing is continued with pure warm water, until the filtrate gives only a slight reddish coloration with silver nitrate. About 110 c.c. will be required. The precipitate which contains all of the barium and a little of the strontium is washed into the precipitating-dish, the portion adhering to the filter-paper dissolved in a little warm dilute nitric acid, and the paper well washed. About 2 c.c. nitric acid (sp. gr. 1.20) should be used. The solution is diluted to 200 c.c. and heated. 5 c.c. of a 30% ammonium acetate solution is added very gradually. Ammonium chromate solution (10%) is then added until the odor of acetic acid has entirely disappeared. After standing one hour, the supernatant liquid is decanted through the filter or on a Gooch crucible, and the precipitate is digested with hot water which is allowed to eool. The precipitate SEPARATION OF CALCIUM, BARIUM, AND STRONTIUM. 179 Is then brought on the filter-paper or crucible and washed with cold water until the filtrate gives a scarcely perceptible test for chromic acid with silver nitrate. 251. Weighing Barium as Chromate. — ^The now pure barium chromate is removed from the paper, the latter burned, and the whole gently ignited in a platinum or porcelain crucible. If the heat is too great the chromic acid will be reduced and the precipitate will become green. The portions of the precipitate adhering to the filter-paper will also be reduced during the incineration of the latter. On gently heating the precipitate, the chromium will be reoxidized. If a Gooch crucible has been used the precipitate may finally be washed with alcohol, dried on the hot plate, and weighed. 252. Separation of Calcium and Strontium as Nitrates. — One c.c. of nitric acid is added to the filtrate from the barium, which is concentrated and the calcium and strontium precipitated by ammonia and ammonium carbonate. If only strontium is present the precipitate of strontium carbonate may be weighed or it may be dissolved in hydrochloric acid, precipitated by the addition of sulphuric acid and alcohol, and weighed as sulphate. If only calcium is present the precipitation as carbonate may be omitted and the calcium immediately precipitated as oxalate and weighed as oxide. If both strontium and calcium are present the mixed carbo- nates are dissolved in nitric acid and the solution evaporated to dry- ness in a small porcelain dish and heated for some time in an air- bath at 130°. The temperature may rise to 180° without decom- posing the nitrates. The dried material is pulverized and treated five times with 5-c.c. portions of a mixture of equal parts of abso- lute alcohol and ether and the solution decanted each time into a small flask. The calcium nitrate is dissolved by the alcohol and ether in which the strontium nitrate is insoluble. After expelling the ether and alcohol from the residue, it is dissolved in water, evaporated to dryness, and again heated for some time at 130° in the air-bath. The mass is pulverized and transferred to the flask with about 15 c.c. of the ether-alcohol. The flask is corked and allowed to stand with occasional shaking for twenty-four hours. The solution is then filtered through a small filter, the strontium 18(J ANALYSIS OP MINERALS. nitrate washed by decantation and then on the filter with 5-c.c. portions of the ether-alcohol about ten or twelve times. The stron- tium nitrate in the flask, on the paper, and in the dish, if it was not all transferred, is then dissolved in a little hot water and the strontium precipitated as sulphate by the addition of sulphuric acid and alcohol. The calcium in the ether-alcohol solution may be precipitated as sulphate and washed with absolute alcohol. It is ignited at a low red heat and weighed as calcium sulphate. The ether and alcohol may also be distilled off, the calcium nitrate dissolved in water, and the calcium precipitated as oxalate and weighed as oxide. As BAEiUM as well as strontium nitrate is insoluble in the ether-alcohol mixture, the carbonates of the three metals may be converted into nitrates and treated in the manner directed for the separation of calcium and strontium. The nitrates of barium and strontium may then be dissolved in water and the barium separated from the strontium by two precipitations of the barium as chromate. Gooch and Browning have suggested the use of amyl alcohol in place of the mixture of absolute alcohol and ether. As the amyl alcohol boils at 138°, it may be rendered anhydrous by boiling out the water. The preparation of abso- lute alcohol or ether is quite laborious. EXERCISE 38. Analysis of Dolomite, (CaO,MgO)C02. Small Amounts of Silica, Iron, Aluminium, and Manganese Are Usually Present. 253. Solution of the Dolomite. — One gram of the powdered material is weighed out, placed in a beaker (about 100 c.c), and dissolved in about 15 c.c. dilute hydrochloric acid. The beaker should be covered with a watch-crystal and the acid poured slowly do'vs'n the lip to prevent loss by spattering. The beaker is warmed on the hot-plate or water-bath until carbon dioxide ceases to be evolved. The watch-crystal is rinsed with a little water and the open beaker heated on the water-bath until the solu- tion is evaporated to a syrupy consistency. It is then stirred with a glass rod until dry. The residue is heated at least one-half hour on the water- bath and then dissolved in water and a little hydrochloric acid and the solution heated for a few minutes. ANALYSIS OF DOLOMITE. 181 iS4. Silica. — The. siliceous residue is filtered off on a small paper and washed with hot water until free from clilorides. The moist paper is placed in a weighed platinum crucible, burned in the usual manner, and the siHceous residue finally ignited for a few minutes over the blast-lamp and weighed. The silica is then volatilized by moistening the weighed precipitate with water, adding 2 or 3 c.c. of hydrofluoric acid and a drop of sulphuric acid, warming until the precipitate is dissolved, and then evaporating to dryness on the hot-plate and igniting. The hydrofluoric acid must leave no residue on evaporation. Finally the residue in the crucible is ignited over the blast- lamp and brought to constant weight. The weight of the residue must be subtracted from the weight of the siliceous residue to obtain the weight of silica. The residue is dissolved in hydrochloric acid and added to the filtrate from the silica. 255. Iron and Aluminium. — Ten c.c. of concentrated hydrochloric acid are added to the acid solution, which should not exceed 200 c.c. It is neutrahzed with filtered ammonia while stirring vigorously and heated nearly to boiling on the hot-plate. If an excess of ammonia is present, very dilute hydrochloric acid is added until the faintest odor of ammonia can be detected over the hot solution or a moist piece of red litmus-paper held over the solution is slowly turned blue. The neutralization and diges- tion should not occupy more than one-half hour and a Jena beaker should be used. The precipitate of iron and aluminium is filtered off and if more than a few flakes are present, it is dissolved in a little warm dilute hydro- chloric acid, the paper washed a few times with hot water, and the solution again neutralized with ammonia, digested on the hot-plate for a short time, and filtered. The precipitate is washed free from chlorides with hot water, the moist paper transferred to the platinum crucible and burned. The precipitate is heated with the blast-lamp for a few minutes and weighed. As this precipitate seldom weighs more than a few miUigrams, the separa- tion of the iron and aluminium is not necessary. 256. Manganese. — If the combined filtrates from the iron and aluminiimi exceed 300 c.c, the solution should be acidified with hydrochloric acid and concentrated on the hot-plate. To precipitate the manganese the solu- tion is made alkaline with ammonia and bromine water or hydrogen per- oxide is added. A few cubic centimeters of hydrogen peroxide are sufficient, while enough bromine water, must be added to produce a yellow color in the solution. It is heated nearly to boihng, ammonia added if necessary to make it alkahne, the precipitate filtered off on a small paper and washed with hot water. The moist paper is burned in the crucible and the manganese weighed as Mn^O,,. 257. Calcium. — The filtrate is again concentrated on the hot-plate to about 250 c.c. after acidifying with hydrochloric acid. The calcium is precipitated by making the solution alkaline with ammonia and adding to the hot solution with vigorous stirring 30 to 40 c.c. of a 5% solution of 132 ANALYSIS OP MINERALS. ammonium oxalate. The solution is digested on the water-bath for about one hour or until the precipitate has settled, leaving a clear supernatant liquid. This is decanted through a filter-paper and the precipitate in the beaker is washed two or three times by decantation with hot water. About 50 c.c. of water are used each time. The precipitate is well stirred up each time and allowed to settle. A little warm dilute hydrochloric acid is poured over the precipitate on the paper and allowed to flow into the beaker con- taining the main portion of the precipitate. The calcium is reprecipitated by neutrahzing the hot solution with ammonia while stirring vigorously. A few cubic centimeters of amtmonium oxalate are added and the precipi- tate digested on the hot-plate as before. Finally the calcium oxalate is filtered off and washed with hot water until free from chlorides. The moist precipitate may be placed in the platinum crucible, gently heated until dry, and the paper burned, then heated more strongly until the carbon is completely incinerated, and finally with the blast-lamp for five to ten minutes. After weighing, the heating with the blast-lamp and weighing are repeated until the weight is constant. 258. Magnesium. — The filtrates from the calcium are combined and, after acidifjdng with hydrochloric acid, evaporated on the hot-plate to 200 or 300 c.c. Excess of disodium phosphate or sodium ammonium phosphate solution and a moderate excess of ammonia are added with stir- ring. After being allowed to stand for twenty-four hours the solution is decanted quite completely from the precipitate, which is dissolved in a few cubic centimeters of dilute hydrochloric acid. A few drops of phos- phate solution are added, then, ammonia with stirring, until a moderate excess is present. The volume of the solution should now not be greater than 50 to 75 c.c. After standing from twelve to twenty-four hours, the precipitate is filtered off and washed with dilute ammonia until free from chlorides. It is then dried, detached from the paper, and the Isitter burned on the platinum wire. The precipitate may also be placed in the crucible fol- lowed by the folded paper. The precipitate is first heated gently, the hd being on the crucible. When ammonia is no longer evolved, the full flame of the Bunsen burner is applied and finally the blast-lamp for a few minutes. 259. Carbon Dioxide by Loss. — If the dolomite is free from organic matter, carbon dioxide may be determined by loss after fusing with borax.--= For this purpose 5 grams of vitrified borax are fused in a platinum crucible with the Bunsen burner. If considerable water is present, it may be more quickly expelled by heating over the blast-lamp. As borax is shghtly vola- tile at this temperature, the material must be brought to constant weight by heatmg to fusion with the Bunsen burner. One gram of the dolomite is placed on the borax, the hd is placed on the crucible, and the contents are * Microcosmic salt may also be used for this determination, as described on page 111. ANALYSIS OF DOLOMITE. 183 brought to fusion with the Bunsen burner. The lid must not be removed until the borax is fused, as it is apt to crack so that small pieces may be thrown out of the crucible. When the dolomite has dissolved in the melted borax and bubbles of gas escape only at considerable intervals, the crucible is cooled in the desiccator and weighed. When the weight is constant, the amount of carbon dioxide is found from the loss in weight of the borax and dolomite. The borax-glass may be removed from the crucible by pressing it between the thumb and fingers on all sides. This generally loosens the melt so that it drops out on gently tapping the crucible. DIRECT WEIGHING OF CARBON DIOXIDE. 260. Apparatus. — Carbon dioxide may also be determined by absorp- tion in caustic potash and ascertaining the increase in weight. For this purpose the apparatus shown in Fig. 18, p. 113, is very convenient and easily made. The glass tube is about 70 cm. long and not less than 12 mm. in diameter. It is drawn out so as to pass through the rubber stopper in the flask. The upper half is filled with granulated calcium chloride, a plug of asbestos or glass wool being first inserted and pushed to the middle of the tube. A few lumps of pumice-stone which have been saturated with copper sulphate solution and dried at about 125° are first dropped in to absorb any hydrochloric acid gas which may pass over. Large lumps of calcium. chloride are then dropped in and then smaller ones, the fine material being at the end where another plug of cotton or glass wool is placed. After being filled in this manner, each end is connected with a calcium chloride tube or other drjdng apparatus and a slow stream of carbon dioxide or hydrochloric acid gas passed through the tube. This is maintained for twenty- four hours, when the acid gas is displaced by a stream of air. The upper end of the tube must be carefully protected from exposure to the atmosphere or moisture. The flask may have a capacity of 100 or 200 c.c. The end of the dropping-funnel should be drawn out and the end of the small tube turned up to prevent the escape of carbon dioxide through the dropping- funnel. A straight calcium chloride tube filled with soda-Ume is fitted to the dropping-funnel by means of a small perforated rubber stopper. 261. For the Geissler caustic potash bulb a solution of caustic potash is made by dissolving one part of the potash in two parts of water. The bulbs are filled not more than two-thirds full with this solution. The cal- cium chloride tube is filled with soda-lime and calcium chloride, the soda- lime being nearest the caustic potash. The calcium chloride should be moderately fine, so that the gases may leave the apparatus well dried. Before placing the soda-lime in the tube a small plug of glass wool or cotton should be inserted. If a caustic potash bulb without the attached calciimi chloride tube is used a separate U-tube or straight calcium chloride tube must be filled as directed and weighed separately. 184 ANALYSIS OF MINERALS. 262. Testing for Air Leaks. — ^When setting up the apparatus, it is well to shellac the small end of the rubber stoppers, as rubber is pervious to carbon dioxide. The small entrance-tube of the Geissler bulb can generally be inserted directly into the rubber stopper in the end of the glass tube, thus saving an extra rubber joint. When the apparatus has been set up, it is tested for leaks by closing the stop-cock in the dropping-funnel and suck- ing out a little air through the caustic potash bulb. Enough air should be sucked out to cause the solution to rise an inch or two in the limb of the Geissler bulb nearest the long glass tube. This column of hquid forms an index of the partial vacuum in the apparatus. The position of the liquid is carefully noted. If the hquid slowly falls, a leak exists which must be found and closed. The most delicate index is formed by the small drops of liquid that generally remain in one end of the small tubes connecting the bulbs of the absorption apparatus. If the apparatus is absolutely gas- tight, a drop of this kind will remain motionless for ten to fifteen minutes. 263. Determination of the Carbon Dioxide. — When the apparatus has been made tight, 1 gram of the dolomite is weighed out and trans- ferred to the flask. After replacing the flask the apparatus is again tested for leaks. Air is now slowly drawn through the apparatus for about half an hour to. remove any carbon dioxide which may be present. The caustic potash bulb is detached, thoroughly cleaned, and after standing in the bal- ance-case for twenty to thirty minutes, is weighed. Whenever 'this bulb is detached, the hole in the stopper in the end of the long glass tube should be closed with a glass plug. After replacing the weighed caustic potash bulb, introduce from 30 to SO c.c. of dilute hydrochloric acid into the dropping-funnel. Open the stop- cock and allow the acid to flow into the flask at such a rate that not more than two bubbles of air per second pass through the caustic potash solution. After all of the acid has been introduced, close the stop-cock in the funnel and warm the flask gently, finally to boiling. A very small flame must bo used at first to prevent the gases from passing too rapidly through the caustic potash solution. It must at all times pass slowly enough to permit the bubbles to be readily counted. When water begins to condense in the lower part of the long glass tube, the heating of the liquid should be stopped, the stop-cock quickly opened, and suction applied to the absorption- bulb. When the first violent flow of air into the flask has ceased, air is drawn through the apparatus at the rate of two bubbles per second for about three-quarters of an hour. For this piu-pose a Bunsen filter-pump may be used. The flow of air may be regulated by a screw pinch-cock. A better device consists of a large bottle filled with water which is led out through a siphon extending to the bottom of the bottle and passing through a doubly perforated stopper. Through the second perforation passes the tube connected to the carbon dioxide apparatus. The flow of water is controlled by a pinch-cock placed on a rubber tube at the end of the siphon. ANALYSIS OF DOLOMITE. 185 The caustic potash bulb is disconnected, carefully cleaned, placed in the balance-case, and weighed. After replacing the caustic potash bulb, the solution of the dolomite is gradually heated to boiling again and air drawn through the apparatus in the same manner as before, after which the caustic-potash bulb is again disconnected and weighed to ascertain if all of the carbon dioxide has been swept out of the apparatus. If the operation has been properly conducted, not more than 2 or 3 mg. will be obtained on the second weighing. When sufficient experience has been gained the operation may be so conducted that only one weighing wiU be necessary. CHAPTER XVI. ANALYSIS OF SILICATES AND SEPARATION OF SODIUM AND POTASSIUM. SEPARATION OF MAGNESIUM FROM THE ALKALIES. In the course of an analysis the other metals are first precipi- tated, leaving magnesium and the alkalies in solution. Magnesium may then be precipitated as phosphate and weighed. During the course of this determination alkalies are introduced into the solu- tion so that it is not available for the determination of these metals. A different method of precipitation of the magnesium must there- fore be adopted or a fresh portion of the unknown must be taken and the magnesium as well as the other heavy metals separated from the alkalies. The latter is by far the most satisfactory method of procedure, and is invariably adopted when sufficient material for analysis is at hand. 264. Separation of Magnesium by Means of Mercuric Oxide. — The magnesium may be removed without introducing alkalies Dy means of mercuric oxide. The solution of the chlorides from which other acids must be absent is treated with mercuric oxide which is free from alkalies. The operation should be con- ducted in a platinum dish. It is best to use freshly precipitated moist mercuric oxide. If much ammonium chloride is present, the solution should first be evaporated to dryness in the platinum dish and the residue gently ignited until ammonium salts are expelled. The residue is then dissolved in a little water and excess of the mercuric oxide added and the material well stirred. The solution is evaporated to dryness with frequent stirring. The residue is heated, at first gently and then more strongly, until mercuric chloride is no longer evolved. All of the mercuric oxide need not be volatilized. The residue is extracted repeatedly with small quantities of hot water and the solution rapidly fil- 186 SEPARATION OP MAGNESIUM FROM THE ALKALIES. 187 tered. The washing must be discontinued as soon as the residue is free from chlorides. By this operation the magnesium is converted into oxide accord- ing to the following equation: MgCl2-l-HgO=MgO+HgCl2. The chlorides of the alkalies are not decomposed in this manner. As the magnesium oxide is slightly soluble, the' solution of the alkalies will contain a little magnesium. Unless the amoimt of magnesium present is less than 1%, the treatment with mercuric oxide must be repeated. The insoluble mixture of magnesium and mercuric oxides may be ignited, the mercury volatilized, and the magnesium weighed as oxide. The magnesium may -also be dissolved in hydrochloric acid after volatilization of the mercury and deter- mined as phosphate. The mercuric oxide used must not leave a residue on volatilizing a portion by heating it to redness in a platinum dish or crucible. Alkalies are frequently present and remain as an alkaline residue after volatilization of the mercury. If mercuric oxide free from alkalies cannot be obtained the amount used must be weighed and a correction made for the amount of alkali present. 265. Separation of Magnesium by Means of Barium Hydroxide. — ■ The magnesium may also be removed by treating the concen- trated solution with a solution of barium hydroxide until it is strongly alkaline. After filtering off and washing the magne- sium hydroxide the excess of barium must be removed from the solution of the alkalies by precipitation with ammonia and am- monium carbonate. • 266. Separation of the Alkalies. — ^The solution of potassium and sodium chlorides is evaporated to dryness in a weighed plati- num dish, finally on the water-bath. The residue must be heated to low redness to completely dry the chlorides before weighing. Two difficulties are met with in this process. The crystals of sodium and potassium chloride tend to enclose water so that on drying the material decrepitates with considerable violence, tending to throw the material out of the dish. The chlorides of sodium and potassium are volatile at a very low temperature. It is therefore necessary after evaporating on the water-bath to cover the platinum dish with a watch-crystal and transfer it to the hot- plate and dry as completely as possible before heating with the Fig. 26. 188 ANALYSIS OF MINERALS. Bunsen burner. The burner must be held in the hand and the flame waved under the dish and removed as soon as any portion of the dish becomes red- hot. If the decrepitation is excessive, the watch-crystal on the dish may be inverted and the platinum dish placed in another of larger size. Any material which is thrown out of the first dish is then deflected downwards into the second one. The weighed chlorides are dissolved in water. A slight amount of insoluble matter which is usually present is filtered off on a small filter-paper and well washed. The paper is thrown into the platinum dish and ignited. The dish with the residue is weighed. This weight is subtracted from the weight of the dish and the dry chlorides. To the solution of the alkali chlorides enough platinic chloride is added to convert both the sodium and potassium chlor- ides into the chlorplatinic salts. The subsequent manipulation is the same as that given for potassium in Chapter VII. The sodium is found by difference. DECOMPOSITION OF SILICATES. The first step in the analysis of a silicate has for its object the breaking of the bond between the silica and the basic elements present, so that the latter may be brought into solution. As the silicates are among the most stable bodies found in nature, the accomplishment of this object is not always. easy. When com- pletely decomposed, or opened up, as the operation is sometimes termed, the analysis of the silicate involves only the ordinary problems of separation and determination. 267. Fusion with Alkali Carbonates. — By far the most largely used and perhaps the best method of accomplishing this object consists in fusing the very finely ground material with sodium carbonate or the so-called fusion mixture, which consists of sodium and potassium carbonates in the proportion of their molecular weights. This mixture fuses at a somewhat lower temperature than sodium carbonate alone. The result of this method of decomposition is that the silica is largely converted into sodium silicate, the silicic acid displacing the volatile carbonic acid in DECOMPOSITION OF SILICATES. 189 the alkali carbonate. The bases remain as • carbonates, or if a given base forms a carbonate unstable at the temperature of the fusion it will remain as oxide. The bases which have marked acid properties, such as aluminium, will combine more or less completely with the alkali. 268. Determination of Silica.— On treating the fused mass with water and acidifying with hydrochloric acid the alkali sili- cate is decomposed with liberation of silicic acid. The remaining alkali salts which are formed are also decomposed. On evapo- rating to dryness, to dehydrate the silicic acid, and again taking up with water and hydrochloric acid, all of the elements present go into solution with the exception of silica and part of the titanium which is frecpently present in silicates. It is seldom that the decomposition of the silicate and the solution of the bases in hydrochloric acid is so complete that the silica is absolutely pure. On the other hand, some of the silica goes into solution even after very prolonged drying. To recover the dissolved silica, the filtrate is evaporated to dryness and the residue again taken up with water and hydro- chloric acid. The most complete separation of silica is effected by heating the dry residue on the water-bath for one-h^vif to one hour. If the residue is dried on the hot-plate or in tt j air- oven at a temperature above 100°, more of the silica tends to go into solution again. The residue after the first evaporation may be dried on the hot-plate. This is a more rapid method, and any silica which goes into solution will be recovered by the evaporation of the filtrate, the residue of which must be dried at 100°. A small amount of silica (1 to 2 mg.) goes into solution even after the second evaporation. This is precipitated with the iron and alu- minium and may be recovered by dissolving these elements in acid-potassium sulphate, dissolving the melt in sulphuric acid, evaporating till fumes appear, diluting with water, and filtering. 269. Determination of Aluminium, Iron, and Titanium. — ^The aluminium and titanium oxides which are usually present with the silica may be recovered by treating the ignite I and weighed pre- cipitate with hydrofluoric and sulphuric acids and evaporating. In this manner the silica may be completely volatilized. Unless 190 ANALYSIS OF MINERALS. sulphuric acid is present, some of the titanium will also be volatil- ized. Miev blasting, the residue is weighed and the correction applied to the weight of the silica. As the bulk of this residue is generally composed of titanium and aluminium oxides, it is advis- able to combine it with the precipitate of aluminium, iron, and titanium hydroxides. After igniting and weighing, the combined precipitate is dissolved by fusing with acid-potassium sulphate. The iron is reduced by passing hydrogen sulphide through the water solution of the fusion. It is then determined volumetrically. The titanium is then oxidized by means of hydrogen peroxide and the amount determined by comparing the intensity of color with the color of a standard titanium solution. (See page 194.) The other metals are separated and determined by methods which have already been described. 270. Decomposition of Silicates for the Determination of the Al- kalies. — As alkalies have been introduced into the fusion mixture, these metals must be determined in a separate portion, which must be decomposed by a different method. One of the best of these methods is that of J. Lawrence Smith. The silicate is decom- posed by gently heating an intimate mixture of the silicate with 1 part of ammonium chloride and 8 parts of pure calcium carbonate. The alkalies as well as some of the calcium are converted into chlorides. Ammonia, the excess of the ammonium chloride, and carbon dioxide are volatilized. On treating with water, the chlo- rides of the alkalies dissolve, while the other bases together with most of the calcium remain undissolved. One advantage of the method is found in the fact that all of the magnesium remains in the insoluble residue, thus separating it from the alkalies. The calcium in the filtrate is precipitated with ammonia and ammo- nium carbonate and filtered off. The filtrate is evaporated to small bulk and a little more calcium precipitated with ammonium oxalate and filtered off. The alkalies in the filtrate, which now contains nothing but these metals as chlorides, are separated and determined by methods already given. 271. Decomposition of Silicates by Fusion with Boric Oxide. — Another method for decomposing silicates by which a solution is obtained in which all of the bases, including the alkalies, may be DECOMPOSITION OF SILICATES. 191 determined, has been devised by Jannasch and Heidenreich.* The flux is powdered boric oxide. It is prepared by recrystalUzing two or three times the commercial article to free it from traces of alkali. The purified crystals are dried and fused in a large plati- num crucible until free from water. By suddenly cooling the fused mass it cracks into pieces of a convenient size for powdering. The powdered material is kept in well-stoppered bottles, as it is hygroscopic. Nearly all silicates are readily decomposed when mixed with 3 to 8 or more parts of this flux and heated, at first gently, and finally with the blast-lamp. The silicate must be very finely powdered, from one-half to one hour being taken for the grinding of 0.5- to 1-gram portions. A platinum crucible holding 45 to 60 c.c. should be used. After intimately mixing the silicate and the flux, gentle heat from a Bunsen burner is applied. Frothing and rising in the crucible is prevented as far as possible by stirring with a short platinum wire which does not reach above the edge of the crucible. When the material has been in quiet fusion for some time in the covered crucible, the flame of the blast-lamp is applied. From twenty to thirty minutes are generally required for the entire operation. Some silicates like andoliisite, cyanite, and topaz are not fully decomposed when heated with the flux by means of the blast- lamp. For these minerals more of the flux must be used, as high as 30 parts to 1 of the mineral being taken. After the preliminary heating with the Bimsen burner, a few additional grams of boric oxide are added and a flame which is fed by oxygen in place of air is applied. An ordinary blast-lamp is used with an opening 2^ mm. wide. It is supplied with gas from at least 5 or 6 ordinary gas- cocks and the flame is made broad and free from luminosity. The heating is continued until the fusion is as transparent as glass. 272. Volatilization of the Boric Acid. — After fusion, the hot covered crucible is immersed in cold water until cold. The con- tents are then placed in a large porcelain or platinum dish and a saturated solution of hydrochloric acid in methyl alcohol is added * Zeit. fiiF anoTR. Chem., Vol. XII, p. 208, 1896. 192 ANALYSIS OF MINERALS. while the dish is covered with a watch-crystal. The crucible is rinsed with more of the methyl chloride and the solution in the dish is gently heated until, with occasional additions of the methyl chloride, solution is complete. This requires from ten to fifteen minutes. The solution is then boiled down to small bulk and evaporated to dryness on a water-bath in which the v/ater is not quite boiling, so that the temperature in the dish is 80° to 85°. More methyl chloride is added and the heating at this temperature is continued until the boric acid is completely expelled. All of the metals will then be in solution as chlorides and may be sepa- rated and determined by methods already given. This method of decomposing silicates has the advantage that a solution of the mineral is obtained which is small in bulk and free from the large amount of alkali salts which are introduced by the carbonate fusion. The separations of many of the metals are much cleaner and sharper in the absence of a large amount of the alkali salts. All of the -metals, including the alkalies, may be determined in one portion. This is a decided advantage when only a small amount of the mineral is at hand. On the other hand, the purification of the boric acid, and more especially the uncertainty as to the complete decomposition of the mineral, are disadvantageous.* EXERCISE 39. Analysis of Feldspar, (K20,Al,03)(SiOj),. Iron, Calcium, Magnesium, Sodium, and Titanium are also Usually Present. 273. Decomposition of the Mineral. — One gram of the finely powdered material is placed in a rather large platinum crucible and mixed with 5 grams of a mixture of 5 parts of sodium and 7 parts of potassium carbonate. The crucible is heated with the Bunsen burner until the mass is in quiet fusion and no more carbon dioxide is evolved. About one-half hour will be required. Seize the crucible with the tongs and cause the contents to solidify in a thin film around the sides of the crucible by giving it a rotary motion while the contents are cooling. Transfer the crucible to a beaker just large enough to allow the crucible to be placed on its side. Add enough water to cover the crucible and heat on the water-bath until the melt is disintegrated. Take out the crucible with a pair of tongs and rinse it well. The lid must also be well washed. * For common errors in the determination of silica, see Jour. Am. Chem. Soc. i4 (1902), p. 262. ANALYSIS OF FELDSPAR. 193 The material in the bcalr is poured in .from the wash-bottle. If an incandescent lamp is u^ed as resistance in circuit with the electrodes, the removal of the acid solution is indicated by the disappearance of light in the lam|_). When the acid solution has been removed in this manner, the deposited metal is washed with distilled water and then with alcohol to remove the water. Ether is sometimes used for the same purpose, but it is much more liable to contain fatty substances in solution, which remain after evaporation of the ether. If ether is used, the metal will dry much more rapidly. The removal of the water is necessary, as moist metals oxidize rapidly and increase in weight. After washing with alcohol or ether, the metal is dried for a few minutes in the steam-bath, or in the air-bath kept at 100°, and after cooling, is weighed. The removal of the alcohol and the drying may be accomplished much more rapidly by burning the portion adhering to the surface. After cooling, the metal is weighed. A trifling amount of oxidation takes place when this method is used. 296. Rotation of the Anode or the Cathode offers several very decided advantages. The metal is generally depositwl in a much smoother and more coherent film, so that a number of metals which cannot be deposited in coherent films with stationary electrodes can be successfully determined by rotating one of the electrodes. This character of the film also allows the use of a much larger current, so that the deposition is much more rapid. The agitation of the solution also increases the rate of deposition because the layer of liquid near the cathode is very soon deprived of its metal by the electric current when the solu- tion is allowed to remain at rest, while when it is kept in constant motion, more metal is constantly being brought up to the electrode. For the same reason, heating the solution increases the rate of deposition because convection currents are set up in the liquid. Rotation of the cathode or anode produces a much more vigor- ous circulation and consequently hastens the deposition, so that 15 minutes is in many cases a sufficient time for the complete deposition of the metal. 297. Electrodes for Rotation. — Cylinders are not suitable electrodes for rotation. Platinum dishes, however, are admir- ably adapted for this purpose. The platinum spiral used as anode 210 ELECTROLYTIC METHODS. can be very conveniently rotated, care being taken to accurately center the spiral and not to rotate it so fast that the solution is thrown out of the dish. Ordinary platinum crucibles may also be used for this purpose. A rubber or cork stopper is selected which fits tightly into the crucible. The shaft of the rotator is passed through a hole in the middle of the stopper, electrical contact being made between the shaft and the crucible. For this purpose a copper wire is wound around the shaft and passes through the hole m the stopper, projecting far enough to Fig. 32 touch the bottom of the crucible. Contact is made between the shaft and a binding post which is connected with the nega- tive terminal. The positive terminal is connected with the anode, which may consist of platinum foil or wire. This arrangement is shown in Fig. 32. The metal is deposited on the outside of the crucible to within about a quarter of an inch of the top. When being weighed the crucible is inverted on the scale pan. 298. The Motors for rotation may be electric or water motors. A velocity of 300 to 600 revolutions per minute will be required. MOTORS FOR ROTATION. 211 Fig. 33. 212 ELECTROLYTIC METF.?DS. A system of pulleys of different sizes is convenient for regulating the speed of rotation. In addition to this means of regulation, the current supplied to electric motors may be varied by chang- ing the resistance in the ch'cuit, and with water motors the amount of water supplied is easily con- trolled. An apparatus designed for utilizing a water motor is shown in Fig. 32. The upright shaft of the motor rotates on a needle point, while the shaft hold- ing the crucible rotates on the ball-bearings of a bicycle hub. An apparatus designed for using an electric motor is shown in Fig. 33. R is the resistance in circuit with the solution being electrolyzed and controls the amount of current used, as indi cated by the ammeter A, which is in the same circuit. The fuses F protect the ammeter from being burned out. The motor M is attached by means of the flexible chord C at the socket T. The motor may l>e instantly stopped by pulling the chord out of the socket T. The motor requires no outside resistance. The switch S controls both the ■ motor circuit and the electrolytic circuit, so that the electrolysis and the rota- tion may be started or interrupted simultaneously. The wiring of this board is shown in Fig. 34, which is a diagrammatic representation of the reverse of the slate board upon which the apparatus is mounted. A volt meter has not been permanently connected with the other apparatus because in ordinary work it is not needed. The resistance R is in series with the electrolytic circuit, but not with the motor circuit. Fig. 34. CHAPTER XVIII. ELECTROLYTIC DETERMINATION OF METAL!^. DETERMINATION OF COPPER. 299. Deposition from a Nitric or Sulphuric Acid Solution. — ^The electrolytic determination of copper is, no doubt, the most accurate as well as the most rapid method of determining this metal. If the solution of copper is free from other metals, a great many of the methods which have been proposed may be used. If, as most gener- ally occurs in practice, the copper must be determined in the pres- ence of other metals but one or two methods are available. By these methods the copper is deposited from a solution containing FREE NITRIC or SULPHURIC ACID. As the copper is apt to be deposited in a spongy form when only sulphuric acid is present, solutions containing only nitric acid, or nitric and sulphuric acids are generally used. Not more than 5% of concentrated nitric acid may be present, and if the same amount of concentrated sulphuric is present, \% of concentrated nitric acid is sufficient. The sul- phuric acid solution is used when the presence of the nitric acid or the ammonium salts formed by its decomposition would interfere with subsequent operations, such as the precipitation of zinc as sulphide. The nitric acid may be entirely omitted if 1 gram of urea is added to the solution containing 7% to 10% of concentrated sulphuric acid. By using a current of about one ampere, the decomposition is then effected in less than two hours.*" 30G. Ssparation and Determination of Lead. — The sulphuric acid solution is used for copper when much lead is present. This metal is then separated as sulphate, in which form it may be fil- tered off and weighed. If only a small amount of lead is present, it may be deposited on the anode from a nitric acid solution, and * Classen, Quan.' Chem. Analysis by Electrolysis, page 160. 213 214 ELECTROLYTIC METHODS. the peroxide washed, dried at 180°, and weighed. If the lead is to be separated and determined in this manner, sulphuric acid should not be introduced, as the lead peroxide will be contaminated with this acid. 301. Separation of Copper from Other Metals. — Zinc, cobalt, NICKEL, CADMIUM, MANGANESE, and moderate amoimts of iron may be present in the nitric acid solution without interfering wit.i the accuracy of the determination of copper. Large amounts oi iron are objectionable because the ferric salts tend to redissolvo the precipitated copper. The difficulty is partially overcome by diluting the solution and increasing the current density. The copper may also be completely separated from iron by precipita- tion with hydrogen sulphide or sodium thiosulphate. The precipi- tate is filtered off, washed, and redissolved in nitric acid. From this solution it may be precipitated electrolytically and weighed. Silver, mercury, bismuth, arsenic, antimony, and tin are deposited with the copper. Silver should be precipitated as chlo- ride before electrolyzing the solution, an excess of hydrochloric acid being avoided. Mercury may be expelled from the ore by roast- ing. "WTien alloys or ores of copper are dissolved in nitric acid the tin and almost all of the antimony remains undissolved while most of the arsenic goes into solution. The removal of these metals may be accomplished by passing hydrogen sulphide through the solution until no more precipitate is formed, and then dissolving the sulphides of arsenic, antimony, and tin in sodium sulphide solution. The copper sulphide may then be dissolved in nitric acid and the copper deposited and weighed. Copper may also be readily and completely separated from bismuth, arsenic, antimony, and iron by precipitation, as cuprous sulphocyanate. The solution of the copper salt is mada slightly acid with hydrochloric acid, and the copper reduced by saturating the solution with sulphur dioxide. A solution of ammonium thiocyanate is then added until precipitation is com- plete, and the mixture heated nearly to the boUing-point for a few minutes. The copper may also be precipitated by the addition of a solution of equal parts of acid ammonium sulphite and ammonium thiocyanate. The precipitate, which is colored when first formed, quickly becomes white, and settles very readily. DETERMINATION OF COPPER. 215 It is washed with warm water until free from the other metals present. The copper is dissolved in a little warm concentrated nitric acid, and then deposited as usual. Copper ores containing arsenic, antimony, and tin may be decomposed by fusion with six times the weight of a mixture of equal parts of sulphur and sodium carbonate or the same amount of dry sodium thiosulphate. The ore must be very finely pow- dered and the fusion conducted in a porcelain crucible. After heating with a small Bunsen-burner flame until sulphur is no longer evolved, the melt is allowed to cool, and then extracted with hot water, and the residue washed with water containing a little ammonium sulphide. By this treatment the arsenic, anti- mony, and tin are dissolved while the copper remains as sulphide, which is dissolved in nitric acid, and the copper in the solution determined. 302. Deposition of Copper from an Ammonium-oxalate Solu- tion. — ^A very excellent method of determining copper is based on its separation from a solution containing excess of ammonium oxalate. Classen* has improved thismethod so that the determina- tion may be carried out in two hours. As most of the metals are precipitated from an oxalate solution, this method cannot be used imless the copper is first separated from the other metals, or the amount present in a pure salt is to be determined. EXERCISE 40. Determination of Copper in Copper Sulphate, CuSO^-sHjO. Clean, dry at 100°, and weigh a platinum dish or platinum cone or cylin- der. Weigh out 1 grain of pure recrystallized copper sulphate. If the cone or cylinder is used,. transfer the copper sulphate to a 250-c.c. bealier, otherwise place it in the weighed dish. Dissolve in a few cubic centimeters of water, add a solution of 4 grams of ammonium oxalate, and dilute the solution to about 120 c.c. Make a saturated solution of oxalic acid by dissolving 4 grams of the crystals in 50 c.c. of water, and filter if necessary. Place the dish on the ring of the electrolytic stand, and adjust the anode so that it is about -J inch below the surface of the liquid. If a beaker is used, it must be placed on a tripod or other support, so that the solution can be heated. The cylinder or cone should be adjusted so that it very nearly touches the bottom of the beaker. Fasten a burette by means of r * Classen, Quart. Anal, by Electrolysis, page 154_ 216 ELECTROLYTIC METHODS. clamp and ring-stand, so that the solution of oxalic acid which is placed in the burette will drop on the split watch-crystal which covers the dish or beaker. A thermometer is also suspended with the bulb in the copper solution, which is heated ^ith a small flame from a Bunsen burner to 80°. The current of electricity used should be so adjusted that about 1 ampere passes through the solution, and the tension at the electrodes is from 2.5 to 3.2 volts. As soon as a film of copper is seen to cover the cathode, the stop-cock of the burette is so adjusted that the oxalic-acid solution falls at the rate Fig. 35. of 10 drops per minute. The copper comes down as a bright-red deposit. If it becomes dark-colored or spongy, the oxalic acid should be allowed to drop faster. The electrolysis will be completed within two hours. When the solution has become colorless, portions may be taken out and tested for copper by acidifying wth hydrochloric acid and adding potassium ferrocyanide. When all of the copper has been deposited, siphon off the acid liquid while water from the inverted wash-bottle is allowed to flow over the copper deposit. The copper is then washed with alcohol, dried at 100°, and weighed. The theoretical percentage of copper in copper sulphate is 25.41. DETERMINATION OF SILVER. 217 EXERCISE 41. Determination of Copper from a Nitric-acid Solution. The copper in copper sulphate may also be precipitated from a nitric- acid solution. This method is well adapted to the determination of copper in the refined metal . About 0.3 gram of the metal is carefully weighed out, and dissolved in 3 c.c. concentrated nitric acid and a little water. The solution is diluted to about 100 c.c, and warmed to 50° or 60°, and the. copper precipitated with a current of 0.5 to 1 ampere. No arrangement need be made to keep the solution warm, as in the preceding method. The deposition of the metal is more rapid, and a better deposit is obtained if the solution is warm at first. After four to five hours, all of the copper will be deposited. If a cylinder made of wire gauze is used, about half this time will be required. If the rotating cathode or anode is used, the amount of nitric acid may be reduced to 1 c.c, while the current may be increased to from 3 to 5 amperes. Most of the copper will be deposited in 10 min- utes. The small amount still remaining in solution will be very rapidly deposited if the solution is nearly neutralized by the addition of ammonia. When all of the copper is deposited, as shown by the ferrocyanide test, the solution is displaced with distilled water while the current is passing, and the deposit washed with alcohol, dried, and weighed exactly as directed in the first method. 303. Deposition of Silver from a Potassium-cyanide Solution. — - Silver is determined electrolytically almost exclusively from a cyanide solution. The potassium cyanide used must be pure. Failure to obtain good results is frequently due to the use of impure cyanide. Another sample of cyanide should therefore be obtained and used. The current density should be less than 0.5 ampere. The electrolysis may be conducted at the ordinary temperature, though when heated to about 65°, the time re- quired for a determination is very much reduced, so that two or three hours is generally sufficient. 3 grams of potassium cyanide are sufficient for 0.5 gram of silver. If other metals are present, the silver may be precipitated as chloride, which is then dissolved in the potassium cyanide. It is doubtful, however, if in this case the result would be more accurate or obtained more quickly than if the sUver chloride were directly weighed. 304. Separation of Silver from Lead and Copper. — Silver may be separated from lead in consequence of the deposition of the lat- 218 ELECTROLYTIC METHODS. ter on the anode as peroxide, when a solution containing 20% of concentrated nitric acid is electrolyzed. The separation of silver from copper is effected by electrolyzirg a solution of the two metals in potassium cyanide by a current giving a tension between the electrodes of about 1.4 volts. When all of the silver has been deposited, the copper may be determined by adding nitric acid to the solution, and increasing the tension between the electrodes. EXERCISE 42. Determination of Silver in Silver Tfitrate. Weigh out 0.5 gram of pure dry silver nitrate, and dissolve in a fevi^ cubic centimeters of water. Dissolve about 2 grams of potassium cyanide in a small amount of water, and pour the silver solution into the cyanide solution. Either the platinum dish or the cylinders may be used as described under the determination of copper. Three Edison-Lalande cells connected in series or other source of electricity giving about 3 volts, should be used The electrolysis requires about six hours, but the time necessary may be reduced by warming the solution to 65°, which is about the temperature the hand can bear for a moment without much discomfort. If the anode is rotated, a current of 2 to 3 amperes may be used, and the deposition will be complete in 10 to 15 minutes. The end of the operation is ascertained by testing a few drops of the solution by adding hydrochloric acid. The test should be made under the hood. Siphon off the solution while add- ing distilled water, rinse with alcohol, dry at 100°, and weigh. If the cyanide solution becomes quite dark on passing the current, the potassium cyanide is impure and the result is not apt to be correct. The determina- tion should be repeated with another sample of the cyanide. The theo- retical percentage of silver in silver nitrate is 63.51. EXERCISE 43. Analysis of a Silver Coin^ 305. Solution of Coin. — When results within 0.1 or 0.2% of the theo- retical have been obtained with silver nitrate, a coin is analyzed. Thor- oughly clean a dime or other silver coin by rubbing with alcohol, ammonia, etc. Cut it into four or five pieces so that each piece weighs about 0.4 gram. "Weigh one of tlie pieces carefully, and dissolve in a few cubic centimeters of a mixture of equal parts of water and concentrated nitric acid. This should be done in a small covered beaker. Warm the beaker on the water-ljath, and when solution is complete remove the watch- crystal, rinsing it if necessary, and let the solution go to dryness. DETERMINATION OF SILVER. 219 306. Silver. — In the mean time weigh out roughly and dissolve in water two grams of potassium cyanide. Dissolve the copper and silver nitrates in a little water, and pour the solution into the cyanide solution. If difficulty is experienced in dissolving the salts of copper and silver which have dried on the sides of the beaker, a little potassium cyanide solution may be added after the water solution has been removed. The silver may be deposited on a platinum dish or on a cylinder. The solution should be warmed to about 65°, and electrolyzed with the current from two Edison- Lalande or similar cells connected in series. The difference of potential between the two platinum electrodes must be measured with a volt-meter and must not exceed 1.5 volts. If the two cells give a higher voltage than this, the potential must be reduced by introducing resistance. A resistance- box may be used, or the connection between the cells and one of the electrodes may be made by means of a few feet of German-silver wire. By drawing more or less of this wire through one of the binding-posts the resistance may be changed until the potential between the two electrodes is a little less than 1.5 volts. The dish or beaker is covered with a split watch-crystal. As the evapo- ration at 65° is considerable, water must be added from time to time. In about five hours the electrolysis will be completed, as may be ascertained by adding water until an unplated portion of the platinum dish is covered. If no silver is deposited on this portion of the dish in half an hour, the deposition of the silver is complete. The solution is now siphoned off into a beaker and replaced with water, care being taken to lose none. The silver is rinsed with alcohol, dried at 100°, and weighed. 307. Copper. — The copper in the filtrate is deposited on a cone or cylin- der after the addition of 5 c.c. concentrated nitric acid. The operation is repeated until analytical duplicates are obtained. The rotating anode may be used in this determination as well as in the determination of the silver. DETERMINATION OF NICKEL AND COBALT. Both nickel and cobalt are most easily determined electro- lytically. The same methods serve for both metals. Where the two metals occur together, they are separated from the other metals present by the methods given in Chapter XIV. Both metals are then precipitated electrolytically, and, after weighing, the deposit is dissolved in nitric acid, one of the metals being precipitated and weighed by methods given in the ' chapter mentioned. 308. Ammonium-oxalate Method. — According to Classen * these metals may be readily deposited from a solution of the ♦ Classen, Quan. Anal, by Electrolysis, page 141. 220 ELECTROLYTIC METHODS. double-ammonium oxalate. Nitrates and chlorides should be removed by evaporation with sulphuric acid. The excess of acid is neutralized with ammonia, and 4 to 5 grams of ammonium oxalate added, which is dissolved by warming the solution. Water is then added until the volume is from 100 to 120 c.c. If the temperature of the solution is kept at 60° to 70°, the elec- trolysis will be complete in three to four hours, a current of about 1 ampere and an electrode tension of 3 to 4 volts being used. 309. Ammonia Method. — By the method of Fresenius and Bergmann * chlorides and nitrates are removed by evaporation with sulphuric acid. The excess of acid is then neutralized with ammonia, and unless at least 6 c.c. of strong ammonia (sp. gr. 0.90) are used for this purpose, 6 grams of ammonium sulphate should be added. 15 c.c. of strong ammonia are then added, or if more than 0.5 gram of metal is present, 20 c.c. should be used. The solution is then diluted to 150-170 c.c, and electrolyzed with a current of 0.7 ampere. The separation is not hastened by warm- ing the solution. The deposition is complete in about six hours. EXERCISE 44. Analysis of a Nickel Coin. 310. Solution of Coin. — Clean a nickel coin thoroughly by rubbing with sand or " Sapolio." Cut it into small pieces weighing about 0.4 gram. After weighing one of these pieces place it in a small beaker (about 150 c.c), and add 10 c.c. dilute nitric acid. Warm the covered beaker on the water- bath until the metal is dissolved. Rinse off the watch-crystal, add 16 c.c. dilute sulphuric acid, and evaporate on the water-bath until no more nitrous fumes come off. 311. Copper. — If a cylinder is available, dilute the solution until the cylinder is nearly covered when immersed so as to very nearly touch the bottom of the beaker. Electrolyze with a current of about 0.8 ampere. If the solution is warmed to about 60°, the copper will be deposited in from four to six hours. If it is convenient to conduct the electrolysis overnight, a current of about 0.4 ampere should be used, and the electrolysis begun with a warm solution, which is then allowed to cool off. The end of the operation is ascertained by testing a small portion of the solution for copper by passing hydrogen sulphide. After boiling for a few minutes to expel *Zeit. Anal. Chem., vol. 19, page 329. DETERMINATION OF TIN. 22i the hydrogen sulphide, this portion may be returned to the beaker. The solution may also be tested for copper by adding water so that a clean por- tion of the cylinder is immersed. If no copper is deposited on this portion after half an hour, all of the metal has been deposited. When all of the copper has been deposited, the solution containing the nickel is siphoned off and replaced with water. The copper is washed with water and alcohol, di-ied at 100°, and weighed. The anode or cathode may be rotated, and the current largely increased, and the time required for the deposition reduced to less than half an hour. 312. Nickel. — The solution containing the nickel should be evaporated if the volume is more than 250 c.c. It is neutralized with ammonia, 15 c.c. of concentrated ammonia added, and the nickel deposited with a current of about 0.7 ampere. About five to six hours will be required. The solution is tested for nickel by adding to a small portion a little colorless ammo- nium sulphide, or passing hydrogen sulphide. If the nickel is deposited slowly, enough ammonia is not present and more must be added. The solu- tion need not be siphoned off. The nickel is washed with water and alcohol, dried at 100°, and weighed. The determination is repeated with other por- tions of the coin until duplicates are obtained'. The rotating anode or cathode may be used for the deposition of the nickel. The current may then be increased to 5 amperes, when the deposition will be complete in less than half an hour. DETERMINATION OF TIN. Tin may be deposited in a bright, crystalline, closely adherent form from solutions of the double-ammonium oxalate containing free oxalic acid. If the tin is obtained as the dioxide, it may be dissolved by fusing with 8 times its weight of caustic soda in a silver dish. The melt is dissolved in water, an excess of oxalic acid added and the solution heated until clear. Hydrogen sulphide is then passed to precipitate the silver which is filtered off, after which the solution is boiled to expel the hydrogen sulphide. Ammonium oxalate is then added, the solution warmed to about 65°, and electrolyzed with a current of 4 to 5 volts and from 0.5 to 1.5 amperes. For 0.3 gram of tin 9 to 10 grams oxalic acid and 4 grams of ammonium oxalate should be used. The tin is com- pletely deposited in about 4|- hours. It is washed as usual with water and alcohol, and dried in the steam-bath. Excellent results are obtained with the rotating anode. 313. Dissolving Tin Deposits on Platinum. — Some difficulty 15 experienced in dissolving the tin which has been deposiled on 222 ELECTROLYTIC MEJ'HODS. the platinum. This metal tends to form an alloy with the platinum so that acids do not readily attack the tin. The difficulty may be overcome by first coating the dish or other electrode with silver or copper or a silver dish may be usexl for these determinations. The silver is deposited from a solution of silver nitrate in potassium cyanide, while the copper may be deposited from a solution of copper sulphate acidified with nitric acid. After a sufficiently thick coating has been obtained by either method the solution is de- canted and the deposited metal washed and dried as usual for the determination of these metals. The best solvent for the tin is prepared by diluting concentrated nitric acid with an equal volume of water and then saturating with oxalic acid. After most of the tin has been dissolved by this solution the nitric acid is washed off with water and the cylinder or dish immersed in concentrated C. P. hydrochloric and allowed to remain for 5 to 6 hours. It is advisable to keep cylinders used for the determination of tin immersed in concentrated C. P. hydrochloric acid when not in use. This acid, will dissolve all the tin if allowed to act for a sufficient time. DETERMINATION OF LEAD. This element cannot be determined by deposition in the metal- lic form, because during the process of washing and drying it invariably oxidizes to a greater or less extent. Advantage is therefore taken of the fact that when a current of electricity is passed through a nitric acid solution of lead, it is converted into the peroxide which is deposited on the anode, and may be dried by heating to 180°-200°. The platinum anode should be roughened according to the suggestion of A. Classen by means of a sand-blast, otherwise the deposit readily flakes off unless depos- ited very slowly and in small amount. By this method the lead is separated from zinc, cobalt, nickel, iron, aluminium, copper, gold, mercury, antimony, and cadmium. Traces of silver and bismuth, if present, are deposited as peroxide with the lead. 314. Conditions of Deposition of Lead. — The amount of nitric acid which must be present depends on the temperature of the solu- tion and the amount of current used. If the electrolysis is carried out at the ordinary temperature, about 10% by volume of nitric DETERMINATION OF LEAD. 223 acid of sp. gr. 1.38 must be present, and a current of .05 amperes used, while if a current of 0.5 amperes is employed 20% by vol- ume of the nitric acid must be present. The depositon of the lead is much more rapid at a temperature of 60° to 65°, but the amount of acid necessary at this temperature for a current of .05 amperes is 2% by volume; while for a current of 0.5 amperes 7%, and for 1.5 to 1.7 amperes 20% is necessary. The small current must be used when only smooth dishes or electrodes are available. If too little nitric acid is present, metallic lead will be deposited on the cathode. Hydrochloric acid should be absent from the solution, and only small amounts of sulphuric acid should be present, as it is partly precipitated with the lead peroxide. The rotating anode or cathode may be used to advantage. The end of the operation is ascertained by adding water, so as to expose a fresh portion of the electrode. If lead is still present, a dark-brown deposit will be formed after a quarter to half an hour. The solution need not be siphoned off. The lead peroxide is washed with hot water and dried for half an hour in an air-bath heated to 200°. The lead dioxide may be dis- solved in hot dilute nitric acid to which a little oxalic acid has been added, or in dilute hydrochloric acid which has been sat- urated with sulphur dioxide. It may also be dissolved in dilute nitric acid after being converted into htharge by heating for a few minutes in the oxidizing flame of the blast-lamp or of a Bunsen burner. 315. Determination of Lead in Ores. — The electrolytic method is well suited to the determination of lead in ores. If possible the ore should be decomposed without the use of nitric acid, which converts any sulphur present into sulphuric acid. As has already been said, the presence of this acid in the solution intro- duces a slight error in the determination. A method of treating the ore, which is suitable for galena, is given by Medicus.* The finely powdered ore is dissolved in concentrated hydrochloric acid. Excess of caustic-potash solution (1:3), and if antimony is present, 1 to 2 grams of tartaric acid are added. The solution is warmed for a few minutes at 100°, and, after cooling and diluting with water, the lead is precipitated by passing carbon dioxide. *Ber. d. Chem. Ges., 1892, p. 2490, ^24 ELECTROLYTIC METHODS. This will require about one and one-half hours. The lead car- bonate is washed with hot water until free from chlorine, and is then dissolved in dilute nitric acid and the paper well washed. The lead in the solution is then deposited as the peroxide, as already directed. DETERMINATION OF ZINC. A large number of electrolytic methods have been proposed for zinc, very few of which have been found practicable. The difficulties attending the gravimetric determination of this metal have led to continued effort in this direction, until methods which appear to be successful have been devised which depend on the deposition of the zinc from a solution kept slightly acid with a weak organic acid. 316. Difficulty of Dissolving Zinc. — In the electrolytic deter- minations of zinc, the same inconvenience is met with as in the determination of tin, that the metal is dissolved with difficulty from the platinum electrode. The same remedy is adopted: namely, to coat the platinum with silver or copper, or to use silver dishes. The zinc may be dissolved, leaving the copper or silver, by warming with dilute sulphuric acid. 317. Ammonium-oxalate Method. — In the method proposed by Classen,* the zinc is deposited from a solution of the double oxa- late of zinc and potassium or ammonium. To the neutral and concentrated solution of zinc about 4 grams of potassium or ammo- nium oxalate are added. Solution is effected by warming, and, if necessary, adding a little water. The solution is diluted to about 120 c.c, warmed to 50° or 60°, and electrolyzed with a cur- rent of 0.5 to 1 ampere with an electrode tension of 3.5 to 4.8 volts. After the current has passed for three to five minutes, a cold saturated solution of oxalic acid, or better, a 6% solution of tar- taric acid is allowed to flow from a burette at the rate of about 10 drops per minute upon the watch-crystal which covers the dish or beaker. The end of the operation is ascertained by test- ing small portions of the solution with potassium ferrocyanide. * Classen, Quan, Analysis by Electrolysis, p. 146. DETERMINATION OF ZINC. 225 The time required is about two hours. The deposit must be washed without interrupting tlic current. 318. Acetic-acid Methods.— A method proposed by Riche and modified by Smith * involves the use of a solution of zinc acidified with acetic acid. To the neutral solution of the zinc, from which ammonium salts should be absent, 1 gram of sodium acetate, | to 1 gram of ammonium oxalate, and 0.3 c.c. of 99% acetic acid are added. The solution is heated to 65°, and a current of about 0.36 ampere and 4 volts passed, and the anode or cathode rotated. Add acetic acid occasionally, as zinc forms the best deposits from a nearly neutral solution. At the end of an hour the current is mcreased to about 0.7 ampere and 5 volts. ^Vater is added so as to cover a clean silver surface with the solution. If no more zinc is deposited and the solution is acid, it is carefully neutralized by the addition of ammonia and the current passed for about fifty minutes longer. The condition of the solution which requires the addition of ammonia is usually indicated by the solution appearing as if full of bubbles. The complete precipitation of tlie zinc is ascertained by testing a portion of the solution ■^^•ith potassium ferrocyanide. The solution is then siphoned off and the depos- ited metal washed with water and alcohol, dried at 100°, and weighed. 319. Determination of Zinc in Ores. — In carrying out the determination of zinc in zinc ores, the finely powdered material is weighed out and dissolved in hydrochloric acid, or, if necessary, in nitric acid. In the latter case the nitric acid is removed by evapo- ration with a little sulphuric acid and the residue dissolved in water. If considerable quantities of copper, arsenic, antimony, etc., are present, hydrogen sulphide is passed through the solution and the precipitate washed with water containing hydrogen sul- phide. If iron, cobalt, and nickel are absent, the zinc may be determined in the filtrate by the method of Classen. Iron does not interfere if the method of Smith is used, since small quantities are not carried down with the zinc, and large amounts would be precipitated as basic acetate on warming the solution. * Jour. Am. Chem. Soc, Vol. XXIV, p. 1073. VOLUMETRIC METHODS. CHAPTER XIX. CALIBRATION OF APPARATUS. 320. Definition. — Volumetric are distinguished from gravi- metric methods by the fact that in the former the amount of a given constituent is ascertained by measuring the volume of a LIQUID, while in gravimetric methods, the element to be determined is brought to a condition where it may be placed on the balance- pan, and WEIGHED. Solutions of reagents which have been made to contain an accurately determined amount of a given chemical substance are called standard solutions. The volume of such a solution, which will just complete the reaction with a given constituent in a known amount of a substance to be analyzed, will give the amount of the constituent to be determined. We may, for instance, weigh out 58.50 grams of pure sodium chloride and dissolve in a. liter of water. The amount of silver contained in a given solution may be ascertained by adding the salt solution until all of the silver has been precipitated as silver chloride, and further addition of the salt solution gives no further precipitate. The volume of the salt solution having been carefully noted, the amount of silver present may be computed from the known amount of salt solution Used. From the equation, NaCl + AgNO^ = AgCl+NaNOs, (23.05 + 35.45-58.50) (Ag- 107.93) we know that 58.50 parts of sodium chloride can precipitate 107.93 parts of silver. As the salt solution contained 58.50 grams of sodium chloride per liter, if 100 c.c. were used to pre- cipitate the silver, then 10.793 grams of silver were present in the unknown solution. 226 VOLUMETRIC APPARATUS. 227 VOLUMETRIC APPARATUS. 321. Flasks. — ^The apparatus which has been devised for meas- uring the solutions consists in the first place of so-called standard flasks. These flasks are made of such a capacity that when filled to a mark made on the stem, they contain definite volumes of liquid, such as 1 liter or 1000 c.c, | liter or 500 c.c, I liter or 250 c.c, ect, 322. Pipettes are tubes constructed to deliver definite volumes having, generally, a bulb at the centre and a mark on the upper part of the stem to indicate the point to which the pipette must be filled to deliver the volume marked on the instrument. The liquid \/ Fig. 36. is sucked up into the pipette, the upper end closed with the finger, and after allowing the liquid to flow down to the mark, the contents of the pipette are allowed to flow into the desired vessel. 228 VOLUMETRIC METHODS. 323. Burettes arc glass tubes of uniform bore, having a stop-cock at the lower end, and marked so that any desired number of cubic centimeters may be measured out. The total capacity is usually Fig. 37. 50 c.c, and the smallest divisions indicate either the fifth or the tenth of a cubic centimeter. It is customary to estimate the tenth of the smallest division so that volumes may be measured with burettes to the /„ or the y^ir of a cubic centimeter. 324. Reading Burettes. — Unless special care is taken in making readings, errors of ^^ of a cubic centimeter may easily be made. Generally the lowest part of the meniscus is the point at which readings are made. As this point is in the centre of the tube, a different reading will be made according to the position of the eye, as will be seen from Fig. 38. One of the simplest devices for securing the correct position of the eye, corresponding to the point h, consists of a piece of paper or card with a perfectly straight edge folded double so as to completely surround the tube. The card is held so that the folded upper edges meet exactly, as shown READING BURETTES. 229 in Fig. 39. The eye is placed in such a position that the edge of the card on the front of the burette, the bottom of the meniscus, and the edge on the back of the burette are in the same line. The reading of the burette is then taken at the upper edge of the card. The eye may also be held in the same straight line with the upper edges of the meniscus. The lowest point of the meniscus Fig. 38. Fig. 39. is then read. This method will give a reading which is slightly too high, but if the burette is always read in this manner, no error will result, as the volumes measured out are always found by taking the difference between two readings. 325. Floats. — Another method of securing accurate readings is by the use of floats, which are small glass tubes sealed at both ends and containing enough mercury to cause them to be very nearly immersed in dilute water solutions. A circle is etched on the float at right angles to its length, and the readings are made by the line which is produced by holdmg the eye in the plane of this circle. While this line gives very exact readings, serious error is often occasioned by the tendency of floats to go down irregularly by a series of jumps instead of following the level of the liquid. Before being used a float should be carefully tested by placing it in the burette filled with water. The water is allowed to flow out at a moderate rate, while the float is closely observed to see if it descends at a perfectly regular rate. 326. Errors in Reading Burettes. — When skill has been ac- quired in using one or the other of these devices and in estimating 230 VOLUMETRIC METHODS. the tenth of one of the smallest divisions on the burette, the accu- racy acquired in practice is still never greater than xtii of ^ cubic centimeter; that is, the reading will be doubtful to the tenth of the smallest division of the burette. As an error of this magnitude is possible at both of the readings, in measuring a given volume of liquid the total error may be at least two-hundredths of a cubic centimeter. Careful experiments show that in careful work Fig. 40. errors of three-hundredths of a cubic centimeter occur. The percentage error of a measurement is greater, therefore, the smaller the volume of liquid measured out, being 3% for 1 c.c, 0.3% for 10 c.c, and 0.1% for 30 c.c. As the possible errors in the other parts of volumetric analysis are not less than 0.1%, about 25 c.c. is the minimum volume of liquid which should be used in a volumetric analysis. Greater accuracy is seldom obtained by using a larger voliome of liquid. CORRECT USE OF PIPETTES. 'ZVi 327. Method of Using Pipettes. — If a pipette is properly used, liquids may be measured with greater accuracy than is possible with a burette. This arises from the fact that the tube in which the initial and final readings are made is so much smaller than in the case of a burette. The errors in using a pipette arise from allowing the liquid to drain for varying intervals of time, and also from leaving a drop of liquid on the lower end of the tube. This drop should be taken off, when the liquid has been brought to the mark on the upper part of the stem, by drawing the point of the pipette along the glass surface of the vessel from which the liquid has been drawn. The liquid should be allowed to flow out of the pipette at its maximum speed, and the point should be touched against a glass surface when the pipette is empty. If it has a mark on the lower stem, the liquid must be allowed to flow to this mark; otherwise the pipette is completely emptied with the excep- tion of the drop which always remains in the point. Some chem- ists blow this drop out of the pipette. This is unnecessary, but the amounts of liquid delivered will be identical if this practice is uniformly adhered to. More uniform results will be obtained if the hole at the point of the pipette is not over 2 mm. in diameter^ so that the contents of the pipette may flow out in about ten seconds. If the opening is too large, it may be diminished by heating the point to redness in the flame of a Bimsen burner. The upper end of the stem as weU as the finger used in closing it must be dry or the flow of liquid cannot be properly controlled. 328. Standard Temperature. — The volumes of the glass- measuring apparatus vary appreciably with changes in temper- ature, while the changes in the volumes of liquids is very marked indeed. It is necessary, therefore, in volumetric work to keep solutions and apparatus at some standard temperature. The most convenient temperature is that of the air in the laboratory. This changes within quite wide limits and varies in different coun- tries. The German chemists have chosen 15° Centigrade as the average temperature of the laboratory. American laboratories are kept much warmer than this, so that 20° is a much better standard for this country. This will be used as the standard temperature for the work described in this book, the glassware being calibrated for this temperature. 232 VOLUMETRIC METHODS. 329. Expansion of Glass, "Water, etc. — The change in volume of a liter flask due to a change in temperature of 5° C. is 0.15 c.c. As this is .015% of the total volume, it is much less than other experimental errors and may, therefore, be entirely disregarded. The change in volume of water is much greater. A liter of water measured at 20° will contract 0.9 c.c. if cooled to 15°, while if heated to 25° it will expand 1.144 c.c. The error in this case for a change in temperature of 5° will therefore be about 0.1%. The change in volume of solutions is generally much greater, the more concentrated the solution the greater being the coefficient of expansion. 330. Errors of Graduation. — A very common source of error in volumetric analysis is the inaccuracy of the graduations of the measuring apparatus used. It is by no means uncommon to find errors amoimting to 1% in apparatus bought of the dealers, while J% would, perhaps, fairly represent the error in using most un- tested apparatus. The errors are due to several causes, but chiefly to the fact that various standards of volume are in use. 331. Definition of Units of Volume. — Volumetric apparatus is usually constructed on the basis of the metric system, the units of volume being the liter and the cubic centimeter. In this system the unit of length was first selected and prepared. The unit of volume was obtained from the unit of length, a liter being defined as the cube whose sides are a decimeter in length. A cubic centimeter is the one-thousandth part of this volume. The unit of weight was obtained by preparing pure water and calling the weight of one liter weighed in a vacuum, at 4° C. a kilogram. 332. Various Standard Liters in Use. — On account of the ease of preparing pure water, measuring its temperature correctly and weighing it with accuracy, the custom has arisen of using accurately weighed water as the standard of volume. This cus- tom is still further justified by the fact that it is comparatively easy to obtain weights Avhich are correct far beyond the accuracy attainable in measuring volumes. The errors found in commer- cial volumetric apparatus have arisen from a desire to avoid the trouble of making the weighings of the water under the standard conditions of temperature and absence of an . atmosphere. In many instances the suggestion made by Mohr has been adopted. STAXDARDS OF VOLUME. 233 He proposed to adopt as a liter the volume of 1 kilogram of water at 4° C, weighed in air. He adopted 17.5° C. as the standard temperature. This liter would differ from the true liter by 1.2 cubic centimeters. Other workers have used as a liter the vol- irnie of one kilogram of water weighed in the air at 15° C, while still others weigh at 17.5° C. or 20° C. With these various standards in use it is not surprising, espe- cially when taken in connection with the possible errors of work- manship, that the use of uncalibrated apparatus frequently pro- duces very erroneous results. The possibility of error is consider- ably reduced if all of the volumetric apparatus in a laboratory has been made by the same manufacturer, since no error will result from the use of a standard other than the true liter if all of the apparatus is graduated according to the same standard, the cubic centimeter being exactly the one-thousandth part of the liter. Apparatus used for gas analysis must, however, be grad- uated according to the true liter. All confusion would be avoided if the true liter were generally adopted. The German chemists have taken the lead in this direction. The German Imperial Standards Commission has made it illegal to use any other than the true liter for official purposes. The directions given for cali- bration in this book will be based on the true liter. CALIBRATION. There are in use two general methods of calibrating volumetric apparatus. By the first method the weight and temperature of the water which just fills the apparatus to be calibrated is found and the volume calculated from the known density of water. By the second method standard bulbs whose capacity has been care- fully determined are used. The water which just fills the bulb is allowed to flow into the flask to be calibrated. Where many pieces of apparatus must be calibrated this method is to be preferred. 333. Calibration of Flasks by Weighing Water. — Flasks which are to be calibrated are first cleaned and dried, then carefully counterpoised or weighed. Distilled water at 20° 0. is poured in until the flask is filled to the mark. The additional weights neces- sary to again counterbalance the flask give the weight of the 234 VOLUMETRIC METHODS. water. The weight of the water in vacuo must now be calculated. The air displaced by a kilogram of water may be taken as a liter, which under ordinary temperatures and pressures weighs 1.2 grams. When in the air the water is lighter by this amount. The weights, however, are also lighter in the air than in a vacuum. The specific gravity of brass weights is 8.4. A kilogram of brass weights would displace \l of a liter of air which would weigh .14 gram. As the buoyant force on the water is 1.2 grams and on the weights .14 gram, the diminution in weight in the air is 1.06 grams (1.20 — .14). A kilogram of water, therefore, weighed in the air with brass weights would weigh 998.94 grams (1000 — 1.06), the weight in a vacuum being exactly 1000 grams. The general correction for air displacement is therefore +.106%. The weight of the water in a vacuum divided by the density at 20° gives the volume. In this manner the volume of the flask when filled to the mark is obtained. If a flask is to be cahbrated to deliver a definite volume it should be rinsed with distilled water and allowed to drain for the same length of time employed in ordinary use. It is then weighed and the calibration carried out as already described. If it is desired to ascertain the point on the stem to which the flask must be filled in order to contain one liter, it should be dried and counterbalanced on the scale-pan, and then weights equal to 997.2 * grams placed on the opposite pan and distilled water at 20° C. poured into the flask until the weights are' again counter- balanced. A mark may now be made with a file on the neck of the flask opposite the lowest point of the meiiiscus, and then a circular mark may be etched on by means of hydrofluoric acid. For this purpose a thin layer of paraffine or beeswax should be melted on the neck of the flask and a circle traced with a sharp steel point while the flask is rotated in a lathe or some similar device. The acid may be applied by means of a piece of filter- paper and should be allowed to remain from three to five minutes. 334. Calibration of Burettes by Weighing Water. — The burette is filled with distilled water at about 20° C. A small flask or * 1000 CO. water at 20° in a vacuum weigh 998.26 grams. From this weight the loss in weight of 1.06 grams due to air displacement must be sub- tracted. CALIBRATION BY WEIGHING WATER. 235 weighing-tube capable of holding 50 c.c. is weighed. The water in the burette is now brought exactly to the zero-poiat and then about 5 c.c. is allowed to flow into the flask. The exact reading of the burette should be taken and the flask weighed. Another 5-c.c. portion of water is then allowed to flow into the flask and the burette again read and the flask weighed. This process is repeated until the last reading on the burette is exactly 50 c.c. and the flask contains the water delivered by the burette between zero and the 50-c.c. mark. The water need be weighed only to the hundredth of a gram, as volumes smaller than .01 c.c. cannot be measured on ordinary volumetric apparatus. The volumes corresponding to the weights of water are now calculated by the method given for the calibration of a flask, .997 gram of water being equal to 1 c.c. The difference between the volume indicated on the burette and that calculated from the weight of the water is the correction to be applied at that point. The correction for intermediate points is found by plot- ting a CTU-ve, as shown ia Fig. 44, in which the ordinates are the numbers 1 to 50 representing the readings on the burette, while the abscissas are the corrections found by the calibration. The assumption is made that the variations in volume are uniform between the points calibrated. Generally this does not lead to serious error. If the curve is very irregular, the burette should be calibrated for every 2 or 3 c.c. CALIBRATION BY MEANS OF THE MORSE-BLALOCK. BULBS. 335. Description of Bulbs. — ^If many pieces of apparatus are to be calibrated the Morse-Blalock calibrating bulbs should be used. These authors recommend the use of 6 bulbs of different sizes combined into 3 pieces of apparatus, as shown in Fig. 41. Three bidbs of capacities 2 c.c, 3 c.c, and 50 c.c, respectively, are all blown on one stem, various parts of which are graduated. Two bulbs of 50 c.c. and 200 c.c. capacity are connected on a single stem, which is also graduated, while the 500-c.c. bulb is on a sepa- rate graduated stem. Bulbs of other sizes could be made as desired. The intention of Morse and Blalock being to adapt the bulbs for use at a considerable range of temperature, they were 236 VOLUMETRIC METHODS. made of such a capacity that the volxune of a bulb at 30* should not be greater than that marked on it, while at 0° the combined vol- ume of the stem and bulb should not be less than that indicated on the bulb. 336. Calibration of the Bulbs. — ^The exact capacity of each bulb Fig. 41. is ascbrtained by filling it with distilled water and weighing the amount delivered on emptying it. The temperature of the water is noted and the volume of the bulb for this temperature is calculated. The temperature of the water need not be that at which the bulb is to be used, since the volume of the bulb having MORSE-BLALOCK CALIBRATING BULBS. 237 been ascertained for a given temperature, by using the coeffi- cient of expansion of the glass of which volumetric apparatus is made, .000026, the volume at any other temperat\ire can be cal- culated. The volume of the graduated stem is ascertained in the same manner. In using a bulb for calibration it is filled with distilled water, which is allowed to flow into the dry and empty flask imtil the bulb and a sufficient number of divisions on the stem are emptied to give the required volume. An example will make this clearer. Weight of Water from Bulb. Calculated Volume. At 17.5°. At 14.3°. At 20°. At 15°. 498.340 gr 499.51 c.c. 498.456 gr. 499.43 c.c. Av. 499. 47 c.c. 499.40 c.c. Weight of Water from Stem. At 14.6°. At 14.7°. 3. 177 gr 3 . 182 c.c. 3.180 gT. 3.185 c.c. Av. 3. 184 c.c. Volume of 1 division of stem = .032 c.c. Bulb +17 divisions of stem =500.00 c.c. at 20° (499.47+17 X.032). Bulb + 19 divisions of stem.=500.00 c.c. at 15° (499.40+19 X.032). 337. Temperature of Water Used in Calibration. — If it is de- sired to calibrate a flask to hold 500 c.c. at 20°, the water con- tained in the bulb and 17 divisions of the stem is allowed to flow into the flask and a mark made on the stem opposite the bottom of the meniscus. The temperature of the water need not be 20°. If the temperature of the water and consequently of the bulb is greater than 20°, the volume of the bulb and 17 divisions of the stem will be greater than 500 c.c, but on cooling to 20° it wiU be exactly 500 c.c. The temperature of the flask will be raised to that of the water, so that the volume of the flask will be greater than 500 c.c, but the flask also will contract on cooling, and as the bulbs are made of the same kind of glass as the flaslvs, the con- traction of the flask will be exactly equal to the contraction of the built. 238 VOLUMETRIC METHODS. The same line of reasoning will show that the temperature of the water may be below the standard temperature without affect- ing the accuracy of the calibration. 338. Calibration of Flasks. — By means of the 50-c.c. and 200-c.c. bulbs, flasks of capacities which are multiples of 50 c.c. may be calibrated. For flasks whose capacities are multiples of 500 c.c, the large bulb is used. Fig. 42. 339. Calibration of Burettes. — ^The 50-c.c. bulb with 2-c.c. and 3-c.c. bulbs are used for calibrating burettes. For this work CALIBRATION OF BURETTES. 239 as well as for calibrating flasks it is necessary to have a bottle holding several liters placed on a shelf which is above the top of the burette or bulb when fixed in clamps for use. A Friedrich- Qw«=s Fig. 43. Greiner two-way stop-cock is attached to the bulb so that water may be withdrawn from the bulb by two openings in the stop- cock. One of these tubes is connected with one arm of a T-tube, the second arm being coimected by a tube to the large bottle on 240 VOLUMETRIC METHODS. t\iv shelf, and the third one is connected with the burette. The latter must be clamped in such a position that the point giving the last reading is above the highest point of the calibrating bulb. All joints must be absolutely water-tight, and the connecting tubes should be of glass as far as possible. A pinch-cock should be placed on the tube leading to the large bottle, which is filled with distilled water. By opening this pinch-cock water may be admitted either to the burette or to the bulb. All bubbles of air must be displaced from the connecting tubes. 340. Cleaning Mixture. — The burette and calibrating bulb must be cleaned so that a continuous film of water is left on emptying them. A most excellent cleaning mixture is made by adding a few grams of potassium dichromate to a half liter of concentrated commercial sulphuric acid. This mixture should be kept in a glass-stoppered bottle to which portions which have been used may be returned unless largely diluted with water. Apparatus filled with this mixture and allowed to stand for some time will be perfectly clean on pouring out the cleaning mixture and washing with water. The cleaning mixture is much more efficient if warmed. When it has become green because of reduced chromium, it must be discarded or more chromate added. Burettes and pipettes must always be kept clean or errors arise in their use because of the drops of liquid which adhere to the glass. If burettes are filled with distilled water when not hi use, they keep cleaner than if left empty. 341. Total Capacity. — The bulb and burette are filled with water which is allowed to flow out at a moderate rate. The burette is again filled with water, and by cautiously turning the two-way stop-cock, water is allowed to flow from the burette into the bulb until the reading in the former is exactly zero. The stop-cock is now turned through 180°, and the 50 c.c. bulb emptied until the meniscus is opposite the mark on the stem. By again turning the stop-cock the water from the burette is allowed to flow into the bulb until the reading on the burette is exactly 50 c.c. The number of divisions of the stem of the bulb fiUed by the water is now noted. From this the total capacity of the burette can be calculated. Duplicates should agree within 1 or 2 divisions of the stem. The 50-c.c. bulb should be so constructed that its CALIBRATIOY OF BURETTES. 24 1 volume is not greater than 49.75 c.c, while the capacity of the stem should be at least .75 c.c, so that the total capacity of the stem and bulb shall be about 50.50 c.c. If the capacity of the burette is greater than 50.50 c.c, the stem may be emptied, and water again allowed to flow La from the burette. 342. Irregularities of Bore. — ^The burette is once more fiUed with water which is brought to the zero-mark. It is then allowed to fill either the 2-c.c., 3-c.c., or both bulbs, depending upon the accuracy desired in the calibration. The reading of the burette is carefully taken, the small bulb emptied, and the opera- tion repeated imtil the burette has been emptied. The capacity of the small bulbs should be a trifle less than that indicated, so that the last reading on the burette may always be less than 50 c.c. The method of calculating the results can best be given by the results of an actual calibration. Total capacity =50 c.c. bulb-|-78 divisions of stem. Total capacity =50 c.c. bulb+80 divisions of stem. Volume of 50 c.c. bulb =49.62 c.c Volume of stem (100 divisions) =.7807 c.c. Total capacity of burette =50.24 c.c. (49.62 +.0078X79). Readings. True Capacity. Corrections. 0.00 0.00 0.00 1.99 1.98 -0.01 3.98 3.96 -0.02 5.92 5.94 +0.02 7.90 7.92 +0.02 9.89 9.89 0.00 11.87 11.87 0.00 13.83 13.85 +0.02 15.81 15.83 +0.02 17.79 17.81 +0.02 19.77 19.79 +0.02 21.71 21.77 +0.06 23.68 23.75 +0.07 25.64 25.72 y0.08 27.60 27.70 fO.lO 29.58 ^9.68 +0.10 242 VOLUMETRIC METHODS. Leadings. True Capacity. Corrections. 31.55 31.66 + 0.11 33.50 33.64 + 0.14 35.48 35.62 + 0.14 37.45 37.60 + 0.15 39.40 39.58 + 0.18 41.38 41.56 + 0.18 43.32 43.53 + 0.21 45.29 45.51 + 0.22 47.28 47.49 + 0.21 49.24 49.47 +0.23 50.00 50.24 +0.24 The last reading being 49.24 and the true capacity at 50.00 being 50.24, the true capacity at 49.24 is found by the proportion 50 : 50.24:: 49.24 :x, where a; =49.47. As the 2-c.c. bulb has been filled twenty-five times with water, the volume of which is 49.47 c.c, the capacity of the bulb must be ^3- of 49.47, or 1.979. As at each successive reading of the burette one bulb-full of water has been taken out, the true volumes at these points are found by multiplying 1.979 by the numbers from 1 to 25. These values are given in the second column. The differences between these values and the readings of the burette are the corrections to be applied at these points. The correc- tions are plotted on co-ordinate paper as abscissa, while the cor- responding burette readings are the ordinates, as shown in Fig. 44. EXERCISE 45. Calibration of Morse and Elalock Bulbs. Clean one of the bulbs by filling with bichromate solution and, after ten minutes, empty it and wash with water. Clamp it in position with the Greiner-Friedrich stop-cock, which is connected with the large bottle of distilled water, as shown in Fig. 42. Fill with water and allow it to flow out at a moderate rate. If the film of water left on the glass breaks, leaving drops, the cleaning with the bichromate solution must be repeated. When the bulb is clean, fill again with distilled water and bring the lower part of the meniscus exactly on a line with the upper mark on the bulb. CALIBRATION CURVE. 243 CORRECTIONS '" CUBIC CENTIMETERS - tti ftp ■¥0.\ -t-O-z 0.0 +0 .1 +02 +03 go +ai +a2 +a3 Fig. 44. 244 VOLUMETRIC METHODS. Let the water flow at a moderate rate into a weighed flask until the menis- cus is at the lower mark on the bulb. Weigh the water to milligrams. Draw out another portion of the water into a flask and take its tempera- ture. Repeat the determination until duplicates are obtained which are identical to at least 1 part in 1000. Calculate the volume using the specific gravity of water at the temperature noted and a correction of +.106% for air displacement. Calculate the volume at 20°. In the same manner ascer- tain the volume of the graduated stem attached to the bulb. Calculate the number of divisions of the stem which must be used to make the volume equal to that marked on the bulb. EXERCISE 46. Calibration of Flasks. Clean and dry the standard flasks you have. Clean all calibrating bulbs to be used according to the directions given in the preceding exer- cise. Clamp the |-liter bulb in position as shown in Fig. 42. Fill the bulb to the upper mark and allow the water to flow into the liter flask until the bulb and the number of divisions on the stem giving 500 c.c. have been emptied. Repeat the operation and immediately make a mark on the neck of the flask opposite the bottom of the meniscus. Cali- brate the other flasks in a similar manner. A thin layer of beeswax or par- afiine should be melted on the neck and a circular hue passing through the file-mark made with a sharp steel point. The flask should be held in a lathe for this purpose. The mark is etched on by dipping a piece of filter- paper into hydrofluoric acid and moistening the mark with the acid. After about three minutes the acid is washed off with water. After wiping the flask dry, the paraffine is melted by waving the neck of the flask in the Bunsen-burner flame. The neck of the flask may then be cleaned by wiping with dry fi_lter-paper. EXERCISE 47. Calibration of Burettes. After cleaning your burette, set it up with the 50-c.c. bulb as shown in Fig. 4.3, p. 2.39. Test the apparatus for leaks by fillingwith water and wiping all joints dry with filter-paper. No drops of water should be visible after fifteen minutes. All joints except o and b maybe made tight by painting the glass tubing with shellac varnish before the rubber tubing is put on. The total capacity of the burette is first found by filling it ivith water by opening the stop-cock c and closing d. The water is carefully brought to the zero-mark by opening a (which should now remain open) and d. CALIBRATION OF BURETTES. 245 The stop-cock d is now turned so that water can flow mto the beakei until it is brought to the lower mark on the 50-c.c. bulb, d is again turned so that water is slowly admitted from the burette until it reads exactly 50 c.c. The number of divisions on the stem of the 50-c.c. bulb is now read. The determination is repeated until duphcates agreeing within 2 or 3 divisions of the stem are obtained. From the values already found for the 50-c.c. bulb, and for one division of the stem, the total capacity of the burette is calculated. The burette is again filled with water and the reading brought to the zero-mark. If the total capacity is not more than 0.2 c.c. over 50 c.c, the burette may be calibrated every 5 c.c. For this purpose the meniscus of the water in the calibrating bulb is brought to the mark below the 2-c.c. bulb and water allowed to flow in from the burette until the 3-c.c. bulb is filled to the mark on the stem. The reading on the burette is now taken and the operation repeated until the burette is empty. Repeat the cahbra- tion until duphcates are obtained for each reading agreeing within less than .02 c.c. 10 portions of approximately 5 c.c, but exactly equal, will now have been withdra^vn from the burette. The method of calcu- lating the re'sult is best given by the following example: A burette whose total capacity was found to be 50.15 c.c. gave as the last reading mth the 5-cc. portions 49.90 c.c. The volume at this point is found by the following proportion: 50.00 : 49.90 : : 50.15 : x, from which x, or the volume of 10 of the 5-c.c. portions withdrawn, equals 50.05 c.c Each 5-c.c portion is therefore equal to 5.005 c.c. This number multiplied by the nmnbers from 1 to 10 gives the true volumes at the burette readings taken. The differences between the readings and the true volumes give the corrections to be apphed at the points cahbrated. To find the corrections at intermediate points, plot a curve with the corrections as the abscissas, and the readings on the burette as ordinates, as shown in Fig. 44. If the difference in the corrections between any two consecutive readings is more than .05 c.c, this portion of the burette should be recalibrated with the 2-c.c. or the 3-c.c. bulb. PROBLEMS. Problem i. The weight of water at 20° C, which just fills a flask, is 249.2 grams. Calculate (a) the volume of the flask at 20°, (6) its volume at 15°, (c) its volume at 25°. Problem 2. A burette was calibrated with a Morse and Blalock bulb, of which the 50-c.c. bulb had a capacity of 49.55, and the stem of 100 divi- sions had a capacity of 0.73 c.c. When obtaining the total capacity of the burette, the bulb was filled and 67 divisions of the stem. When calibrating the bore of the burette, 5-c.c. portions were withdrawn from the burette by means of the 2- and 3-c.c. bulbs. The following readings were obtained : 5.05, 10.09, 15.09, 20.19, 25.19, 30.19, 35.15, 40.19, 45.19, 50.10. Calculate the total capacity of the burette and the corrections at the readings given. 246 VOLUMETRIC METHODS. Problem 3. Another burette was calibrated with the same bulb. For the total capacity, the bulb full plus 70 divisions of the stem was obtained. When calibrating the bore of the burette, the following readings were taken : 5.00, 10.11, 15.09, 20.09, 25.08, 30.09, 35.05, 40.09, 45.10, 50.10. Calculate the total capacity of the burette and the corrections at the readings given. The accuracy of the calibration of these burettes may be judged by a comparison of the volumes calculated for the 5-c.c. portions withdrawn from the burettes. These values should not differ by more than .03 c.c. CHAPTER XX. ACIDIMETRY. 343. A Standard Solution has already been defined as one which has been made so as to contain an accurately determined amount of a given chemical substance in a definite volume. It has been found that the acidity of a solution depends on the amount of replaceable or acid hydrogen present, irrespective of the amount of other substances present. A standard solution of an acid may therefore be defined as one which contains a known amount of acid or replaceable hydrogen in a definite volume. In practice it has been foxmd very convenient to use standard solutions of acids which contain 1 gram * of replaceable hydro- gen per liter. Such a standard solution is called a normal solu- tion. The strength of any standard solution may be given in terms of a normal solution. It may for instance be ^, J, 3^^, etc., normal, or 3, 3J, 1.362, etc., times normal. 344. Normal Acids. — ^The amount of a given acid to be dissolved in a liter to make a normal solution will vary directly as its molec- ular weight, and inversely as the number of acid or replaceable hydrogens in the molecule. Hydrochloric acid having a molec- ular weight of 36.458 (CI =35.45, H = 1.008), a liter of a normal solution of this acid must contain 36.458 grams of the acid. The molecular weight of sulphuric acid is 98.076 (S =32.06, O4 =64.00, H2 =2.016). As the molecule, however, contains two replaceable hydrogen atoms, one-half of the molecular weight, or 49.038 grams, must be dissolved in a liter to make a normal solution. Normal solutions of all monobasic acids must therefore contain per liter the weight of acid equal to the molecular weight in grams, while * Since the system of atomic weights having oxygen as 16.00 as the basis has come into use, 1.008 grams of hydrogen per liter has become the standard. 247 24S VOLUMETRIC METHODS. one-half the molecular weight in grams must be taken per liter for dibasic acido. As the acid value of a solution depends wholly on the amount of replaceable hydrogen present, and as the designation of the strength of a standard acid in terms of normal gives the amount of hydrogen per liter, the weight of the acid radicle need not be known for calculating the resiilts of a titration. If caustic soda is titrated with normal hydrochloric acid, one liter of the acid will neutraUze exactly 40.058 grams of NaOH(Na = 23.05, = 16.000, H = 1.008), according to the equation HCl+NaOH=NaCl+H20. If normal sulphuric acid were used the reaction would take place according to the equation H2S04+2NaOH=Na2S04+2H20. Two molecules of caustic soda or 80.116 grams per liter would be neutralized by one molecule of sulphuric acid or 98.076 grams per liter. But half of the latter amount having been taken for a liter of normal sulphuric acid, only one molecule of the caustic soda or 40.058 grams per liter of acid will be neutralized. It may be said then in general that a liter of a normal acid will neutralize 40.058 grams of caustic soda. If sodium carbonate were being titrated only half of the molec- ular weight must be taken, as one molecule of this base is capable of replacing two atoms of acid hydrogen, the molecular weight in grams, 106.10, being able to replace 2 grams of acid hydrogen. One liter of normal acid is therefore equal to 53.05 grams of sodium carbonate. One-thousandth of this weight or the amount per cubic centimeter is more convenient for use in calculation, being .040058 gram of caustic soda and .05305 gram of sodium carbonate. If the strength of the acid must be expressed in fractions or deci- mal parts of normal, the value per cubic centimeter will be found by multiplying the value per cubic centimeter of a normal acid by the fractions or decimals. The value in caustic soda per cubic centimeter of -^ normal acid will be ^\ of .040058 of .0040058. When many titrations of a given substance must be made the calculation is simpler if the acid is so made that 1 c.c. is equal to an amount of base, which can be expressed by a simple number such as .050 gram of sodium carbonate. If such an acid is used for titrating any other substance, the calculation is not as simple as when the value of the acid is expressed in terms of normal. ACIDIMETBY. 249 345. Percentage Given by Number of Cubic Centimeters of Acid Used. — The calcxilation of the result for a single substance may be made very simple indeed if the acid is made of such a strength that the percentage is given by the number of cubic centimeters used in titrating a given weight of the substance. For instance, 1 gram of sodium carbonate requires for its neutralization .6872 gram of hydrochloric acid. This follows from the fact that 1 molecule of sodium carbonate (106.10) is neutralized by 2 mole- cules of hydrochloric acid (72.916). If a standard solution of hydrochloric acid is made of such a strength that it contains 6.872 grams of acid per liter, then 100 c.c. of this acid will just neutral- . . ize 1 gram of pure sodium carbonate. If mixtures of sodium carbonate and neutral salts are analyzed, the percentage of sodium carbonate will be given by the number of cubic centimeters of the acid necessary to neutralize 1 gram of the mixture. The strength of the acid in terms of normal may be found from the ratio of 6.873 to 36.458, which gives .1885 normal. As burettes holding 50 c.c. are commonly used, it will generally be found more convenient when mixtures containing more than 50% of sodium carbonate are to be analyzed to make the acid twice as strong. Then the percentage will be twice the number of cubic centimeters used. The same object will also be accomplished if acid of the same strength is used and ^ gram of the material is weighed out for each determination. The percentage may also be given by the number of cubic centirrieters of a normal acid, provided a definite weight of the material is taken. In the case of sodium carbonate, 100 c.c. of a normal acid will be equal to 5.305 grams of the pure material, since, as has already been shown, one liter of the normal acid is equal to 53.05 grams of sodium carbonate. If 5.305 grams were weighed, the number of cubic centimeters of normal acid used would be equal to the percentage of sodium carbonate. It would be more convenient to weigh out 2.652 grams and multiply the number of cubic centimeters used by 2. If a fifth-normal acid were used, one-fifth or one-tenth of 5.305 grams would be taken. 250 VOLUMETRIC METHODS. INDICATORS. 346. Effect of Carbon Dioxide on Indicators. — ^The results obtained by titrating a given base with a given standard acid will differ according to the indicator used to give the end-point. This difference is due to the varying degree of sensitiveness of the indicators. For instance, a given amount of sodium carbo- nate will require about half as much standard hydrochloric acid for its neutralization if phenolphthalein is used as the indicator as when methyl orange is employed. This arises from the fact that when hydrochloric acid is added to a solution of sodium carbonate soditim chloride is produced and carbonic acid liberated. As carbonic acid is too weak an acid to change the color of methyl orange, the addition of the strong hydrochloric acid is continued until a slight excess is present, which gives the acid color to the indicator. Phenolphthalein, on the other hand, is sensitive to carbonic acid. Sodium carbonate reacts alkaline with phenol- phthalein until, by the addition of a strong acid like hydrochloric, one-half of the sodium has been converted into sodium chloride, and the remainder is present as the bicarbonate. Another drop of hydrochloric acid liberates a little carbonic acid, which changes the color of the phenolphthalein. If the solution is concentrated or hot, carbon dioxide will escape, so that if the solution is boiled the titration of sodium carbonate will give the same result with either indicator. As carbon dioxide is present in the air, in ordinary distilled water, and in most solutions of acids as well as alkalies, imless special precautions are taken to exclude it, the analyst must be thoroughly familiar with the behavior towards this acid of all indicators used. As has already been stated, methyl orange is almost entirely unaffected by it, while phenolphthalein is sen- sitive to one of the two hydrogen atoms of the acid. Carbonic acid may therefore be said to be monobasic to phenolphthalein in cold dilute solution. Its basicity is zero in boiling solution with phenolphthalein, as well as with methyl orange both hot and cold. Much valuable work on indicators has been carried out by R. T. Thomson,* who has divided them into three classes. * C. N., V. 47, p. 123; 185, v. 49, pp. 32, 119. J. S. C. I., v. 6, p. 195. INDICATORS. 251 The first class, of which methyl orange is the best example and the one most commonly used, comprises lacmoid, dimethyl amido- benzene, cochineal, and congo red. This class is most sensitive to alkalies and can be used with strong acids only, being unaffected by weak acids such as carbonic, silicic, etc. The second class, of which phenolphthalein is the type and the one most commonly used, includes also turmeric. These indicators are extremely sensi- tive to acids, and are therefore especially adapted to the titration of weak acids among which most organic acids must be placed. Strong bases must be used with these indicators, and carbon dioxide must be carefully excluded. The third class, of which litmus is the type, includes rosolic acid and phenacetolin. This class of indicators is about midway between the other two in sensitiveness, being fairly sensitive to both weak acids and weak bases. 347. Basicity of Acids with Various Indicators. — ^The table on page 252 prepared by Thomson gives the number of molecules of a univalent base, such as caustic soda or potash, neutralized by a molecule of each acid in the presence of the indicator at the head of the column. A blank indicates the absence of a sharp end-point. The same results are obtained with calcium or barium hydrate where insoluble compounds are not formed. 348. Litmus, although one of the earliest indicators used, has by no means been superseded by the many indicators recently introduced, especially by the workers in organic chemistry. It is considerably more sensitive than methyl orange, not only being more sensitive to weak acids, but giving an end-point more readily distinguished by most workers. Carbon dioxide interferes seri- ously, so that to obtain the best results this acid must be quite completely excluded. This may be readily done by boiling the solution. This operation does not destroy the indicator, but as it is considerably more sensitive cold than hot, the solution should be cooled before the titration is completed. For this reason methyl orange is more commonly used when considerable carbon dioxide is present. As obtained in commerce litmus consists of small cubes com- posed of powdered chalk saturated with the coloring-matter, which is a vegetable extract. The coloring-matter is not a single 252 VOLUME TRW METHODS. Acids. Methyl Orange. Phenolphthalein. Litmus. Name. Formula. Cold. Cold. BoUing. Cold. Boiling. Sulphuric Hydrochloric Nitric H2SO4 HCl HNO3 H2S2O3 H2CO3 H2SO3 H2S H3PO4 HjAsO, H3ASO3 HNO2 H4Si04 H5BO3 H2Cr04 H2C2O4 HC2H3O2 HC4H,02 H2C4H4O4 HC3HA H2C4H40e H3C„H,0, 2 1 1* 2 1 1 1 indicator destroyed 1 2 1 1 2 1 dilute 2 1 dilute 2 2 1 t 2 2 1 1 2 1 2 3 2 1 1 2 2 2 2 1 1 2 1 2 1 nearly 1 nearly 2 nearly 1 2 2 1 1 Thiosulphuric .... Carbonic Sulphurous Hydrosulphuric. . Phosphoric Arsenic 2 Arsenious Nitrous Silicic Boric Chromic ■■ Oxalic 2 Acetic Butyric Succinic Lactic Tartaric Citric * Concentrated nitric acid sometimes contains oxides of nitrogen, producing on dilution nitrous acid which destroys methyl orange. 1 1 in a solution containing 30% of glycerine, or a few grams of mannitol. substance. Various methods have been proposed for obtaining the pure coloring-matter for use as a sensitive indicator. One of the simplest and best methods consists in extracting the cubes two or three times with hot 85% alcohol. By this treatment a violet coloring-matter is removed. The residue is then extracted several times with water, the first portion being rejected, as it contains the bulk of the alkali carbonates. The remainder of the water-extract is acidified with dilute sulphuric acid and boiled for some time to completely expel carbonic acid. The excess of sulphuric acid is then neutralized with barium hydrate. The neutral color is violet, which a drop of alkali turns to a decided blue, and a drop of acid to a decided red. The solution should be INDICATORS. -53 preserved from mould by the addition of a few drops of chloroform. If kept in a closely stoppered bottle, the color disappears, but is restored on exposure to the air. The solution is best pre- served in a wide-mouthed bottle partially filled, and protected from dust by a cover of filter-paper, or a loose cork stopper, through which passes a glass tube with which to withdraw the solution as needed. 349. Cochineal solution is very similar to litmus solution. When acidified the color is red, which is turned to violet by alkalies. It is not as sensitive to carbonic and weak organic acids as litmus. This gives it an advantage over the latter, as carbon dioxide need not be so rigidly excluded. It cannot be used, however, in the presence of even traces of compounds of iron or aluminium c acetates. It is used largely in the titration of ammonia in the Kjeldahl method for determining nitrogen. The solution is made by digesting 1 part of the crushed cochineal with 10 parts of 25% alcohol. 350. Methyl Orange is the sodiimi or ammonium salt of para- dimethyl-aniline-azo-benzene-sulphonic acid. It is a synthetic product, and can generally be purchased in a sufficiently pure condition for use. It is dissolved in water, 1 gram per liter giving a solution of which 1 drop is sufficient for most titrations. A large amount of this indicator decreases the sharpness of the end- point. An amount just sufficient to give a faint yellow color to the alkaline solution should be used. The solution should be kept in a bottle provided with a stopper, through which a drawn-out glass tube passes so that the indicator may be introduced by drops. Methyl orange is very largely used because it is comparatively unaffected by carbonic acid as well as by sulphuretted hydrogen, hydrocyanic, silicic, boric, arsenious, oleic, stearic, palmitic, and carbolic acids, etc. The base in the salts of these acids may be titrated without removal of the acid. It has found a very ex- tended use in the standardization of acids by means of pure sodium carbonate. In the presence of carbonic acid the change from the yellow alkaline to the red acid color is somewhat gradual-. It is difficult to remember the exact shade corresponding to the neutral point, especially as a slight reddish tint is imparted by 254 VOLUMETRIC METHODS. the carbon dioxide. It is therefore advisable to prepare a solu- tion for comparison which is given this tint by using distilled water saturated with carbon dioxide which has been freed from mineral acid by passing through a solution of sodium bicarbonate. The beaker containing this solution should be of the same size and contain the same volume of water as the beaker in which the titration is made. The same amount of indicator is also added. At the end-point of the titration a color a little deeper than that of the carbon dioxide solution must be obtained. The- indicator is most sensitive in a cold solution. The presence of alcohol greatly decreases its sensitiveness. It cannot be used in solutions containing nitrous acid, as this acid destroys the indi- cator. If only small amounts are present, the titration may be completed before the indicator is destroyed. If large amounts are present, excess of standard alkali must be added, and the solution titrated back with standard acid. 351. Phenolphthalein is a synthetic organic product which may readily be purchased in a pure condition. It is dissolved in 90 to 95% alcohol, 1 or 2 grams per liter. It is advisable to first distil the alcohol after adding a little caustic soda or potash, as the commercial product is seldom free from organic acids. One or two drops of this solution are used. In neutral or acid solutions it is colorless, but the faintest excess of alkali gives a sudden change to purple-red. Carbon dioxide must be entirely absent, as in its presence the indicator reacts acid until it is con- verted into the bicarbonate by the alkali. Small amounts of carbonates in presence of alkali hydrates may be titrated accu- rately by means of phenolphthalein. Acid is added until the pink color has disappeared. All of the alkali hydrate will then be neutralized, and the alkali carbonate converted into bicarbo- nate. On adding methyl orange and standard acid until the red acid color of this indicator is developed the bicarbonate will be decomposed. From the amount of acid added for this purpose the amo\mt of carbonate originally present may be calculated. The amount of carbonate must be small, as even the bicarbonate gives a slight color to the phenolphthalein.* Ammonia must * Kiister, Zeit. i. anorg. Ch., 13, 127. INDICATORS. 255 also be absent, as in its presence the end-point is not sharp. Phe- nolphthalein is especially valuable for the titration of weak acids, especially the organic acids. Many of these acids being insoluble in water are dissolved in alcohol and then titrated. This does not seem to be quite as accurate as titration in water solution. A strong base must always be used with phenolphthalein. As water solutions of caustic soda and potash always contain carbon dioxide, an alcoholic solution of the latter is sometimes used, and more frequently a solution of barium hydrate. 352. Best Acid for General Use. — As has already been said, the acidity of a solution depends only on the amount of replace- able hydrogen present in a given volume, irrespective of the nature of the acid radicle. As has been shown, the acid value of the different hydrogen atoms differs when two or more are present in the same molecule, so that the acidity of such a solu- tion may differ widely with different indicators. For general use, therefore, a strong acid should be chosen so that the strength of the solution may be the same with all indicators. A stable non-volatile acid is also desirable so that a solution once stand- ardized may be permanent. These considerations limit us to the three commonly used strong acids, hydrochloric, sulphuric, and nitric. Of these three only the first two have found any very general use. The standardization of nitric acid offers some difficulties because it forms no insoluble weighable salt. It is also an oxidizing agent, so that titrations may be disturbed by secondary reactions. On the other hand, both sulphuric and hydrochloric acids are very stable and may also be precipitated and weighed with great accuracy. Choice must therefore be made between these two acids for a carefully standardized perma- nent acid to be used in the standardization of alkalies and other acids. For general use hydrochloric acid offers several decided advantages over sulphuric acid. It is appreciably stronger, so that the end-point with most indicators is considerably sharper than with sulphuric acid. This is partly due to the fact that there are two atoms of hydrogen in the sulphuric acid molecule, so that when the solution is partly neutralized an acid sulphate is formed whose hydrogen atom is not so strongly acid as the hydro- gen atom first displaced. The use of sulphuric acid is also lim- 256 VOLUMETRIC METHODS. ited by the fact that it forms insoluble sulphates with the alkaline- . earth metals. The gravimetric standardization of hydrochloric acid by precipitation as silver chloride is subject to fewer sources of error than the standardization of sulphuric acid by precipita- tion as barium sulphate. 353. Most Desirable Strength of Standard Acids. — Various strengths of acid are in use among chemists, many using tenth- or fifth-normal acids, while among commercial chemists the use of half- or full-normal solutions is quite common. The use of such concentrated solutions is defended by the argument that a sharp end-point may always be obtained with one drop of the acid. This is advantageous in commercial laboratories, where titrations must be made rapidly, and as large amounts of material must be weighed out to use a proper volume of acid, the weighings may be more quickly made. This is evident from the consideration that an error of 5 mg. on 5 grams is the same percentage as an error of 1 mg. on 1 gram. As methyl orange must frequently be used in the presence of carbon dioxide, the strong acid gives a sharp end-point even with one drop of the standard solution. The use of the more dilute solutions is defended by the argu- ment that a larger volume of the solution must be used for titrat- ing a given weight of material, and therefore the reading of the volume on the burette is more accurate. The indefinite dilution of the solution does not lead to greater and greater accuracy, since a point is soon reached where a greater and greater volume of the solution must be taken to give a distinct end-point with the indicator. This limit is usually reached with fifth- or tenth-normal solutions. With moderate care a distinct end-point may be obtained even with methyl orange in the presence of carbon dioxide by means of one drop of fifth-normal acid. Another decided advantage of the fifth-normal acid is found in the mod- erately small coefficient of expansion with change in temperature, which with stronger acids becomes quite considerable. In gen- eral, therefore, more accurate work can be done with the more dilute acid, while time is saved at the expense of accuracy by the use of strong acids. The final decision must therefore be made on the requirements of the work at hand. For the exercises given in this book fifth-normal solutions will be uniformly used. PROBLEM. 257 PROBLEM. Problem 4. Calculate the number of grams of impure potassium car- bonate which must be weighed out, so that when the solution is titrated with N/5 acid, the number of cubic centimeters used shall be equal to the per cent of potassium carbonate. Calculate the strength of acid in terms of normal to be made up, so that if 1 gram of impure potassium car- bonate is weighed out, the per cent of potassium carbonate will be equal to the number of cubic centimeters used multiplied by two. CHAPTER XXI. STANDARD ACIDS. STANDARDIZATION. One of the most important points in volumetric work is the accurate standardization of the solutions. No subsequent care can overcome an error in this part of the work. Although many- excellent methods of standardizing acids have been proposed, considerable skill and experience are necessary in order to secure absolutely reliable results. 354- Sodium Carbonate. — One of the oldest as well as most reliable methods of standardization is by means of sodium car- bonate. It may be purchased in a high state of purity, so that it frequently needs only to be heated to completely dehydrate it and decompose any bicarbonate which may be present. For this purpose it need not be heated over 300°.* It is better to heat to this temperature than, as has long been the custom, to heat to dull redness and below the fusing-point of the carbonate, as small amounts of the carbonate are apt to be converted into the hydroxide by this treatment. The carbonate must first be tested for impurities. It should dissolve completely in water. A portion of 2 or 3 grams is tested for chlorides by dissolving in water, acidifymg with dilute nitric acid, and adding silver nitrate. Only a slight opalescence is per- missible. Another 2- or 3-gram portion is dissolved in water, acidified with hydrochloric acid, and tested for sulphuric acid by means of barium chloride. A very pure sodium carbonate may be obtained by heating PURE SODIUM BICARBONATE to 270°-300° for onc-half to one hour. The bicarbonate must be tested for impurities by the methods * Lunge, Zeit. f. angew. Chem., 1897, 522. 258 STANDARD ACIDS. 259 given for the carbonate. Small amounts of chlorides and sulphates may be washed out by repeated treatment with small amounts of cold distilled water. The bicarbonate may also be recrystallized. Distilled water heated to 80° in a Jena beaker is saturated with the bicarbonate by adding small amounts and stirring. The saturated solution is filtered through a folded filter in a hot-water funnel, and the solution cooled and vigorously stirred as the salt crystallizes out. It is filtered off in a funnel closed with a platinum cone or porcelain plate, washed with small portions of cold water and dried in a porcelain dish. Impure sodium carbon- ate may be purified by making a cold saturated solution, filtering if necessary, and saturating with carbon dioxide which has been washed by passing through a solution of sodium bicarbonate. The precipitated bicarbonate is filtered off and washed as already directed. It is advisable to preserve the bicarbonate as such and convert suitable portions into carbonate immediately before weighing it out for use. Better results will also be obtained if separate small portions are weighed out for each titration instead of making a standard solution, portions of which are measured out. The errors of the volumetric apparatus are thus avoided. 355. Sodium Oxalate (Sorensens).* — Sodium carbonate of the highest purity may be prepared from sodium oxalate which is purified as follows: A saturated solution of the sodium oxalate is prepared. After making it slightly alkaline with caustic soda the solution is allowed to stand in a beaker or cylinder until it is entirely clear. Various impurities, especially calcium oxalate, are precipitated. The solu- tion is filtered and then concentrated on the water-bath, until one-tenth has vaporized. On cooling, the sodium oxalate crystal- Uzes out, leaving potassium oxalate, chlorides, sulphates, and caustic soda in solution. The sodiiun oxalate is filtered off and washed several times with cold water. The sodium oxalate is redissolved and after concentrating reprecipitated. This is re- peated until the mother hquor is free from sulphuric acid and contains so httle caustic soda that 100 c.c. will be neutralized to litmus by a drop of normal acid. The sodium oxalate is dried at 240° C. and is not hygroscopic. Prepared in this manner, • Zeit. f. Anal. Ch., 26, 639. / 259a VOLUMETRIC METHODS. sodium oxalate is crystalline. It may be obtained in a less dense form by precipitation with alcohol. The hot satm-ated solution is filtered and allowed to drop into alcohol. The precipitate is is filtered off, washed and dried at 240° C. The purity of the preparation may be ascertained by the following tests.* Water. — 10 grams should not lose more than 0.001 g. in weight after dr)ang for twenty-four hours at 100°. Sodium Carbonate and Sodium Acid Oxalate. — Introduce into a conical Jena flask about 250 c.c. of water and 10 drops of phenol- phthalein solution and evaporate to about 180 c.c. while passing in a current of air free from carbon dioxide. Allow to cool to the ordinary temperature and add 5 grams of sodium oxalate, shaking carefully and maintaining the current of air. The oxalate slowly dissolves. If the solution becomes red, not more than 4 drops of decinormal acid should be required to render it colorless. If the solution is colorless, not more than 2 drops of decinormal sodium hydroxide should be required to produce a distinct red color. Chlorides and Sulphates. — Decompose 10 grams of sodium oxalate by heating in a platinum crucible. If the illuminating gas contains sulphur, an alcohol lamp should be used. The sodium carbonate is dissolved in dilute nitric acid free from chlorine and the carbon filtered off. One-half of the solution is tested for chlorides by the addition of silver nitrate. The other half of the solution is tested for sulphates by means of barium chloride. Clhorides and sulphates should be absent. Iron and Potassium. — Decompose 10 grams of sodium oxalate by gently igniting in a platinum crucible, removing the last traces of carbon by gently igniting with the blast-lamp. The residue should be dissolved in water and transferred to a plat- inum dish. There should be at most only a scarcely weighable trace of undissolved ferric oxide. The solution is filtered, if nec- essary, and acidified with hydrochloric acid free from iron. It is then evaporated in the platinum dish to dryness on the water- bath and the residue dried for two hours at 120°. The residue must dissolve clear in water and should give (o) With sulphocyanate only a faint reaction for kon. (6) With sodium cobalt nitrite solution no test for potassium. * S. P. L. SSrensen, Zeit. f. Anal. Chem. 42, 572. STANDARD ACIDS. 2596 Organic Impurities. — In a clean test-tube heat 1 gram of sodium oxalate with 10 c.c. pure concentrated sulphuric acid as long as gas is evolved, first gently and then more strongly until the acid begins to boil. When cool, the color of the sulphuric acid is compared with that of another 10-c.c. portion similarly treated, but without the addition of the sodium oxalate. The sodium oxalate should not impart more than an exceedingly faint brownish tinge to the sulphuric acid. The purified and tested sodium oxalate may be converted nto sodium carbonate as follows: The sodium oxalate is dried to constant weight at 100° C. in a platinum crucible. It is gently heated at first with the lid on the crucible. The decomposition of the sodium oxalate into sodium carbonate is generally complete in fifteen minutes to half an hour. The small amount of carbon which generally remains is burned by heating the half-covered crucible until the sodium carbonate begins to melt. The weight of sodium carbonate is calculated from the weight of the sodium oxalate taken, one gram of sodium oxalate giving .7910 gram of sodium carbonate. A factor for calculating the acid direct from the sodium oxalate may also be used, 1 gram of sodium oxalate being equal to .2722 gram hydrochloric acid and .3660 gram sulphuric acid. 355a. Crystallized Oxalic Acid. — Extended use has also been made of crystallized oxalic acid. A solution of an alkali, such as caustic soda or potash, is first standardized by means of the pure oxalic acid. The acid solution to be standardized is then care- fully compared with the standardized alkaline solution. This method has the disadvantage that the errors of two titrations are introduced into the standardization. Carbon dioxide must also be absent as litmus or phenolphthalein must be used as the indicator. Two difficulties are met with in preparing pure oxalic acid — the removal of traces of the alkalies and the drying of the crystal- lized product so as not to lose water of crystallization. The first object is attained by dissolving the commercial article in a mixture of equal parts of alcohol and ether, in which the alkali oxalates as well as other impurities are insoluble. After filtering the solution, the bulk of the alcohol and ether are distilled off, water is added, and the distillation continued. Finally the water solution is transferred to a porcelain dish and kept near 260 VOLUMETRIC METHODS. the boiling-point until the last traces of the alcohol and ether as well as any ethereal salts which may have been formed are vola- tilized. The steam will then have no odor. If the oxalic acid begins to crystallize out of the hot solution during this process, small quantities of distilled water are added from time to time. When the volatile matter has been expelled, the solution is allowed to cool with constant stirring. The crystals are filtered off in the usual manner and washed with small quantities of cold distilled water. They are then spread out on unglazed porcelain and exposed to the air until dry. The material is preserved in glass- stoppered bottles. The crystals must not be exposed to the air after they are dry, as shown by their non-adherence to a clean dry glass surface. The crystals then have the formula HA04.2H,0. 356. Anhydrous Oxalic Acid. — ^To obviate the uncertainty of the degree of hydration, anhydrous oxalic acid has been somewhat used. Drying at 60° to 80° for several hours is said to secure an anhydrous product. At 100° considerable oxalic acid is volatil- ized. As the anhydrous material is quite hydroscopic, it must be cooled in a desiccator and weighed as soon as cool. The pure recrystallized product prepared as already directed must be used for this purpose. Solutions of oxalic acid are not entirely stable, especially if dilute. For standardizing purposes it is therefore advisable to weigh out convenient amounts, and, after drying and weighing, to dissolve in water and titrate immediately. 357. Potassium Tetroxalate. — ^In place of the crystallized dehy- drated oxalic acid, the use of crystallized potassmm tetroxalate KHC2O4.H2C2O4.2H2O has been recommended. This salt may easily be prepared by making a saturated solution of oxalic acid and filtering if necessary. One-fourth of this solution is neutral- ized with pure potassium carbonate and added with stirring to the remainder of the oxalic acid solution. The crystals which separate out are recrystallized from hot water several times and dried as directed for crystallized oxalic acid. The salt is pre- served in well-stoppered bottles. 358. Potassium Dichromate. — The suggestion of Richter * to use pure potassium dichromate which may be titrated against a _ * Zeit. f. anal. Ch., 21. 205. STANDARD ACIDS. 261 solution of caustic potash, using phenolphthalein as the indicator, would seem to be of value. Potassium dichromate may readily be prepared pure and anhydrous. 359. Gravimetric Standardization. — The precipitation and gravimetric estimation of the acid radicle offers a most excellent method of standardizing acids, especially sulphuric and hydro- chloric. The details of the precipitation and weighing of sul- phuric acid as barium sulphate, and of hydrochloric acid as silver chloride, have already been given. A very simple gravimetric method consists in evaporating to dryness in a weighed platinum dish a measured volume of the acid solution which has been neutralized with ammonia. Both ammonitmi chloride and ammo- nium sulphate may be dried at 100° without appreciable volatili- zation. Very pure ammonia may be obtained for this purpose by distilling the concentrated solution and absorbing the gas in distilled water. Test-tubes may be used for this purpose, and the distilled ammonia should be used immediately. The weighed ammonium salt may be still further tested for impurities by vola- tilizing the weighed salt by gently heating the dish with the Bunsen burner and weighing the residue. 360. Standardization by Specific Gravity. — Standard solutions of acids, especially sulphuric acid, may be made with almost as high a degree of accuracy as by the methods already given, by carefully taking the specific gravity of the concentrated acid and diluting a measured or weighed amount to a definite volume. The C. P. concentrated sulphuric acid should first be diluted to about 30%. Acid of this concentration is not hydroscopic, and the change in specific gravity with change in concentration is greater than with the stronger acid. The specific gravity is carefully taken at the standard temperature, 15° C, by means of a Westphal balance or carefully calibrated pycnometer. The greatest care should be taken to bring the temperature to exactly 15° C. with a tested thermometer. If the acid is to be measured out it must be kept at this temperature as carefully as when the specific gravity is being taken. This difficulty may be obviated by weighing the acid. This method has the additional advantage that the weight can be taken far more accurately than the volume, especially if it is corrected for air displacement. The weighing 262 VOLUMETRIC METHODS. need not be made closer than a few milligrams, as an error of 10 mg. is negligible. The calculation is also very simple. The number of grams of acid required per liter for the solution to be made is divided by th", percentage given in the table for the spe- cific gravity found. The quotient multiplied by 100 is the num- ber of grams of the concentrated acid to be weighed out and diluted to a liter. As the measuring out of concentrated hydrochloric acid is accompanied by much more uncertainty, it is advisable to measure out a little more of this acid than the calculated amount, so that after standardization it may be diluted to the exact strength. A solution of hydrochloric acid of approximately constant strength may be obtained by boiling either the dilute or concentrated acid. Equilibrium is soon reached so that, when boiling at about 760-mm. pressure, both acid and water are volatilized in equal amount from the residue, which contains 20.2% of hydrochloric acid and has a specific gravity of 1.10.* 180 grams of this acid diluted to 1 liter give an acid slightly stronger than normal. 361. Dilution of Acid to Exact Strength. — If the sulphuric acid is not made by diluting a strong acid whose specific gravity has been very carefully taken, it should also be made so as to be somewhat stronger than the acid required. At least 1^ liters should be made, placed in a bottle of suitable size and well mixed by shaking. After carefully standardizing the acid by two of the methods already given, it should be diluted to the exact strength. Perhaps the simplest method of calculating the results of each standardization is to find the number of grams of acid present per liter. The average of the standardizations, which agree within at least 0.2%, is taken and divided into the number of grams per liter required. For normal hydrochloric this will be 36.46 grams and for normal sulphuric 49.04 grams. The quotient multiplied by 1000 gives the number of cubic centi- meters to be diluted to 1 liter. This number subtracted from 1000 gives the number of cubic centimeters of water to be measured out into the liter-flask, which is then filled to the mark with the acid to be diluted. The strength of the diluted acid is verified * Roscoe and Dittmar, Jour. Chem. Soc. XII, 12S, 1S60. STANDARD ACIDS. 263 by one of the methods already used. This is especially necessary if the amount of water added per liter is large. Frequently the most satisfactory method of standardization is by the use of sodium carbonate, this being a volumetric method, and is carried out luider the same conditions that will be met in regular work. EXERCISE 48. Preparation of Standard rf/5 Hydrochloric Acid. To 75 c.c. of concentrated hydrochloric acid an equal volume of water is added and the solution boiled in a beaker or a porcelain dish for a few minutes. About 36 grams of this acid are weighed out and transferred to the liter flask, which should be clean, but need not be dry. The beaker or other vessel is rinsed out into the flask. Distilled water is added until the flask is filled to the mark. The solution is poured into a large bottle,* which, if not dry, should first be rinsed with a little of the acid. The flask is rinsed with a little distilled water, and a second liter of acid made and poured into the large bottle. The bottle is filled from a third liter of the diluted acid. After thoroughly mixing the solution by shaking, the acid is standardized by two of the following methods, duplicate determinations being made by each method. STANDARDIZATION. 362. First Method. By Evaporation with Ammonia. — Measure out with a burette or a cahbrated pipette 50 c.c. of the acid into a weighed plati- num dish. Fit a small flask or a test-tube -with a glass deKvery-tube bent tAvice at right angles. Place 10 to 15 c.c. of strong ammonia in the flask and distil off the ammonia, using a small flask or test-tube containing 15 to 20 c.c. distilled water as a receiver. The end of the delivery-tube should dip into the distilled water. Neutralize the acid with this ammonia, evapo- rate to dryness on the water-bath, and weigh the ammonium chloride. Bring to constant weight by heating on the water-bath for fifteen to twenty minutes and weighing. When the weight is constant volatilize the ammo- nium chloride by gently heating the dish -with the Bunsen burner, and weigh again. Deduct this weight from the weight of the dish and the residue dried on the water-bath to obtain the weight of the ammonium chloride. Calculate the weight of hydrochoric acid present in 50 c.c. by the proportion, Mol, wt. NH4CI : Mol. wt. HCl : : weight NH^Cl : weight HCl. * The 2^-liter glass-stoppered bottles in which the concentrated C. P. acids are purchased are very convenient for this purpose. 264 VOLUMETRIC METHODS. 363. Second Method. By Titration with Sodium Carbonate. — Place about 4i grams of pure sodiiim bicarbonate in a weighed platinum crucible. Place it in a sand-bath so that the crucible is nearly immersed in the sand. Insert a thermometer in the bicarbonate and heat the sand-bath until the thermometer registers 270° to 300° for one-half to one hour. If an air-bath is at hand which can be raised to this temperature, the bicarbonate may be heated in it for the same length of time. It may then be placed on a weighed watch-crystal. The carbonate is cooled in the desiccator and weighed. The heating is repeated for half-hour periods until constant weight is obtained. The weight of the carbonate is then brought to exactly 2.650 grams by taking out small portions with a spatula. Transfer the carbonate to a beaker, rinse the crucible with water; dissolve the material in water, and transfer to a 250-c.c. flask. After diluting to the mark and shaking thoroughly, fill a burette vnth the solution and titrate against the hydrochloric acid solution, using methyl orange as the indicator. The titra- tion should be made the same day the carbonate is dissolved, as the solution is not entirely stable because of action on the glass. Measure out 2.5 c.c. of the alkaline solution into a 100-c.c. beaker, and add one drop of the indicator. 50 c.c. of distilled water are placed in a similar beaker, one drop of methyl orange added, and the solution saturated with carbon dioxide which has been passed through a solution of sodium bicar- bonate. To the beaker containing the alkali hydrochloric acid is added until, when compared with the neutral solution, a pink tinge is noticeable. A drop of alkah is then added, and if the neutral tint is not restored, the solution is made alkaline and again neutralized by cautious additions of the acid. This process is repeated until no difficulty is experienced in so closely judging the acid tint that one drop of the alkali will restore the neutral tint. When a satisfactory end-point has been obtained both burettes are carefully read. The titration is repeated until duplicates agreeing within less than 0.1 c.c. are obtained. When skill has been attained in making the titration, more accurate results can be obtained by weighing out portions of the carbonate just suffi- cient for one titration. A considerable amount of the bicarbonate is con- verted into carbonate and brought to constant weight. Portions weighing from 0.4 gram to not over .5265 gram are weighed out and transferred to beakers. After dissolving in about 50 c.c. of distilled water and adding a drop of methyl orange the solutions are titrated with the hydrochloric acid. Calculation. — By dividing the weight of sodium carbonate by the num- ber of cubic centimeters of acid used in each case the weight of sodium carbonate per cubic centimeter will be obtained, which for an exactly X/S normal solution is .01061 gram. The amount found per cubic centimeter should be divided into .-01061, and the quotient multiplied by 1000 to find the number of cubic centimeters to be diluted to one liter to make the acid exactly fifth-normal. STANDARD ACIDS. 265 364. Third Method. By Determination of Chlorine as Silver Chloride. — Prepare two Gooch crucibles and dry on the hot plate. Measure out two 25-c.c. portions of the acid into Erlenmeyer flasks and dilute to 200 c.c. Add silver nitrate solution with constant agitation until all the chlorine is precipitated. Heat nearly to boiling, and shake vigorously until the supernatant Uquid is clear. Decant the clear hquid through the Gooch crucibles, wash two or three times by decantation with hot water, transfer the precipitates to the crucibles, and wash wdth hot water until free from silver. Dry on the hot plate and weigh. If the duplicates differ from each other by more than 1 mg., repeat the determination, using the same crucibles without removing the precipitates. Calculation and Dilution of Acid. — From the average weight of silver chloride obtained from 25 c.c. of acid the strength of the acid is obtained by the following proportion : Mol. wt. AgCl : mol. wt. HCl : : weight AgCl : X, where X is equal to the number of grams of hydrochloric acid in 25 c.c. This number divided into 7.292 and the quotient multiplied by 25 r7.292 ,n . . . , (1000) gives the number of cubic centimeters to be diluted to 1000 to give a fifth-normal solution. This number subtracted from 1000 gives the nimiber of cubic centimeters of water required. If the liter-flask is wet, it should be rinsed two or three times with a little of the acid to be diluted. The necessary amount of water is carefully measured out with a burette and allowed to flow into the flask, wliich is then filled to the mark with the acid and thoroughly shaken. ' The diluted acid is emptied into a bottle capable of holding all of the acid to be diluted. If it is not dry, it is rinsed with a little of the diluted acid. The remainder of the acid is diluted in the same manner and poured into the large bottle. The strength of the diluted acid must be verified by one of the methods already used. This is especially necessary if the amount of water added is large. DETERMINATION OF SODIUM HYDROXIDE, CARBON- ATE, AND BICARBONATE. These three compounds of sodium evidently cannot occur together, because the hydroxide and bicarbonate would react to form the normal carbonate. We have therefore to consider the determination of sodium hydroxide and carbonate when occurring together, as well as the analysis of mixtures of the bicarbonate and the carbonate. 365. Titration of Carbonic Acid. — One method of analyzing the first combination has already been outlined in the discussion 266 VOLUMETRIC METHODS. of the properties of indicators. Using phenolphthalein as the indicator, all of the hydroxide will be neutralized by standard acid, while the carbonate will be converted into bicarbonate. On adding methyl orange and continuing the addition of acid until the solution is pink the bicarbonate formed will be decom- posed. The amount of acid added during the second part of the titration will be half that required to completely decompose the carbonate. The difference between the two amounts of acid used win be the amount necessary to neutralize the sodium hydrox- ide. This method will give correct results only if the amount of carbonate is small and the solution is dilute and cold. The escape of carbon dioxide because of a local excess of acid must be abso- lutely prevented by vigorous stirring. The method is well adapted to the analysis of commercial caustic soda. 366. Precipitation of Carbonic Acid. — If a large amount of sodium carbonate is present, the total amount of alkali present is determined by titration with standard acid, using methyl orange as the indicator. From another portion the carbon dioxide is precipitated by the addition of barium chloride solution. The alkali is in this manner converted into chloride, and on titrating the solution only the alkali hydroxide remains to react with the acid. Any excess of barium chloride will remain as such or react with the sodium hydroxide, giving sodium chloride and barium hydroxide. It has been shown, however, that the excess of barium chloride must be small, as otherwise the percentage of caustic soda found will be low. Using phenolphthalein as the indicator, the alkali hydroxide may be titrated in the presence of the barium carbonate, if care is exercised to stir the solution thoroughly so as not to have, at any time, an excess of acid pres- ent in any part of the solution. The acid reaction of the solution is not permanent because of the slight solubility of the barium carbonate. The titration is therefore conducted quite rapidly and a momentary disappearance of the red color is taken as the end-point. The solution, after the addition of the barium chloride, may also be diluted to a definite volume, and after allowing the pre- cipitate to settle, a measured portion of the clear liquid may be withdrawn and titrated. The amount of acid necessary to neu- TITRATION OF THE ALKALIES. 267 tralize the carbonate present is found by subtracting the amount of acid necessary to neutralize the sodium hydroxide from the amount used in titrating the total alkali. 367. Determination of Bicarbonates. — If a mixture of sodium carbonate and bicarbonate is to be analyzed, the total alkali may be found by titration with acid, using methyl orange as the indi- cator. As the precipitation of the carbon dioxide by barium chloride would tend to produce free acid according to the equation 2NaHC03 +2BaCl3 =2BaC03 +2NaCl +2HC1, thus preventing the completion of the reaction, a measured excess of standard alkali solution must first be added. After the precipitation of the carbon dioxide by the addition of barium chloride the alkali present is titrated with standard acid. The difference between the amount of alkali added and that found after the precipitation of the carbon dioxide is that neces- sary to convert the bicarbonate present into carbonate or to neu- tralize the equivalent amount of hydrochloric acid liberated from the barium chloride. The standard alkali used in this case may be either ammonia or caustic soda. The former is disadvanta- geous because of its disturbing effect on phenolphthalein, while it is difficult to obtain either alkaline solution free from carbon dioxide. Perhaps the simplest way of overcoming this difficulty is to conduct a blank determination by precipitating the carbonate from a measured volume of the alkaline solution and titrating as in the actual determination. EXERCISE 49. Determination of Sodium Hydroxide, Sodium Carbonate, and Water in Caustic Soda. Weigh out in a weighing-bottle from 3 to 5 grams of caustic soda. Do not take the sample from the top of the bottle, where it has been exposed to the moisture and carbon dioxide of the air. Transfer to the weighing- bottle with as little exposure as possible and stopper quickly. When weighed, transfer to a beaker, rinse the bottle mth water, and dissolve the soda. Transfer to a 250-c.c. flask and dilute to the mark with boiled water. 368. I'irst Method. Titration of the Carbonic Acid. — Take out 25 c.c. with a pipette and dilute lo about J 00 c.c. with boiled water which has been thoroughly cooled. Titrate, while stirring vigorously, with N /5 acid after add- ing 2 or 3 drops of phenoliDlithalein until the pink color has just disappeared. 268 VOLUMETRIC METHODS. Add 1 or 2 drops of methyl orange and continue the titration until a faint pink appears. Repeat until duplicates are obtained. Twice the amount of acid used in the methyl-orange tritation is the amount necessary to neutralize the sodium carbonate. 1 c.c. of N/5 acid equals .01061 gram NajCOj. The difference between the amounts of acid used in the phenol- phthalein and the methyl orange titrations is the amount necessary to neu- tralize the sodium hydroxide. 1 c.c. of N/5 acid equals .008012 gram NaOH. Calculate the amount of NajCOj and of NaOH present in 250 c.c. The difference between the sum of these weights and the weight of the caustic soda gives the weight of water present. Calculate the per cent of each constituent. 369. Second Method. Precipitation of the Carbonic Acid. — Measure out 25 c.c. of the soda solution with a pipette into a beaker, add 1 or 2 drops of methyl orange, and titrate to faint pink with N/5 acid. Measure out 50 c.c. into a 100-c.c. flask and add with shaking a little dilute solution of barium chloride drop by drop as long as a white precipi- tate forms. Dilute to the mark with boiled and cooled water, shake thor- oughly, and allow to settle. Measure out 50 c.c. and titrate with the N/5 acid, using phenolphthalein as the indicator. Repeat both titrations until duplicates are obtained. The second titration gives the amount of acid necessary to neutrahze the caustic soda, while the difference between the first and second gives the amount of acid necessary to neutralize the sodium carbonate present in 50 c.c. of the solution. Calculate the percentage of NaOH, NajCOj, and HjO as directed under the first method. DETERMINATION OF HARDNESS OF WATER. 370. Temporary Hardness or Alkalinity. — The hardness of water is caused by the presence of salts of calcium and magnesium in solution. If these metals are in solution as bicarbonates, they may be precipitated as carbonates by boiling the water, the excess of the carbon dioxide being expelled. The hardness of water which may be removed in this manner is called temporary hard- ness. The amount of calcium and magnesium held in solution in this manner may be ascertained by titrating the solution with a standard acid, using methyl orange or erythrosene and chloroform as the indicator. The result is given in parts of calcium carbon- ate per 100,000, giving the so-called degrees of hardness. As the bicarbonates of calcium and magnesium react alkaline to methyl orange and some other indicators, this titration also gives the alkalinity of the water. If sodium or potassium carbonates are present, they will also react alkaline, but a correction for this error will be obtained in determining permanent hardness.* * The method of determining hardness recommended in the Standard Method of Water Analysis, compiled by the Am. Pub. Health Ass. is given on p. 422. DETERMINATION OF HARDNESS OF WATER. 269 371. Permanent Hardness or Incrustants. — The presence of chlorides and sulphates of calcium and magnesium produce per- manent hardness, so called because these salts cannot be removed by boiling the water. On adding excess of sodium carbonate and evaporating to dryness, the sulphates and chlorides of calcium and magnesium are converted into the carbonates of these two metals and the sulphate and chloride of sodium. The bicarbonates of calcium and magnesium are also decomposed. On filtering and washing, the carbonates of calcium and magnesium, being insol- uble, remain on the paper, while the excess of the sodium car- bonate as well as the sodium chloride and sulphate pass into solu- tion. The excess of sodium carbonate having been ascertained by titration, the amount needed to decompose the chlorides and sul- phates of calcium and magnesium is found by difference. If the so-called soda reagent is used, the magnesium will be precipitated as hydroxide. The amount of permanent hardness is also calcu- lated as parts of calcium carbonate per 100,000 or per million. As the principal salt producing permanent hardness is calcium sulphate, which tends to produce a very hard scale in boilers, the term incrustants is also used to designate these constituents of the water. If the alkalinity of the filtrate is greater than that due to the sodium carbonate added, alkali carbonates must have been present in the water. Permanent hardness is then absent, and the amount of alkali carbonates present computed as calcium car- bonate must be subtracted from the temporary hardness already found.* EXERCISE so. Determination Oi Hardness of Water, Temporary and Permanent. Make an N/50 solution of hydrochloric acid b}'- leasuring out 100 c.c. oi the N '5 acid into a liter flask and diluting to the mark. Make an N/50 solution of sodium carbonate by diluting 100 c.c. of the N/5 alkali to a liter, or weigh out 1.061 grams of the pure dry salt, dissolve in water, and make up to a liter. 372. Temporary Hardness or Alkalinity. — Measure out 100 c.c. of the water to be tested into a porcelain dish and add one drop of methyl orange. Measure out 100 c.c. of distilled water into a similar porcelain dish and add one drop of the indicator. Add the N/50 acid drop by drop until a faint pink color is obtained. Titrate the water in the other dish with the acid until the color matches that in the dish conti-xning the distilled wa..er. From the amount of acid used, less the amount added to the distilled water (.1 to * See method given in Chapter XXIX on Water Analysis. 270 VOLUMETRIC METHODS. .3 c.c), the alkalinity or hardness is calculated in parts per 100,000 or parts per million. Erythrosene and chloroform may also be used as the indicator in this titration. The end-point with this indicator is sharper and its use is espe- cially to be advised when alum is likely to be present in the water. 0.1 gram of the erythrosene is dissolved in 1000 c.c. of water. 100 c.c. of the water to be tested is placed in a 250-c.c. glass-stoppered bottle; 2.5 c.c. of the indicator * and 5 c.c. of chloroform are added. The bottle is shaken thoroughly, so that the chloroform forms an emulsion. The standard acid is added and the shaking continued until the color is discharged. The alkalinity is calculated from the amount of acid used. 373. Permanent Hardness or Incrustants. — 100 c.c. of the water is meas- ured out into a porcelain evaporating-dish. Add a measured amount of the N/50 sodium carbonate solution sufficient to make the water strongly alkaline. Evaporate to dryness, take up with a small amount of water, filter, and wash the dish and filter-paper until free from alkali. Titrate the excess of the alkali with N/50 acid, using methyl orange as the indicator- The difference between the amount of alkali added and that titrated back is the amount used iu precipitating the calcium and magnesium, which produced the permanent hardness. The calculation is made in the same manner as that of temporary hardness. If the amount of alkali titrated back should be greater than the amount added, the water contains alkali carbonates which have been reckoned as temporary hardness. The amount of excess should therefore be deducted' from the temporary hardness. Per- manent hardness is absent. The so-called soda reagent gives somewhat more accurate results because the salts of calcium and magnesium are rendered less soluble by this reagent. It is composed of equal parts of N/10 sodium carbonate and sodium hydrox- ide. It may be prepared by weighing out 2.653 grams of pure sodium carbonate or measuring out 250 c.c. of N/5 sodium carbonate solution and 250 c.c. of N/5 sodium hydroxide solution (page 282) , and diluting to one liter. 200 c.c. of the water to be tested is measured into a Jena Erlen- meyer flask. After boiling 10 minutes to expel free carbonic acid, 25 c.c. of the soda reagent is added. It is again boiled down to a volume of 100 c.c, cooled and rinsed into a 200-c.c. graduated flask and the volume made up to 200 c.c. with boiled distilled water and thoroughly mixed by shaking the flask. The solution is filtered through a dry paper, and after rejecting the first 50 c.c, 100 c.c. is titrated with the N/50 acid, using ery- thrcsine as the indicator. The calculation is made in the same manner as before except that the 25 c.c. of tenth, normal soda reagent must be considered equal to 125 c.c. of the N/50 acid. When the permanent hardness is small, better results are obtained by determining total hardness by means of the soda reagent, and getting per- *When erythrosene is referred to as the indicator, the .01% solution of the sodium salt of erythrosene is implied. STANDARD SULPHURIC ACID. 271 manent hardness by difference. For this purpose the exact amount of acid used in titrating the alkalinity is added to 100 c.c. of the water being tested, thus converting the temporary hardness to permanent hardness, which is determined by the method already described. EXERCISE 51. Standard N/5 Sulphuric Acid. 374. Specific Gravity. — To 100 c.c. distilled water add 40 grams con- centrated sulphuric acid. Cool to 15° C. and take the specific gravity with the Westphal balance. In setting up this balance, the screw A is adjusted so that the two points at C are exactly on a level when the thermometer D Fig. 45. is suspended in the air. By the thumb-screw B the height of the beam is regulated so that the thermometer does not touch the bottom of the small cyUnder. The -.veights are of four sizes. The largest size hung on the hook on the end of the beam gives units, while on any other space gives tenths, the next size gives hundredths, the next size thousandths, and the smallest ten-thousandths of specific gravity. The instrument in Fig. 45 reads 0.0885, the largest size weight being absent. The sulphuric acid is poured into the cvlinder and the weights placed on the various divisions until the two points at C are again exactly on a level. If the thermometer now reads 15°, the 272 VOLUMETRIC METHODS. specific gravity is given by the weights on the beam. The percentage of H2SO4, corresponding to the specific gravity found, is obtained from the table of specific gravity of sulphuric acid given on p. 533. The specific gravity times the per cent of HjSO, present gives the number of grams of HjSOj per cubic centimeter. The number of cubic centimeters of acid to be measured out to give 9.808 grams of HjSOj is now computed. The acid, still being at 15°, is transferred to a burette, measured out into a liter flask, diluted to the mark and thoroughly mixed. Another liter is prepared in the same manner and mixed with the first in a 2-liter bottle. Instead of measuring out the concentrated acid, it may be weighed out. For this purpose it need not be kept at 15°. The amount to be weighed out for one liter is found by dividing 9.808 by the per cent found in the table and multiplying the quotient by 100. If this solution has been care- fully prepared, it will be very near the calculated strength. STANDAEDIZATION. First Method. — Same as first method in Exercise 48. Second Method. — Same as second method in Exercise 48. 375. Third Method. Precipitation as Barium Sulphate. — Measure out by a burette 50 c.c. of the sulphuric acid into a beaker. Dilute to 400 c.c, heat to boiling, and add barium chloride solution drop by drop while stirring the solution vigorously until all of the sulphuric acid is precipitated. Digest until the solution is clear. Wash by decantation, dry, ignite, and weigh. Repeat the determination until duphcates are obtained agreeing within .2%. Compute the weight of H2SO4, multiply by 20, and divide 9.808 by the product. This gives the number of cubic centimeters of the acid to be diluted to 1000 c.c. The acid is diluted as directed in Exercise 48. DETERMINATION OF SPECIFIC GRAVITY. The strength of acids as well as of a great variety of solutions is commonly ascertained by a determination of the specific grav- ity. This method is especially advantageous when dealing with solutions of pure substances. 376. Specific Gravity may be defined as the ratio between the weight of a given volume of a substance and the weight of an equal volume of some standard substance. When dealing with liquids or solids, water is almost invariably used as the standard of com- parison. In terms of the metric system of weights and measures, specific gravity may be defined as the ratio of the weight of a cubic centimeter of a substance to the weight of a cubic centi- meter of water. As the weight of the latter is 1 gram, specific DETERMINATION OF SPECIFIC GRAVITY. 273 gravity may be defined as the weight in grams of 1 cubic centi- meter of a given substance. When using the metric system of weights and measures, specific gravity is therefore identical with density, which is defined as the weight of unit volume of a sub- stance. 377. Standard Temperatures. — ^With the English units of weights and measures, specific gravity and density are not identical. The identity is not exact even with the metric system, except when the water is used at its maximum density (4° C.) and is weighed in vacuo. Considerable confusion has arisen because in deter- mining specific gravity the weight of water in the air and at other temperatures than 4° C. has been assumed to be 1 gram per cubic centimeter. Even when the exact weight of the water has been ascertained, confusion may arise if the conditions used are not given. The temperature at which both the substance and the water are weighed should always be given. This is ordi- narily done by placing a fraction after the figure giving the specific gravity. The numerator of the fraction gives the temperature of the substance, while the denominator gives the temperature of the water. 1.563-V^, for instance, means that a given volume of the substance at 15° weighs 1.563 times as much as the same volume of water at 4° C. If these weights have been reduced to vacuo, we may also state that the density of the substance at 15° is 1.563; that is, the weight of 1 cubic centimeter is 1.563 grams. For substances whose specific gravity is small, the correction obtained by reducing the weight to vacuo is very small, being about one-tenth per cent. The error introduced by considering the weight of 1 c.c. of water at 15° to l;e 1 gram is .08%, while at 20° the error is .17%, and at 25° it is .29%. As both of these corrections are positive, the error introduced by neglecting both of them is the sum of the two. The simplest and most conve- nient method of determining specific gravity consists in weighing both the substance and the water in the air at ordinary temper- tures, such as 15° or 20°. When the specific-gravity tables have been calculated from determinations made in the same manner, no error is introduced unless the volume of solutions are calcu- lated from their Aveight or weights from volumes. Such calcula- tions require true densities. 274 VOLUMETRIC METHODS. 378. Pycnometers. — The most reliable method of determining specific gravity involves the use of some form of pycnometer. The simplest form of this instrument consists of a small bottle closed with a glass stopper containing a capillary opening through which the excess of liquid may escape when inserting the stopper after filling the bottle. It is first weighed empty and dry, again weighed after filling with water, and finally the weight of the bottle filled with the substance being tested is taken. The latter as well as the water must be brought to the standard temperature before filling the bottle. As most liquids transmit heat quite readily and have a fairly high coefficient of expansion, there is consid- erable difficulty in getting the pycnometer filled at the desired temperature. After being filled the bottle may be placed in a vessel of water which is maintained at the standard temperature. If allowed to remain sufficiently long in the water, the bottle and contents ultimately attain the same temperature as the water. The time necessary for this purpose varies with the viscosity and other properties of the liquid. To obviate uncertainty in this respect, pycnometers are made with a thermometer ground so as to serve as a stopper, the bulb being in the liquid while the stem projects out of the bottle. A capillary tube is fused into the side of the pycnometer and is provided with a hollow glass ground cap. The liquid is cooled down below the standard temperature before filling the bottle, which is then placed in water maintained at this temperature. As the liquid expands the excess escapes through the. capillary tube. When the thermometer in the pyc- nometer indicates the desired temperature, the excess of liquid on the end of the capillary tube is removed and the cap put on. The bottle is removed from the water and wiped clean and set aside until the contents have reached room temperature, when it is weighed. The weight of distilled water which just fills the bottle at the standard temperature is obtained in the same manner. If the true density is desired, the weight of the water is corrected for air displacement, as explained on page 11, and the volume in cubic centimeters calculated from the density of the water at the temperature used. The weight of an equal volume of water in grams at 4° C. in vacuo is numerically equal to the number of cubic centimeters found by the calculation given. This weight, DETERMIXATIOX OF SPECIFIC GRAVITY. 275 divided into the weight of the bottle full of the liquid being tested, gives the density or the specific gravity compared with water at 4° C. 379. The Westphal Balance. — The manipulation of this instru- ment is very simple, and the results are almost as accurate and reliable as those obtained with the pycnometer. It consists of a beam suspended on a laiife-edge and so adjusted that when a glass plummet is suspended from a hook at one end, the weight on the other end is just counterbalanced, as indicated by the two points at C, Fig. 45, which are then exactly on a level. If they are not, the balance may be adjusted by means of the thumb-screw A. As commonly made, the plummet weighs 15 grams and displaces 5 grams of water at the standard temperature. The arm from which the plummet is suspended is divided into 10 equal divi- sions, and a set of weights in the form of riders is provided, such that the largest is just equal to the weight of the water displaced by the plummet, generally 5 grams. If the plummet is suspended in water, it will lose 5 grams in weight, which will be exactly counterbalanced by the large weight when hung from the hook at the end of the beam. This indicates a specific gravity of one. When the plummet is suspended in other liquids, the large weight indicates tenths when placed on other divisions of the beam. Similarly, riders weighing 0.5, 0.05, and 0.005 gram indicate hun- dredths, thousandths, and ten thousandths of specific gravity when they are hung on the beam so as to counterbalance the buoyant force of the liquid being tested on the glass plummet. The temperature of the liquid is taken by means of a thermometer within the plummet. 380. Hydrometers, or specific-gravity spindles, are instruments the construction of which is based on the principle that the buoyant force of a liquid is directly proportional to its density or specific gravity. The hydrometer is essentially a long, closed, glass tube, more or less enlarged at the lower end and so weighted with shot or mercury that it will assume a perpendicular position in a liquid in which it does not entirely sink. As the weight of the instru- ment is constant, the depth to which the hydrometer sinks in a given liquid may be used as an index of the specific gravity of the liquid. A paper scale is inserted in the stem of the instru- 276 VOLUMETRIC METHODS. ment and fixed in such a position that the specific gravity may be read directly on the scale at the surface of the liquid. A thermometer is sometimes made a part of the instrument, so that the temperature at which the readings are taken may be noted. These instruments are very convenient and rapid in use, but are not well adapted for accurate work. Specific gravity is often designated by degrees of one of the several scales which have been devised and which are almost invariably made use of in the construction and graduation of hydrometers. The degrees Twaddell are the simplest and most scientific of these scales. On this scale a degree corresponds to .005 specific gravity above unity. 7° Twaddell is therefore equal to 1.035 sp. gr.; 30° is equal to 1.150, etc. The spindle is usually made quite large, with a very small stem, so that determinations of specific gravity may be made with very great accuracy. The entire scale of 150° is usually divided equally among 6 spindles. This instrument is very generally used in Great Britain. 381. The Baume Scale is generally used in other countries. This instrument was originally graduated as follows: The point to which the spindle sank in water at 17|° C. was marked 0, and the point to which it sank in a 10 per cent solution of common salt was marked 10°. This space was divided into 10 equal divi- sions, and the remainder of the stem marked off in spaces of the same length. This gave the scale for liquids heavier than water. For liquids lighter than water, the point to which the spindle sank in a solution of 1 part of salt in 9 parts of water was marked 0, and the point to which it sank in water was marked 10. One- tenth of this space was taken as a degree, and the remainder of the scale divided into degrees of this size. This method gave degrees of unequal value in terms of specific gravity. On account of the varying purity of the salt used, the degrees on different instruments varied greatly in value, so that 66° Baumd might indicate in the case of sulphuric acid a minimum strength of 89.5% H2SO4, or a maximum of 95% H2SO4. As a remedy for this confusion, several formulas have been proposed by which the degree Baume may be calculated from the specific gravity, or the latter calculated from the former. The so-called rational formula is d= ^-pQ , in which d= specific gravity and B° = the degree DETERMINATION OF SPECIFIC GRAVITY. 2ll 144 Baume. In Holland the formula 6, = —-. — =r-„ and in the United 144— B° 145 States the formula d = — — — =-„ has been adopted as the standard 145— B° Baum6 scale for liquids heavier than water, and for liquids lighter 140 than water, the formula d = T^ — ^ has been adopted as the standard American scale. Under these circumstances it is evident that the Baum^ scale is the worst which could have been adopted, and that degrees Baum6 is by no means a definite statement of specific gravity unless it is known what particular scale has been used. The formula giving the relation between specific gravity and degrees BaumS should be printed on the scale, and only the standard American scale should be used in the United States. PROBLEMS. Problem 5. How many cubic centimeters of sulphuric acid having a specific gravity of 1 .3567 at 15° must be diluted to a liter to make N/5 acid? How many grams of the same acid must be used for making a liter of N/5 acid? Problem 6. Four grams of caustic soda were weighed out, dissolved in water, and diluted to 500 c.c. Fifty c. o. of this solution required, when titrated with N/5 acid and phenolphthalein, 46.36 c.c. of acid, and when methyl orange was added, 0.94 c.c. was required to give the end-point with this indicator. Calculate the percentage of sodium hydroxide, sodium car- bonate, and water in the caustic soda. Problem 7. One gram of sodium phosphate was -dissolved in water and titrated with N/5 acid, using phenolphthalein as the indicator, 0.67 c.c. of the acid being used. Methyl orange was then introduced and acid again added imtU the end-point was reached, 13.92 c.c. of the N/5 acid being used in the second titration. Calculate the percentage of crystallized tri- sodium and disodium phosphate present. Problem 8. Ten grams of a mixture of sodium carbonate and sodium bicarbonate was dissolved in water and diluted to a liter. To 50 c.c. of this solution, 25 c.c. of N/5 sodium hydroxide was added, and then enough barium chloride to precipitate the carbonate. On. titrating with N/5 hydro- chloric acid, 15 c.c. was used. Another 50-c.c. portion required 40 c.c. when titrated with the N/5 acid and methyl orange. Calculate the per- centage of sodium carbonate and bicarbonate present. Problem 9. In titrating 100 c.c. of a sample of water, 5 c.c. of N/20 sodium hydroxide was required with phenolphthalein. Another 100-c.c. 278 VOLUMETRIC METHODS. portion required 15 c.c. of N/20 hydrochloric acid with methyl orange. Calculate (a) the amount of carbon dioxide present and the alkalinity in parts per million; (6) calculate the amount of calcium oxide necessary to soften 1000 gallons of the water; (c) calculate the amount of lime- water necessary to soften the same quantity of water. Problem lo. To 100 c.c. of the same water, 50 c.c. of N/50 sodium-car- bonate solution was added, and after evaporation, 35 c.c. of N/50 hydro- ehioric acid was required to titrate the excess. Calculate (a) the permanent hardness in parts per million ; (b) the amount ol' soaium carDonate neces- sary to soften 1000 gallons of the water. After adding the softening agent and allowing the water to settle, or filtering it, how could (a) the presence of an excess of lime be found by titration, (&) a deficiency of lime, (c) an excess or deficiency of the soda- ash? CHAPTER XXII. STANDARD ALKALIES. 382. Standard Alkaline Solutions not Permanent. — While standard solutions of hydrochloric and sulphuric acids may be kept for months and even years without change in strength, no suitable alkaline solution has been found which in practice may be preserved without deterioration. This arises mainly from two causes, the absorption of carbon dioxide from the air and the action of the alkaline solutions on the glass of the containing bottles. In the case of ammonia neither of these causes of deterioration is serious, but another difficulty is met with in the volatility of the base. Moreover, it is a weak base and cannot be used with phenolphthaiein. The strength of all basic solutions must there- fore be frequently checked by titration with standard acid. The presence of carbon dioxide may in many cases be disregarded by using methyl orange as the indicator in titrations. As this is not always possible on account of the nature of the acid, standard alkaline solutions must frequently be prepared and kept free from carbon dioxide by methods described later. 383. Caustic Soda has been foimd to be the alkali best suited for general work. It is a strong base, forms soluble salts with all acids, and may be prepared of any desired strength, while the pure commercial article is cheaper than caustic potash. A solution of the highest purity may be made by dissolving metallic sodium in pure water. This metal may now be readily obtained in pure condition. Caustic soda made from the metal may also be purchased. The pure commercial caustic soda almost invari- ably contains small amounts of chlorides, sulphates, silica, and alumina, besides considerable amounts of carbon dioxide and water. About 20% more than the calculated amount of NaOH should therefore be weighed out, and dissolved in the required 279 280 VOLUMETRIC METHODS. amount of water. After titrating this solution against standard acid it is diluted to the required strength- On account of the presence of carbon dioxide, methyl orange must be used as the indicator, or litmus accompanied by boiling to expel the carbon dioxide. 384. Removal of Carbon Dioxide. — The alkaline solution must be kept in a bottle closed with a rubber stopper or a glass stopper coated with paraffine or vaseline. If the solution is to be used for titrating organic acids with phenolphthalein as the indicator, the carbon dioxide must be removed by means of barium chloride or hydroxide, and the solution subsequently protected from the carbon dioxide of the air. This may be accomplished either by pouring a layer of kerosene over the solution or by passing the air which enters the bottle through a soda-lime tube. In either case the soda solution must be withdrawn by means of a siphon. The arrangement shown in Fig. 46 is very convenient. The caus- :©>Kis»o Fig. 46. tic-soda solution is made up in a bottle of suitable size and, after the addition of enough barium chloride or barium hydroxide solu- tion to precipitate all of the carbon dioxide, is allowed to stand until the precipitate has settled. The clear solution is then siphoned off into the bottle A, and if a layer of petroleum-oil is not STANDARD ALKALIES. 281 placed on top, the tube B is filled with isoda-lime and inserted into the stopper. The long arm of the siphon C is pushed through the second hole of the stopper so that a space of about one inch is left between the end and the bottom of the bottle. The alkali may then be withdrawn, leaving undisturbed any barium carbon- ate which settles to the bottom. The siphon is filled by apply- ing suction to the tube D, which is filled with soda-lime, or pressure may be applied at B. The long arm of the siphon may be adjusted so that the end is opposite the zero-mark on the burette, and any excess of alkali drawn over will flow back iiito the stock-bottle. 385. Alcoholic Caustic-potash Solution is frequently used foi the titration of organic acids with phenolphthalein as the indi- cator. Such a solution is of twofold advantage. Many of the organic acids are insoluble in water, but dissolve readily in alco- hol. Alcoholic solutions of caustic potash are free from carbon dioxide, because potassium carbonate is almost absolutely insolu- ble in alcohol. On dissolving the caustic potash in the alcohol, the potassium carbonate remains undissolved. After standing . twenty-four hours the clear solution may readily be siphoned off from the potassium carbonate, which adheres quite firmly to the bottle. The alcohol used must first be freed from aldehyde or organic material extracted from the oak-barrel, which turns yellow on the addition of alkali and interferes with the titration. For this purpose a stick of caustic potash is added, and after stand- ing for some time the alcohol is distilled off. The caustic-potash solution must be protected from the carbon dioxide of the air by the method suggested for caustic soda or some similar device. 386. Water Solutions of Barium Hydroxide have been largely used when the presence of carbon dioxide would be objectionable. Any carbon dioxide which enters the solution is immediately precipitated as the insoluble barium salt. A solution saturated with barium hydroxide at the ordinary temperature is about two^ fifths normal. After shaking the solution with excess of the com- mercial barium hydroxide until no more dissolves, the clear liquid should be siphoned off and diluted a little before standardizing so as to prevent the crystallizing out of the barium hydroxide during occasional periods of low temperature. This alkali gives a very sharp end-reaction with phenolphthalein. 2S2 VOLUMETRIC METHODS. 387. Ammonia. — On account of its volatility strong solutions of ammonia cannot be us3d. Even with tenth- or fifth-normal solutions the loss of alkali is considerable if the solution is kept in a half-filled bottle and is poured out into the burette. During this process the air above the liquid is largely replaced, and as the pressure of the ammonia vapor is proportional to the concentration of the ammonia in solution, the loss in any case is considerable. The action of the ammonia on the glass is com- paratively slight, especially if the same bottle is used continually. The amount of carbon dioxide absorbed is also inconsiderable. If therefore the ammonia solution is siphoned out of the bottle, a fifth- or tenth-normal solution will show only inappreciable changes in strength from day to day, and if kept in an old bottle may be used for considerable periods without change of titre. EXERCISE 52. Preparation of N/s Caustic Soda and Determination of Ammonia in Ammonium Chloride. Weigh out roughly about 10 grams of pure caustic soda, dissolve in water, and dilute to 1 liter. Place in a bottle provided with a rubber stopper. Standardize by titrating against standard hydrochloric or sulphuric acid, using methyl orange as the indicator. Use at least 25 c.c. of alkali for each titration, If the acid is exactly N/5 normal, divide the number of cubic centimeters of alkaU by the number of cubic centimeters of acid used; The quotient multiplied by 1000 gives the number of cubic centimeters of the alkali to be diluted to 1000 to make a N/5 solution. Verify the strength of the diluted solution by titrating against the standard acid. Weigh out 250 mg. of ammonium chloride and transfer to a 300-c.c. Jena beaker. Dissolve in 50 c.c. of water and add 50 c.c. of the caustic- soda solution. In a similar beaker place 50 c.c. of water and 50 c.c. of the caustic-soda solution. Place both beakers on the hot plate and warm, finally to boiling, until no more ammonia is evolved. Cool both beakers, and titrate the alkali remaining in each vfith the standard acid, using methyl orange as the indicator. The difference between the amounts of acid used to neutralize the alkah in the beakers is the amount necessary to neu- tralize the ammoniain the ammonium chloride. 1 c.c. of N/5 acid = .003407 gram of ammonia or .0107 gram of ammonium chloride. Compute the percentage of ammonia as well as of ammoniiim chloride in the sample analyzed. KJELDAHL METHOD FOR NITROGEN. 283 THE KJELDAHL METHOD OF DETERMINING NITROGEN. In 1883 Kjeldahl described a method for determining nitro- gen "in organic substances which, with the numerous modifications which have since been suggested, is now the method most largely used for the determination of nitrogen, not only in organic com- pounds of the greatest diversity of composition, but also in nitrates and other inorganic compounds. There are very few nitrogen compounds in which by one or the other modification the nitro- gen cannot be determined with a degree of accuracy and with a rapidity almost unequalled by any other quantitative method. 388. Oxidation of Organic Matter and Conversion of Nitrogen into Ammonium Sulphate. — By the original method the organic material was digested hot with concentrated sulphuric acid, by which the carbon was oxidized to carbon dioxide, while the nitrogen remained as ammonium sulphate. The oxidation was completed by the addition of powdered potassium permanganate. The concentrated acid was then diluted somewhat, and neutralized with caustic soda. The ammonia liberated was distilled off, and determined by means of standard sulphuric acid. In many cases the combustion of the carbon was very slow. Various modifica- tions have been introduced to overcome this difficulty. The addi- tion of mercury, mercuric oxide, or copper oxide or sulphate greatly shortens the time of combustion, probably by acting as carriers of oxygen, the metal being alternately oxidized by the sulphuric acid and reduced by the carbon. The addition of phosphorus pentoxide or potassium sulphate serves the same purpose by raising the temperature to which the solution may be heated. When mercury has been added to the sulphuric acid it must be re- moved by means of sodium sulphide before the ammonia is dis- tilled off, otherwise some of the ammonia combines with the mercuric oxide and cannot be expelled by distillation. 389. Reduction of Nitrates. — By the original Kjeldahl method the conversion of nitrates to ammonia was very uncertain. To overcome this difficulty, salicylic acid or phenol was dissolved in the sulphuric acid. This solution first converts the nitrate into nitro-phenol. This compound is readily reduced to amido-phenol. 284 VOLUMETRIC METHODS. which is completely hydrolyzed by caustic soda with liberation of ammonia. These reactions take place according to the follow- ing equations; * CeHs.OH+HNO, =CeH,.OH.NOj+HjO; C6H,.OH.N02 +3Hj =CeH,.0H.NH2 +2H,0; CeH^.OH.NHj +NaOH =C„H4.0H.0Na+ NH,. The reduction is greatly assisted by the addition of sodium thio sulphate or zinc. The addition of pure cane-sugar is also advan- tageous when substances of high nitrogen content are oxidized. 390. Kjeldahl Digestion-flasks. — The digestion of the nitrog- enous substance with sulphuric acid is carried out in pear- shaped, long-necked flasks of Bohemian or Jena glass which are made for this purpose. The flasks are made of various sizes, since some material foams considerably on being heated with the sulphuric acid, and therefore requires the use of a larger flask. Some workers also desire to obviate the transference of the sul- phuric-acid solution from the digestion-flask to a special distilla- tion-flask. If this method is pursued, a digestion-flask capable ^ Fig. 47. of holding about 500 c.c. must be used. If a separate distillation- flask is employed, the digestion-flask need not ordinarily have a capacity of more than 200 c.c. 391. Digestion. — During the digestion the flask is placed so as to be inclined at an angle of 45°, in order that there shall be no loss of material by spattering, nor any contamhiation by the falling * These reactions are in reality more complicated than represented, as several sulphuric acid molecules are combined with the benzene radicle. The reactions as given simply show the transformation of the nitric-acid radicle into ammonia. KJELDAHL METHOD FOR NITROGEN. 285 in of foreign material. As an additional precaution some workers close the mouth of the flask by means of a stopper made by drawing out a glass tube which is somewhat larger than the neck of the digestion-flask. The drawn-out end is sealed, and after cutting off about an inch of the glass tube the stopper is inserted with the drawn-out end in the neck of the flask. Racks for holding six or eight flasks in an inclined position are used where much work must be done. These racks are pro- vided with a Bunsen burner under each flask. If the material froths much, the flask is at first heated with a small flame and over a wire gauze. The heat is gradually increased until finally the flask is heated with the bare flame, until the acid is colorless or a light yellow. The digestion requires from one-half to several hours. As fumes of sulphuric acid and sulphur dioxide are copi- ously evolved, the digestion must be conducted under a hood. 392. Distillation. — ^After cooling, the acid is diluted somewhat and transferred to the distillation-flask, if a separate flask is used for this purpose. The distillation-flasks should have ^^ a capacity of 500 to 600 c.c. and should be made of >^ hard Bohemian or Jena glass. They may be round- bottomed or of the Erlenmeyer form. To prevent the carrying over of the caustic alkali in the form of spray, a trap of some kind must be interposed between the distillation-flask and the condenser. The simplest form consists of a glass bulb 4 to 5 cm. in diameter fiUed with glass wool. In another trap shown in Fig. 48, which is very much used, the upper exit tube extends into and is somewhat curved towards the side of the bulb. The rate of distillation is much increased by Yig 48 covering the bulb with thin asbestos paper, whicli may be made to adhere by simply moistening with water. It is then pressed with the hand around the bulb. The trap is connected to the distillation-flask by means of a rubber stopper and to the condenser-tube by means of a piece- of thick-walled rubber tubing. The tube of the condenser must be of block tin. Glass cannot be used, as it is dissolved by the steam and ammonia vapor, increasing the alkalinity of the distillate. The condenser-tube is curved downwards, where it is joined to ^ 286 VOLUMETRIC METHODS. the trap. A number of these block-tin tubes are frequently passed through the same condenser, which consists of a copper or iron tank of convenient size, as shown in Fig. 49. The DISTILLATE is received in an Erlenmeyer flask of 300- to 400-c.c. capacity, in which a measured amount of standard sul- phuric acid has been placed. If the condenser is supplied with a sufficient amount of cold water, it appears to be unnecessary for the end of the condenser-tube to dip into the standard acid. It is customary, however, to attach a large-sized glass tube to the exit of the tin condenser-tube by means of a piece of rubber tubing. \^W^ Fig. 49. The end of this glass tube should touch the surface of the acid. All of the ammonia is usually carried over when 150 c.c. of liquid has distilled over. Until experience is gained it is advisable to continue the distillation after the Erlenmeyer flask containing about 150 c.c. of distillate has been removed. The distillate is then allowed to flow into a small beaker containing a drop of the indicator. If the distillate is still alkaline, the contents of the beaker are poured into the Erlenmeyer flask. The bumping which so frequently accompanies the boiling of an alkaline solu- KJBLDAHL METHOD FOR NITROGEN. 287 tion may generally be prevented by the addition of a few grams of granulated zinc. Bumping is also very much reduced if caus- tic soda free from carbonate is used. A small piece of paraffine will prevent foaming of the liquid. 393- Titration of Ammonia. — The excess of acid in the distil- late is titrated back with standard alkali, which may be caustic soda, potash, or ammonia. The indicator used may be methyl orange, litmus, or cochineal. As the various reagents used may contain more or less nitrogen, a blank must be conducted when- ever one or more of the reagents is used for the first time. For the blank determination pure cane-sugar is digested, and care must be taken to add each of the reagents in the amount used in the actual analysis. The amount of nitrogen found in this deter- mination is deducted from each result subsequently obtained with the reagents tested. WILFARTH'S MODIFICATION APPLICABLE TO SUBSTANCES FREE FROM NITRATES. One gram of the finely powdered substance is placed in the digestion-flask, 1 gram of metallic mercury or about 0.7 gram of mercuric oxide is added, and 25 c.c. of a solution of 200 grams of phosphorus pentoxide in 1 liter of concentrated sulphuric acid (sp. gr. 1.84). T^ter digesting and transferring the acid to the distillation-flask with about 200 c.c. of water the mercury is pre- cipitated by the addition of 25 c.c. of a solution of 40 grams of potassium sulphide in 1 liter of water. The acid is then neutral- ized by the addition of a saturated solution of caustic soda, of which about 50 c.c. will be required. After the addition of a few grams of granulated zinc the ammonia is distilled into the sul- phuric acid, of which 50 c.c. of N/5 acid is usually ample. The excess of acid is titrated back with N/5 or N/10 alkali. GUNNING'S MODIFICATION APPLICABLE TO SUBSTANCES FREE FROM NITRATES. Gunning's improvement consisted in the addition of potassium sulphate to the sulphuric acid, by which the time of digestion is considerably reduced. Sodium sulphate may also be used as 288 VOLUMETRIC METHODS. well as sodium pyrophosphate. A very excellent digestion mix- ture consists of 20 c.c. of the mixture of phosphorus pentoxide and sulphuric acid, given in WUfarth's modification, to which 10 to 15 grams of potassium sulphate are added, either in the beginning or after the first violent oxidation is ended. Excellent results may be obtained with this mixture without the addition of mercury, though the oxidation is considerably hastened by the addition of the usual amoimt of the latter. The decomposition is frequently complete in thirty minutes and rarely requires more than an hour. The precipitation of the mercury, neutraliza- tion, distillation, etc., are carried out as given in the Wilfarth modification. GUNNING-JODLBAUER MODIFICATION APPLICABLE TO NITRATES. In this modification the digestion is carried out with sulphuric Acid which contains 40 grams of phenol or salicylic acid per liter. 1 gram of the nitrogenous substance is placed in the digestion- flask and dissolved in 30 c.c. of the phenol-sulphuric acid. The solution should be kept cold and thoroughly shaken to hasten the solution. 2 to 3 grams of zinc-dust and 1 gram of mercury are then added while the flask is shaken and cooled. After stand- ing one to two hours the flask is heated, at first gently, then more strongly. 10 to 15 grams of potassium sulphate may then be added, and the flask heated imtil the solution is colorless or nearly so. The remainder of the process is conducted as directed for the other modifications. FORSTER'S MODIFICATION APPLICABLE TO NITRATES. Having failed to obtain aU of the nitrogen from nitrates by the previous modification, Forster introduced the use of sodium thiosulphate as a reducing agent. He treated ^ gram of the nitrate in the digestion-flask with 15 c.c. of a 6% phenol-sulphuric acid, and after the nitrate was completely dissolved 1 to 2 grams of sodium thiosulphate were added. T\Tien the decomposition of the thiosulphate is complete 10 c.c. of concentrated sulpliuric KJELDAHL METHOD FOR NITROGEN. 289 acid and 1 gram of mercury are added. The decomposition is completed in the usual manner. The thiosulphate mxist not be added before the addition of the phenol-sulphuric acid, or nitrogen wiU be lost. The solution need not be cooled as in the Jodlbauer modification. The oxidation is generally complete in IJ hours. 394. Substances Decomposed with Difficulty. — In regard to the substances which cannot be determined by the Kjeldahl method and its modifications Dyer * states that the nitrogen in aromatic nitro-compounds caimot be determined by simple reduc- tion with zinc. Phenol or salicylic acid must be present. The same thing is true of many of the azo compoimds. Hydroxyl- amine and acetaldoxime required the addition of sugar before all of the nitrogen was converted into ammonia by the Jodlbauer method. The cyanide and ferrocyanide of potassiirai could be analyzed by the Gunning modification, while the ferricyanide required the Forster modification. Phenylhydrazine could not be correctly analyzed by any modification. H. C. Sherman f states that where large proportions of chlorides and nitrates exist together all modifications fail. EXERCISE S3. Detennination of Nitrogen in Milk by the Kjeldahl-Gunning Method. 395. Standard Solutions. — For this determination a fifth normal acid and alkah will be needed. The fifth nonnal sulphuric- or hydrochloric-acid solution already prepared may be used. The fifth-norrdal solution of caustic soda may also be used, or a fifth-normal solution of ammonia may be pre- pared as follows: About 15 c.c. of concentrated ammonia (sp. gr. 0.90) are diluted to 1 liter. A second hter is made in the same manner as the first. The entire solution is well mixed in a large bottle, and then titrated against the N/5 standard acid, using cochineal or methyl orange as the indicator. The solution of ammonia, which is somewhat stronger than fifth-normal, is then diluted to the exact strength, and placed in a bottle connected with the burette by means of a siphon, as described on p. 2S0. The soda-lime tube may be omitted. A simpler arrangement, which answers fully as well for this purpose, consists of a suitable-sized bottle, closed *Chem. Centr., Ill, XVII, 433. t J. Am. Chem. Soc. XVII. 567. 290 VOLUMETRIC METHODS. by a two-holed rubber stopper through which a siphon passes, the long arm of which is outside the bottle. To the end of the long arm is attached, by means of a short piece of rubber tubing, a piece of glass tubing. A pinch-cock is placed on the rubber tube. By opening this pinch-cock and forcing air through the second opening in the stopper, the siphon may be filled. The burette is filled by opening the pinch-cock and allowing the so- lution to flow out. 39b. Weighing the Sample. — The sample of milk to be analyzed must be thoroughly mixed by shak- ing the bottle. 10 c.c. is withdrawn with a pipette and allowed to flow over the asbestos contained in a Hofmeister capsule. The asbestos must be first ignited, and after placing in the capsule, the whole is weighed. After the milk has been added the whole is again weighed. Neither of these weights need be taken closer than several milli- grams. A second or even a third sample of milk is weighed out in the same manner. The milk Fig. 50. jg evaporated to dryness on the water-bath, the capsule crushed in a clean porcelain mortar and transferred to the digestion-flask. The milk may also be evaporated in the flasks, the weight then being obtained from the specific gravity, or the average weight of 10 c.c. may be obtained by first weighing several portions in small beakers. The evapora- tion in the flasks is slow unless the steam is displaced by a stream of air. The water may also be boiled out after the addition of the digestion- mixture. 397. Digestion. — Weigh out 20 grams of phosphorous pentoxide in a beaker, and add 500 c.c. of concentrated C. P. sulphuric acid. Stir until solution is complete,- and transfer to a glass-stoppered bottle. Transfer 20 c.c. portions of this solution to the digestion-fiaslis containing the milk, and also to two other flasks in each of which about a gram of granulated sugar has been placed. Add 10 grams of potassium sulphate to each flask, and heat gently over the wire gauze with the Bunsen burner. The heating is conducted under the hood. Increase the heat as the frothing ceases, finally remove the -wire gauze, and heat strongly until the solution is color- less or only a light yellow. While the solution is digesting a saturated solution of caustic soda is made by dissolving 200 grams in twice this amount of water. 40 or 50 c.c. of the standard acid is measured out into each of the Erlenmeyer flasks to be used as receivers. If the condensers of the distilling apparatus have not been recently used, they arc cleaned by passing through them the steam from a flask of water strongly acidified with sulphuric acid imtil KJELDAHL METHOD FOR NITROGEN. 291 the distillate is no longer alkaline to cochineal solution. If 100 c.c. of the laboratory supply of distilled water reacts alkaline -mth. the cochineal solution, ammonia-free water should be made by condensing the steam from a flask of water acidified with sulphuric acid. This steam should be condensed by passing it through one of the block-tin tubes of the Kjeldahl condenser. This distilled water should then be used for diluting the con- centrated acid and whenever needed during the titrations. 398. Distillation. — When the digestion-flasks have cooled, about 60 c.c. of water are added to each, and if the small-sized digestion-flasks have been employed, the solutions are transferred to the distillation-flasks, using about 150 c.c. of the ammonia-free water to rinse out the digestion-flasks. The acid is neutralized wth the concentrated caustic-soda solution, of which about 50 c.c. will be required. 2 or 3 grams of granulated zinc are added and the flask is immediately connected with the condenser to prevent loss of ammonia from the hot alkaline solution. The distillation of the ammonia in the first flask is begun immediately, and the acid in the other flaslcs is diluted, transferred, and neutralized in rotation. 399. Titration. — When about 150 c.c. of distillate has passed into the first Erlenmeyer flask, it is replaced by a small beaker containing a little of the indicator. If 10 or 15 c.c. of the distillate does not react alkaline with the indicator, 2 or 3 drops of the cochineal are added to the solution in the Erlenmeyer flask and the excess of acid titrated with the standard alkaline solution. If the latter has not been compared within a day or two ■with the standard acid, a measured volume must be titrated against the acid. The other Erlenmeyer flasks should be removed and the solu- tions titrated in the order in which the acid solutions were transferred to the distillation-flasks. After deducting the average amount of acid neu- tralized by the ammonia obtained in the blank determinations, the amount of nitrogen found is computed by the factor 1 c.c. of N/5 acid = .002802 gram of nitrogen. To obtain the percentage of casein, the percentage of nitrogen is multiplied by the commonly used factor 6.25 or by the more recently obtained and probably more correct factor 6.43. EXERCISE 54. Determination of Nitrogen in Potassium Nitrate by the Fbrster Modification of the Kjeldahl Method. For this determination it is advisable to use a sample of pure potas- sium nitrate as a check on the process. A sample of commercial sodium or potassium nitrate may then be analyzed. Twenty grams of phenol are dissolved in 340 grams of concentrated sulphuric acid. A solution of potassium sulphide is made by dissolving 20 grams cf the commercial article in 500 c.c. of water, ^-gram portions of the nitrate are weighed out and transferred to dissection-flasks and dis- 292 VOLUMETRIC METHODS. solved in 15 c.c. of the phenol-sulphuric-acid solution. After the nitrate is completely dissolved, 1 to 2 grams of sodium thiosulphate are added. When the reaction due to the latter is ended, 10 c.c. of concentrated sul- phuric acid is added and about 1 gram of metallic mercury. The mercury may be dropped from a glass tube the end of which is partially closed by fusion or drawing out. The size of this opening may readily be so adjusted that one drop shall weigh about 1 gram. A blank determination is made by treating about 1 gram of sugar with the reagents as already given. The digestion and the subsequent operations are carried out exactly as given in the preceding exercise, except that after the acid solution has been trans- ferred to the distillation-flask, 25 c.c. of the potassium -sulphide solution are added. 1 c.c. of N/5 acid is equal to one five-thousandth of the molecular weight of potassium nitrate or the same proportion of the molecular weight of sodium nitrate. CHAPTER XXIII. TITRATION OF BORIC AND CARBONIC ACIDS. ESTIMATION OF BORIC ACID. The titration of boric acid cannot be carried out by means oi any indicator in the ordinary manner of titrating an acid, on account of the very weak acid properties of this substance. Because of this peculiarity, the base in many of the salts of boric acid may be readUy titrated by means of a strong mineral acid and methyl orange as the indicator without the removal of the boric acid, which is entirely neutral towards methyl orange. Because of the difficulty of carrying out the gravimetric methods which have been devised, the discovery of a method of increasing the acid properties of this acid so that it can be titrated with phenol- phthalein as the indicator has been very welcome to chemists. 400. Thompson's Glycerine Method. — The method proposed by R. T. Thompson * consists in the addition of enough glycerine to constitute 30% of the solution when the titration is complete. In such a solution the boric acid acts as a monobasic acid. To secure a solution which shall contain no free alkali nor acid except boric acid, the borate is dissolved in water and the solution neu- tralized with sulphuric acid after the addition of a drop of methyl orange. A drop or two of acid is then added and the solution boiled in an open dish for a few minutes to expel carbon dioxide. A dilute solution may be boiled in this manner for fifteen minutes without the loss of more than a faint trace of boric acid.f Dilute caustic-soda solution is now added until the pink color of the methyl orange just disappears. Glycerine is then added in such amount that after completion of the titration at least 30% shall be * Jour. Soc. Chem. Ind., XII, 432. t L. de Koningh, Jour. Am. Chem. Soc, 1897, 385. 293 294 VOL UME TRIG ME THODS. present. After the addition of a drop of phenolphthalein the solu- tion is titrated with fifth-normal caustic soda which is free from carbon dioxide. The usual pink end-point of the phenolphthalein is obtained when a slight excess of the alkali has been added. More glycerine should then be added, and if the pink color of the phenolphthalein fades, more alkali must be added. A blank deter- mination should be carried out to determine the acidity of the glycerine. For this purpose the same amount of glycerine should be diluted with an equal volume of water and the solution tit- rated with the standard alkali. If ammonia is present, it must first be removed by adding a slight excess of sodium carbonate and evaporating the solution down to half its bulk. By this treatment any iron present will be precipitated and should be filtered off. The neutralization of the boric acid with caustic soda takes place according to the fol- lowing equation: HBO, +NaOH =NaB02 +11,0. 1 c.c. of normal caustic alkali is therefore equal to .035 gram of B2O3. 401. Jones's Mannitol Method. — Another method of titrating boric acid has been based on the fact that in the presence of manni- tol the acidity is sufficiently increased to allow of its titration with phenolphthalein as the indicator.* The solution may be neu- tralized in the same manner as given in the glycerine method. Phenolphthalein is then added and standard alkali introduced until the solution is red. A pinch of mannitol is then thrown in, which bleaches the red color. Enough alkali is added to produce a slight alkaline reaction, followed by mannitol until a permanent red color is produced which cannot be discharged by further addi- tion of mannitol. 1 or 2 grams of mannitol is usually sufficient. The author of this method, instead of using methyl orange as the indicator to secure a solution free from acids except boric, utilizes the fact that a mixture of potassium iodide and iodate is entirely unaffected by free boric acid, while strong mineral acids liberate iodine, thus destroying the free njineral acid according to the following equation : 6HC1 +5KI +KIO3 =6KC1 -fSI^ +3H2O. * L, C. Jones, Am. J. Sci., 1S93, pp. 147-153. TITRATION OF BORIC ACID. 295 The free iodine is removed by adding sodium thiosulphate solution, which combines with the iodine, producing only neutral compounds as follows: 2Na2S203 +I2 =2NaT + Na2S406. The amount of free mineral acid present may be estimated by de- termining the amount of free iodine, as will be shown in the next chapter. The solution of the borate, the volume of which should be about 50 c.c. and which should contain about 0.1 gram of boric acid, is made slightly acid with hydrochloric acid. To precipitate carbonic acid 5 c.c. of a 10% solution of barium chloride is added. A mixture of a solution of potassium iodide and iodate is made in a separate beaker. About 2 grams of the iodide and ^ gram of the iodate are generally sufficient to destroy all the mineral acid present, unless the amount is excessive. Starch solution is added to the iodate solution, and if a blue color develops, it is re- moved by the cautious addition of dilute sodium thiosulphate solu- tion. A drop of the borate solution added to the iodate should produce a blue color, indicating the presence of free hydrochloric acid, and the certainty of the presence of all boric acid in the free state. The two solutions are then mixed, and the iodine liberated is removed by the addition of sodium thiosulphate. The colorless solution now contains only starch, neutral chlorides, potassium tetrathionate, iodide, and iodate, and the total amount of the boric acid to be determined in the free condition. Phenolphthalein is now added and the boric acid titrated, the mannitol being added, as already directed. The end-reaction by this method is sharper than that in Thompson's method. 402. Decomposition of Organic Matter containing Boric Acid. — Boric acid must frequently be determined m food products, and the separation of the organic matter is first necessary. Thompson proceeds as follows: 100 grams of the sample is made strongly alkaline with lime-water, evaporated to dryness in a platinum dish, and cautiously heated until the organic matter is thoroughly charred so as to give a colorless solution. 20 c.c. of water is added, and the solution acidified with hydrochloric acid, when all but carbon will be in solution. The whole is transferred to a 100-c.c. flask and 0.5 gram calcium chloride added. A few drops of 296 VOLUMETRIC METHODS. phenolphthalein are introduced, then caustic soda until a slight pink color is produced, and finally 25 c.c. of lime-water. By this treatment the carbon dioxide as well as other weak acids, such as phosphoric acid, which may be present are precipitated as calcium salts. This solution is made up to 100 c.c, thoroughly shaken, and filtered through a dry paper. To 50 c.c. of the filtrate normal sulphuric acid is added imtil the pink color disappears, then a drop or two of methyl orange is introduced and acid is again added until the yellow color is just changed to pink. Fifth- normal caustic soda is then added drop by drop until the Uquid assumes the yellow tinge, excess of soda being carefully avoided. The solution is boiled until the carbon dioxide is entirely excelled. It is then cooled and the boric acid titrated, using phenolphthalein and glycerine or mannitol as the indicator. DETERMINATION OF CARBONIC ACID IN WATER. The determination of the percentage of carbonates and bicar- bonates when existing as alkali salts, and in the case of the former in the presence of caustic alkali, has already been given. The determination of carbonic acid when present as an insoluble car- bonate is best carried out gravimetrically, as given in Chapter IX. Carbon dioxide existing in solution, as in natural or artificial mineral waters, is generally determined volumetrically. It exists in three cojiditions in such waters: as free carbonic acid; as semicombined acid, or bicarbonates of the alkali or aJkaline-earth metals; and as combined acid, or carbonates of the alkali metals. 403. The Percentage of Free Carbon Dioxide may be deter- mined by titrating 100 c.c. of the water with a standard solution of caustic soda or sodium carbonate after the addition of a few drops of phenolphthalein. The alkali should be dilute, tenth or twentieth normal being convenient. When sufficient alkali has been added to give a faint pink to the indicator, all the carbon diojdde has been converted into bicarbonate and a slight excess of carbonate is present, which produces the color with the indicator. The titration must be repeated, adding at once nearly all the alkali used in the first titration as some carbon dioxide may have been lost during the stirring of the acid solution. TITRATION OF CARBONIC AClD. 297 If free carbon dioxide is present, normal carbonates must be absent, since sucli salts would react with the free acid to form bicarbonates as follows: Na2C03+ H2CO3 =2NaHC03. Only bicarbonates, therefore, can exist in a solution containing free carbonic acid. On the addition of caustic soda the following reaction takes place: NaOH+H,C03=NaHC03+H20. ~ 40.058 parts of caustic soda are therefore equivalent to 44 parts of carbon dioxide. 1 c.c. of N/10 caustic soda will therefore be equal to .004400 gram of CO^. With sodium carbonate solution the following reaction takes place : Na^COg +H2CO3 =2NaHC03. 106.1 parts of sodium carbonate are therefore equal to 44 parts of carbon dioxide. 1 c.c. of N/10 sodium carbonate solution will therefore be equal to .0022 gram of COj. The use of an alkaline solution of which 1 c.c. is equal to 1 mg. of carbon dioxide is very convenient. 2.412 g. of sodium carbonate are equal to 1 g. of carbon dioxide f = 2.412 j. Probably the most con- venient method of procedure is to dissolve 2.412 g. of pure anhydrous sodium carbonate in 1 liter of distilled water from which the carbon dioxide has been expelled by boiling. 1 c.c. of this solution will be equal to 1 mg. of carbon dioxide. If the water is neutral to phenolphthalein only bicarbonates can be present. The water may be tested for the presence of FREE MINERAL ACIDS by titration with the standard solution of alkali, using methyl orange as the indicator. The difference between the amounts of alkali used in this titration and the one in which phenolphthalein is used as the indicator gives the amount of alkali necessary to neutralize the free carbonic acid. In the presence of free mineral acid semi-combined or combined carbonic acid would be absent. 404. The Amount of Semi-combined Carbon Dioxide may be determined by titrating 100 c.c. of the water with dilute standard 298 VOLUMETRIC METHODS. acid, using methyl orange as the indicator. The titration with acid and methyl orange gives the amount of semi-combined carbon di- oxide onl}' when the water is acid or neutral to phenolphthalein, as normal carbonates are then absent. If the water is alkaline to phenolphthalein free carbon dioxide is absent and only carbon- ates and bicarbonates can be present. By titrating 100 c.c. of such a water with phenolphthalein and dilute (N/20) acid until the pink color is removed and then adding methyl orange and continuing the addition of the acid until the pink color of the methyl orange is developed, the amount of combined and semi-combined acid may be calculated. The combin ;d carbon dioxide is calculated from the number of cubic centimeters of acid used in the first stage* of the titration, 1 c.c. of N/20 acid being equal to .0022 gram of carbon dioxidu (~kk~), thv following reaction taking place;' HCl +Na2C03 =NaHC03 +NaCl. The number of cubic centimeters of acid used in the second stage of the titration minus the amount used in the first stage gives the amount of semi-combined carbon dioxide, 1 c.c. of N/20 acid being equal, as before, to .0022 gram of carbon dioxide. 405. The Total Amoirnt of Carbon Dioxide may be obtained by adding together the amount of free carbon dioxide to that semi- combined; or whan free carbon dioxide is absent, by adding the amount of semi-combined to the amount combined. The amount of free and semi-combined carbon dioxide is frequently deter- mined by Pettenkofbe's method. In this process a measured amount of a standard solution of calcium or barium hydroxide is added to a measured volume, generally 100 c.c, of the water to be -tested. The free and semi-combined carbon dioxide is precipi- tated as barium or calcium carbonate with consequent decrease of alkalinity of the solution. The excess of alkali is titrated back with dilute acid. The total amount of carbon dioxide is calculated from the excess of barium hydroxide plus the initial alkalinity of the water : Ba(0H)2+ H2CO3 = BaCOg + 211 fi ; Ba(0H)2+ CaCHCOa)^ = BaC03+ CaC03+2H20. 406. If Sulphates or Carbonates of the Alkalies are present, the precipitated barium salts carry down some of the alkalies unless TITRATION OF CARBONIC ACID. 299 the barium was piesent as chloride, so as to leave the alkali metals as chlorides. For this reason a little neutral barium chloride solution should be added. 407. Determination of Total Amount of Carbon Dioxide in the Presence of Magnesium. — If magnesium bicarbonate is present, an error is introduced into the determination as shown by the following equations: Mg(HC03)2 +Ba(0H)2 =BaC03 +MgC03 +21Ifi; MgC03+Ba(OH),=BaC03+Mg(OH)2. • As the magnesium hydroxide is insoluble the alkalinity of the solution has been decreased by 2 molecules of barium hydroxide, whereas with other bicarbonates but 1 molecule is neutralized. To remedy this difficulty Pettenkofer added 2 c.c. of a saturated solution of ammonium chloride so that the magnesium hydroxide should remain in solution. According to Trillich* this is oidy a partial remedy of the difficulty. He proposes to omit the addi- tion of the anunonium chloride and apply a correction, to the result from a separate determination of magnesium.. The DETERMINATION WAS CARRIED OUT according to Petten- kofer as follows: 100 c.c. of the water are placed in a flask and to it are added 3 c.c. of a 10% solution of calcium or barium chloride and 2 c.c. of a saturated solution of ammonium chloride, as well as 45 c.c. of a solution of barium hydroxide of such a strength that 1 c.c. =1 mg. of carbon dioxide. The flask is then corked and allowed to stand for twelve hours. 50 c.c. of the clear liquid are withdrawn and the excess of alkali titrated with standard acid. Trillich omits the addition of ammonium chloride, so that the magnesium hydroxide is entirely precipitated. The amount of magnesium hydroxide is calculated from a gravimetric deter- mination, and its equivalent in barium hydroxide subtracted from the amount of barium hydroxide neutralized by the water. 408. Precipitation of Carbon Dioxide as Calcium Carbonate. — When water contains a considerable amornit of carbon dioxide so that there is danger of loss during the manipulation, a sample should be collected in a measuring-flask which contains a measured volume of an alkaline solution of calcium chloride made as follows: ' * Zeit. f. angew. Chem., 1889, pp. 117, 337, 300 VOLUMETRIC METHODS. To one part of crystallized calcium chloride 5 parts of water are added and 10 parts of ammonium hydroxide sp. gr. 0.96. This solution is placed in well-stoppered bottles and allowed to stand several days. The clear supernatant liquid is drawn off as needed. On the addition of this reagent to carbonated water all of the carbonic acid present is precipitated as calcium carbonate. The precipitate is at first amorphous, but on standing for at least twelve hours it becomes crystalline. This change may be hastened by warming on . the water-bath. The precipitate must not be filtered off imtil it has been digested in this manner or stood at least 12 hours. It is then filtered off and washed free from alkali. The carbon dioxide may now be determined in the calcium car- bonate by various methods. A very simple method consists in dissolving the carbonate in excess of standard acid and titrating back the excess with standard alkali. 409. If Bottled Carbonated Waters are to be Examined, the bottles may be opened so as not to lose carbon dioxide by a very simple device. An ordinary brass cork-borer of suitable size is taken, and several small holes are drilled a little farther from the cutting end than the length of the cork to be pierced. A rubber tube is securely fastened to the upper end of the borer, and con- nected to a convenient absorption apparatus containing some of the alkaline calcium chloride solution. The cork of the bottle is pierced with the cork-borer by holding the latter firmly in the hand and turning the bottle around. When one of the small holes in the borer has been brought below the cork the carbon dioxide wUl pass out through the cork-borer and into the absorp- tion apparatus. When no more carbon dioxide is evolved the bottle is warmed by placing it in lukewarm water, which is then brought nearly to a boil. The absorption apparatus is then dis- connected, and the carbon dioxide swept out of the connecting tube by means of air freed from carbon dioxide. Th^j contents of the bottle are added to the liquid in the absorption apparatus, the solution digested on the water-bath for some time, and finally the calcium carbonate is filtered off and washed. The calciimi carbonate may then be titrated with standard acid and alkali. 410. Determination of Carbon Dioxide in Gases. — ^Tien car- bon dioxide occurs in considerable amoimt in mixtures of gases TITRATION OF CARBONIC ACID. 301 it is determined by gasometric methods. AVhen it occurs in amounts considerably less than 1% it is customary to determine it gravimetrically by passing the gas through an appropriate series of absorption-tubes so that the carbon dioxide may finally be weighed. More frequently, however, the carbon dioxide is determined volumetricaUy. 411. Pettenkofer's Method. — ^Many methods of procedure and forms of apparatus have been proposed, but nearly all of them follow in general the method suggested by Pettenkofer, which consists in absorbhig the carbon dioxide in a measured volume of standard alkali and titrating back the excess. Pettenkofer used barium hydroxide and oxalic acid as the standard solutions. He took a large-sized bottle, aspirated the air to be examined through it until the air already present was completely displaced, and added a measured amount of the standard alkali. On shaking the bottle vigorously for some time the carbon dioxide was com- pletely absorbed. The alkali was poured into a cylinder or bottle, the precipitate allowed to settle, a measured volume with- drawn, and the excess of alkali titrated with the oxalic acid, using phenolphthalein as the indicator. 412. Several Difficulties and Errors are met with in this mathod. The smrface of the glass bottle is capable of holding barium hydroxide so that more alkali is apparently removed from the solution than corresponds to the carbon dioxide present. To overcome this difficulty various remedies have been proposed. The bottle may be rinsed with some of the barium hydroxide solution, washed free from alkali with distilled water, and then dried. A better method much used consists in coating the inside of the bottle with a thin layer of paraffine. This method is suc- cessful where a coat of paraffine can be obtained which does not peel off. Still another method consists in shaking the bottle with moist freshly-precipitated baritun carbonate. A thin coating of this salt adheres quite firmly to the glass, and acts quite efficiently as a protective agent. 413. Apparatus. — ^Another source of error is found in the contamination of the barium hydroxide solution with carbon dioxide while being poured from the large bottle into a suitable vessel for allowing the precipitate to settle. In this case the 302 VOLUMETRIC METHODS. most serious contaminating influence is the breath of the operator. A form of apparatus recently described by Sewaschen * seems well adapted to overcome this difficulty. The large bottle A has a capacity of about 6 liters. The exact volume is carefully deter- mined, most conveniently by the Morse-Blalock bulbs. The vol- ume is taken to a mark c on the neck, to which the rubber stopper B must always be inserted. The small bottle C should have a capacity of very nearly 100 c.c, the exact capacity having been a A 6OOOCC B*=f n Fig. 51. carefully determined when the stopper D is inserted to the mark d on the neck, the bottle being filled so that no air-bubbles are left under the stopper and the glass rod h inserted. The volume of the neck from the point d must also be determined and added to that of the large bottle. The neck of the small bottle must be free from any projecting ridge, and must fit quite snugly into the opening in the stopper B. If a suitable 100-c.c. bottle is not at hand, a stout 100-c.c. Florence flask may be used. The projecting rim on the neck may be quite readily cut off by means of a dia- mond, a small flame, or other glass-cutting device. 414. Manipulation.— To use this apparatus, the bottle A is cleaned and dried and the inner surface treated by one of the methods already given. The air to be examined is aspirated through the bottle after the stopper B is inserted. While the large bottle is being filled with air the standard solution of barium hydroxide is measured out into the small bottle, the stopper D * Hygjenische Rundschau, No. 9, 1897. TITRATION OF CARBONIC ACID. 30S inserted, and finally the rod b. The large bottle is turned on its side and the neck of the small bottle inserted into the stopper B after removing the stopper D. On turning the large bottle upright the barium hydroxide solution flows out of the small bottle and comes in contact with the air i^ the large bottle. The latter is shaken occasionally for about two hours, when the absorption of the carbon dioxide will be complete. By inverting the large bottle the barimn hydroxide solution will flow into the small bottle, which is then disconnected from the large bottle, the stopper D inserted, and the barium carbonate allowed to settle, which requires from twelve to twenty-four hours. Measured volumes are then withdrawn and titrated with the standard acid. 415. Barium Hydroxide has been used most largely as the alkaline solution in which to absorb the carbon dioxide. The barium hydroxide is seldom free from traces of the alkalies, which exert a disturbing influence on the titration when oxalic acid is used and barium carbonate is present. The potassium or sodium oxalate formed in the solution reacts with the barium carbonate as follows: K2CA+ BaCOg =BaC204+ K2CO3. The potassium carbonate is converted into oxalate by the further addition of oxalic acid. Carbon dioxide is liberated, and the potassium oxalate reacts with more of the barium carbonate, according to the equation given. More oxalic acid is therefore used than corresponds to the amount of free alkali originally present. This action is entirely prevented by the addition of a little barium chloride to the solution, which converts any potassium or sodium salts present into chlorides. The solution must be carefully pro- tected from carbon dioxide by one of the methods already given. It is standardized by titration against standard acid. 416. Oxalic Acid has beeii most largely used to titrate the excess of the alkali. 2.8636 * grams of the pure recrystallized acid dissolved in one liter of distilled water gives a solution 1 c.c. of which is equal to 1 mg. of carbon dioxide. As dilute solutions of oxalic acid are not stable, a fresh solution must be made as needed by weighing out the crystallized acid. * Note to student. — Show how this number is calculated. 304 VOLUMETRIC METHODS. 417. Sulphuric Acid of equivalent strength may also be used with advantage. The presence of barium carbonate does not inter- fere so much with the titration when sulphuric acid is used. Some workers who use this acid titrate the excess of the barium hy- droxide without removing any of the barium carbonate. Alkalies, if present, do not interfere with the titration if sulphuric acid is used. It may be made by diluting a stronger solution. 45.45 c.c. of a normal solution or 227.27* c.c. of a fifth-normal acid diluted to a liter will give an acid 1 c.c. of which is equal to 1 mg. of carbon dioxide. The acid may also be made by diluting concentrated acid so as to make an approximate solution, which is then standardized by titration with the barium hydroxide solu- tion, which in turn is standardized by titration against weighed amoimts of pure crystallized oxalic acid. Phenolphthalein is used as the indicator in all titrations. 418. The Results of the Analysis must be expressed in per cent by volume, the amount of carbon dioxide in the air being given in parts by volume in }0,000. The temperature of the air and the barometric pressure must be taken at the time the large bottle is filled. In very careful work the tension of water vapor in the air must also be taken. The volume of the air at 0° and 760 mm. pressure must then be calculated according to the formula V{P-v) ^o°.76o = (1 I 0036667^)760 ' "^^^^^ ^ ^® ^'^^ volume of the large bottle used, P is the barometric pressure in millimeters, and t the temperature when the sample of air is taken, p is the tension of water vapor in millimeters of mercury. The amount of carbon dioxide, being obtained in milligrams, must be reduced to cubic centimeters by multiplying the number of milligrams found by .508. EXERCISE 55. Detennination of Carbon Dioxide in the Air. 419. Standard Solutions. — Prepare a saturated solution of barium hy- droxide by treating 15 to 20 grams of barium hydroxide with warm dis- tilled water. Place in a well-stoppered bottle, and when the insoluble matter has settled, draw off by a siphon or large pipette about 250 c.c. and dilute to 1 liter. Protect the solution from the carbon dioxide of the air by placing the solution in a bottle arranged as shown in Fig. 50, p. 290. * Note to student. — Show how this number is calculated. TITRATION OF CARBONIC ACID. 305 Measure out 227.27 c.c. of N/5 sulphuric acid by means of a burette and dilute to 1 liter. If standard sulphuric acid is not at hand, weigh out 2.8647 grams of oxaUc acid and dilute to a liter. If the oxalic acid is used, about 0.2 gram of barium chloride must be added to the barium hydroxide solution. The bottles described on p. 302 are prepared : 420. Collecting the Sample. — The rubber stopper is inserted in the large bottle A, and into the hole of the stopper is inserted a smaller two-holed rubber stopper through which two glass tubes pass, one of which extends to the bottom of the bottle, while the other passes through the stopper only. The tube extending to the bottom of the bottle is connected to a long tube passing out of a window. The smaller glass tube is connected to a Bunsen suction-pmnp, which is allowed to draw air through the bottle for half an hour. The small bottle is filled with the barium hydroxide solution, the rubber stopper inserted so as not to enclose any air-bubbles, and after pressing it down to the mark, the excess of barium solution is removed and the glass plug inserted. When the large bottle has been filled with the outside air, the stopper containing the glass tubes is removed. The stopper is also removed from the smaU bottle, which is inserted into the hole in the stopper of the large bottle, which for this purpose is placed on its side. The temperature is taken from a thermometer which has been hung by the large bottle. The barometer is also read. The large bottle is shaken at intervals for about two hours. It is then inverted, the small bottle containing the barium hydroxide solution is removed, and the stopper inserted. 421. Titration. — In the meantime the standard solutions are compared, the titrations being carried out rapidly to prevent absorption of carbonic acid. Phenolphthalein is used as the indicator. If oxalic acid is used, the solution in the small bottle must stand overnight before being titrated. If sulphuric acid is used, it may be titrated when convenient, the glass plug being withdrawn from the stopper, and 25 c.c. of the solution withdrawn by means of a pipette and titrated immediately. Two or three titrations should be made with 25-c.c. portions. The difference between the amount of acid used for these 25-c.c. portions and the amount used in the stand- ardization for 25 c.c. is found and multiplied by the ratio of the volume of the small bottle to 25. This gives the number of cubic centimeters of acid equivalent to the carbon dioxide in the sample of air taken. Calculate the number of cubic centimeters of carbon dioxide by multiplying the num- ber of milHgrams by .508. Calculate the volume of air taken at 760 mm. and 0° Divide the former by the latter and multiply by 10,000 to find the number of parts of carbon dioxide per 10,000. OXIDATION AND REDUCTION METHODS. CHAPTER XXIV. POTASSIUM PERMANGANATE AND BICHROMATE SOLUTIONS. 422. Oxidation. — A number of very excellent volumetric methods are based on the fact that many substances exist in two states of oxidation. In the case of arsenic, we have arsenious oxide, AS2O3, and arsenic oxide, AS2O5. Sulphur forms several compounds of varying oxygen content, as sulphurous acid, H2SO3, and sulphuric acid, H2SO4. Substances capable of giving up oxy- gen, such as POTASSIUM PERMANGANATE, KMn04, couvert the former into the latter. The oxidizing substance itself may not contain oxygen, but may be able to oxidize by acting on water or some inter- mediary substance. Iodine is such a substance which is capable of converting sodium arsenite into sodium arsenate, as follows: Na3As03+ 12+ H2O = Na3As04+ 2HI. An alkali must be present to unite with the hydriodic acid so that the action might be represented as follows: Na3 ASO3 + 12 + 2Na0H = NajAsO^ + 2NaI + H^O . The mechanism of the reaction is probably quite complicated, as many reactions are possible under the circumstances. It may, however, be carried out so that 2 atoms of iodine are equal to 1 atom of oxygen. The result of the oxidation may not even be the addition of oxygen, as when ferrous chloride, FeCla, is converted into ferric chloride, FeCl3. Perhaps the simplest definition of oxidation is the change in valence of an element by the addition of negative atoms or radicles or the removal of positive atoms or radicles. The change in valence may not always be an increase, though it generally is. The oxidation of ammonia may result in a decrease 306 AVAILABLE OXYGEN. 307 of valence. It is apparent that the reactions taking place during oxidation and reduction are more complicated than those involved in acidimetry and alkalimetry. 423. Normal Oxidizing and Reducing Solutions. — The strength of all oxidizing and reducing solutions may nevertheless be expressed in as simple terms as the acid and alkaline solutions. The basis selected is oxygen, the strength of all standard oxidizing solutions being expressed in terms of available oxygen or its equivalent in chlorine, iodine, etc. A normal oxidizing solution is defined as one which contains 8 grams of available oxygen or its equivalent' per liter. Hydrogen has been selected as the basis of reducing solutions, 1 liter of a normal reducing solution containing 1 gram of available hydrogen or its equivalent. 424. Available Oxygen. — ^The distinction between available and non-available oxygen must be carefully made, just as the acid hydrogens of a molecule must be distinguished from the non-acid atoms in acidimetrJ^ Of the 4 atoms of oxygen in the molecule of potassium permanganate but 2| are available. The decomposition of 2 molecules of this compound may be represented as follows: 2KMnO, = K2O + 2MnO +50. The oxidation is generally carried out in acid solution, the potas- sium oxide and maiiganous oxide combining with the acid as follows : Kfi + 2MnO + 3H2SO4 = K2SO4 + 2MnS0, + SH^O. Five atoms of oxygen, or 80 parts by weight, being free for oxida- tion purposes, one-tenth of the gram molecular weight of 2 mole- cules of potassium permanganate, or 31.63 grams must be dis- solved in 1 liter to make a normal solution. The decomposition of potassium dichromate is similar: KA2O7 =K,0 +Cr203 +30. This substance is also used in acid solution, the potassium and chromium oxides reacting with sulphuric acid as follows: K,0 + CrA+4H2SO, =K,S04+ Cr2(S04)3+4H20. As 3 atoms of oxygen, or 48 grams, are available, one-sixth of the gram molecular weight of potassium dichromate must be dis; 308 VOLUMETRIC METHODS. solved in a liter to obtain a solution containing 8 grams of avail- able oxygen per liter. Iodine reacts as follows in the presence of a substance capable of absorbing oxygen: l2+H20=2HI+0. Two atoms of iodine liberate 1 atom of oxygen. The weight of 1 atom of iodine in grams must therefore be dissolved in a liter to give a normal solution. 425. Indicators. — No indicator is known which directly shows the presence or absence of available oxygen or hydrogen in a solution. Potassium permanganate serves as its own indicator. As it has an intense reddish-purple color in water solution, and when reduced in acid solution forms the colorless potassium and manganous sulphates, the least excess of the oxidizing solu- tion is indicated by the characteristic permanganate color. As long as even a trace of a reducing substance is present the color fades out by reduction, of the permanganate. As POTASSIUM BICHROMATE is ycllow and on being reduced gives the green chromic salts it is impossible to detect the presence of a small amount of the yellow dichromate in the presence of the large amount of green chromic salts. The dichromate solu- tion is generally used with a ferrous iron solution as the reducing substance. In acid solution potassium dichromate instantly converts any ferrous iron present into ferric iron. A very delicate test for ferrous iron is found in the blue color which even very small amounts produce in a solution of potassium eerricyanide. A solution of ferrous iron to which dichromate solution has been added may be tested for the presence of the former by taking out a drop of the solution and adding it to a drop of the indicator. When, after the successive addition of small amounts of dichromate solution, a test for ferrous iron is no longer obtained, the titration of the ferrous iron is complete. Solutions of IODINE are deep red to yellow according to the concentration. After reduction the iodine forms colorless salts. If the substance oxidized also forms colorless salts the titration may be conducted by using the iodine color as the indicator. This gives fairly accurate results, but a much more delicate indi- cator for iodometric work is found in starch solution, which POTASSIUM PERMANGANATE. 309 gives an intense blue color when the least excess of free iodine is present. Because of the tlelicacy of this method of obtaining the end-point, iodometric titrations are among the most accurate known to the chemist. POTASSIUM PERMANGANATE. 426. Decomposition of the Permanganate by Manganese Diox- ide. — ^This salt may now be purchased in a high state of purity. As it is not hygroscopic, solutions made by weighing out the C.P. article and dissolving in water are accurate within less than i%. Sulphates, nitrates, and chlorides are the common impurities which must be absent from the C.P. article. It is almost impossi- ble, however, to prepare a sample of this salt which is free from small amounts of manganese dioxide. The injurious effect of this impurity on solutions of potassium permanganate has been clearly shown by Morse and his pupils.* He has proved that the permanganate is slowly decomposed by even traces of the dioxide with liberation of oxygen and formation of more of the dioxide. As the rate of decomposition is proportional to the amount of the dioxide, the decomposition is accelerated with the age of the solution. The remedy consists in the complete removal of the dioxide. This is easily accomplished by filtering the per- manganate solution through asbestos which has been digested with strong aqua regia and well washed. Organic matter, dust, etc., which may reduce the permanganate with formation of the diox- ide must be rigorously excluded from the solution after filtration. All bottles and glassware with which the solution may come in contact must be freed from organic matter by washing with dichromate cleaning mixture. As distilled water usually contains ammonia and organic matter, it is well to allow the permanganate solution to stand some time before filtration so as to entirely decompose this organic matter. The distilled water may also be pin-ified by adding a very little potassium permanganate, boiling until the permanganate is decomposed, and then filtering off the manganese dioxide through asbestos. If the permanganate solution is made up in this manner and kept out of strong sunlight it is very * Am. Chem. Jour., 18, 411; 23, 313. '■^^^ VOLUMETRIC METHODS. permanent, readily keeping its strength for months. Portions of the solution withdrawn from the bottle should not be returned to it. STANDARDIZATION OF POTASSIUM PERMANGANATE SOLUTIONS. 427. By Pure Iron Wire. — ^As has been said, permanganate solutions made by weighing out the pure salt are within at least J% of the calculated strength. A higher degree of accuracy is obtained by carefully standardizing the solution. The most largely used and perhaps the most reliable method consists in the use of known amounts of iron. Soft-iron wire containing from 99.6 to 99.8% of iron may readily be obtained. The exact percentage of iron may be obtained by a gravimetric determination of the iron as given in Exercise 11, or of the total carbon, as given on page 374, the iron being found by difference. Formerly the practice was com- mon of dissolving the iron in dilute sulphuric acid, the air having been excluded by means of carbon dioxide. Under these circum- stances the iron is converted into ferrous sulphate, which was then titrated with the permanganate. As one of the impurities in the iron is carbon which is combined to a greater or less extent as a carbide of iron, the solution of the wire in acid leads to the formation of hydrocarbons, which are oxidized by the permangan- ate. As one part of carbon may reduce ten and one-half parts of potassium permanganate, and if the carbon is present as a hydro- carbon it may reduce twice this amoimt of the permanganate, the error from this source, even if the amoimt of carbon present is very small, may be very considerable. Titrations of solutions of iron wire have been made which indicate as much as 104% of metal in the wire if calculated on the assumption that the only reducing substance present is ferrous iron. Blair recommends the oxidation of the hydrocarbons by means of potassium chlo- rate * or potassium permanganate. f If this method is pursued the iron need not be dissolved in an atmosphere free from oxygen, as it must be reduced after oxidation of the organic material. If the chlorate is used the iron is dissolved in hydrochloric acid, a few crystals of potassium chlorate added, and the excess of * Blair, The Chemical Analysis of Iron, 4th Ed., p. 220. t Ibid., p. 95. POTASSIUM PERMANGANTE. 311 chlorine expelled by boiling. If permanganate is used the iron may be dissolved in sulphuric acid. 428. By Standard Iron Solution. — A very excellent reagent for standardizing permanganate solutions is recommended by Blair.* 100 grams of pure wrought iron free from manganese and arsenic, and in which the phosphorus has been determined, is weighed out and dissolved in nitric acid. The solution is evap- orated to dryness and the residue heated in a platinum crucible or dish as hot as possible with the blast-lamp. The ignited material "is groimd in an agate mortar, dissolved in hydro- chloric acid, evaporated to dryness to dehydrate the silica, taken up with hydrochloric acid, and the silica filtered off. The filtrate is diluted to about 4 liters. Portions of this solu- tion of 15 to 25 c.c. are weighed in small flasks and the iron precipitated and weighed as ferric oxide. The weight of the phosphoric acid must be subtracted from the weight of the oxide. The percentage of phosphorus in the iron multipliedby 1.6 f will give the percentage by which the weight of the ferric oxide should be reduced. After determining in this man- ner the exact amount of iron in a weighed portion of the solution, the permanganate may be standardized by taking weighed por- tions of the ferric chloride solution, reducing the iron, and titrating. 429. Reduction of Iron Solutions. — Various methods of reduc- ing the solutions of iron are in use. The most common sul stance used is GRANUi ATED MOSSY ZINC. The sohition of the iron is placed in a flask of about 250 c.c. capacity. It may be closed with a Bimsen valve, made by cutting a slit a quarter of an inch long with a sharp knife in a small rubber tube which is then closed at one end with a glass plug, while the other end is passed over a short piece of glass tubing Avhich passes through the rubber stopper * Blair, The Chemical Analysis of Iron, 4th Ed., p. 218. t Note to student. — How is this number obtained ? Fig. 52. 312 VOLUMETRIC METHODS. into the flask, as shown in Fig. 52. The gases within the flask can readily pass out through this valve, while the air cannot enter. The air in the flask is expelled by adding a little sodium carbonate. A few grams of zinc are then added, and the solution warmed nearly to boiling. The sides of the flask must not be heated with the flame of the burner above the point filled by the solution. If hydrochloric acid is present, the complete reduction of the iron is indicated by the change in color from the yellow of the ferric chloride to the light green of ferroxis chloride. Sulphuric acid is then added and the solution heated until all of the zinc is dissolved. Cold, recently boiled water is then added and the solution allowed to cool. When the stopper is removed from the flask it should be replaced as soon as possible, a little sodium carbonate having been added to expel the air. 430. Exclusion of the Air. — Another very convenient method of excluding the air consists of inserting into the flask a one-holed rubber stopper through which passes a glass tube bent to a slightly acute angle, so that when the flask is in- clined the long end of the glass tube may be vertical, as shown in Fig. 53. This end dips into a saturated solu- tion of sodium bicarbonate. On heat- ing the flask the air is expelled, and on cooling a partial vacuum is formed so that the bicarbonate solution is drawn up into the flask. On meeting ^!p:f the acid solution, carbon dioxide is evolved, which prevents the remain- der of the solution from entering the flask. After the addition of zinc and Fig. 53. reduction of the iron, sulphuric acid is introduced to dissolve the remainder of the zinc. The carbon- ate solution is remo^^ed, boiled and cooled distilled water is added, and the iron solution titrated. 431. Titration of Iron in Zinc. — As it is difficult to obtain zinc which is entirely free from iron, the amount of zinc used must be roughly weighed. The same amount of zinc is weighed out and POTASSIUM PERMANGANATE. 813 treated with sulphuric acid in a flask exactly as in reducing the iron. After all of the zinc has been dissolved the solution is titrated with the permanganate. The amount of solution used in this blank determination is subtracted from the volume of permanganate used in titrating the iron. The REDUCTION of the iron solution by means of stannous CHLORIDE, AMMONIUM SULPHITE, and HYDROGEN SULPHIDE are more especially adapted to the reduction of iron in irori ores, and will be described under that heading on page 318. These methods are equally well adapted to the reduction of iron solutions for the purpose of standardization. 432. Ferrous Ammonium Sulphate may also be used for standardizing permanganate solutions. This salt is remarkably permanent. When dry it may be kept for a long time without any appreciable oxidation of the iron. It also retains its water of crystallization with considerable persistence. If a sample of this salt is at hand which is free from ferric iron and has been com- pared with a standard of known purity, it offers a very convenient method of standardizing permanganate solutions, as the salt may be weighed out, dissolved in water, and, after the addition of dilute sulphuric acid, may be titrated with the permanganate immediately. The iron begins to oxidize very soon after being dissolved, so that the solution must be titrated immediately. As this salt may be titrated just as readily with bichromate solutions, its purity may be tested by titrating a weighed portion with a standard solution of potassium dichromate. 433. Oxalic Acid and the Oxalates, such as sodium oxalate and potassium tetroxalate prepared and tested exactly as given in Chapter XXI for standardizing alkahes and acids, may be used for standardizing permanganate solutions. It is evident that this class of substances affords a means of comparing stand- ard acids with standard oxidizing solutions. Testing the purity of the oxahc " acid by titrating it with a standard acid is of value only when the absence of alkaUes from the oxalic acid is assured. This is readily accomplished by igniting a portion of the acid. It should volatilize completely." The titration of oxalic acid and oxalates with permanganate solution is generally carried out in a warm solution acidified with sul- phuric acid. In the cold the oxidation of the oxalic acid is 314 VOLUMETRIC METHODS. very slow. It may also be accelerated by the addition of man- ganous sulphate solution. If the solution is warmed, the temper- ature must not be allowed to exceed 60°, as at a higher temperature small amounts of oxalic acid may be volatilized. At this temper- ature the beaker can be taken in the hand without serious discom- fort. The permanganate solution must be added slowly and with vigorous stirring to the oxalic acid solution. Much manganese dioxide must not be allowed to form in the solution or an excess of permanganate wUl be required, as the manganese dioxide decom- poses some of the permanganate with evolution of gaseous oxygen. 434. Direct Measurement of Oxygen. — ^An excellent method of standardization which does- not require the preparation of a pure substance is that of G. Lunge.* In this method a measured volume of the permanganate solution is decomposed by means of hydrogen peroxide and sulphui-ic acid according to the following reaction: 2KMn04+ 3H2SO4+ 5H,02 =K2S04 + 2MnS04+ SH^O + SOj. Twice the amount of available oxygen present in the permanga- nate is evolved in the gaseous form. This gas is measured and the weight of KMnO^ calculated. Lunge's nitrometer is used for carrying out the determination. The accuracy of the determination depends largely on the accuracy with which the nitrometer has been graduated or calibrated. As the solutions used are already saturated with air the evolution of the oxygen is very quickly complete. The gas is measured in the nitrometer over mercury. The details of the process and a diagram of the nitrometer are ^ven in Exercise 56, page 315. 435. Standard Sulphuric Acid may be used for standardizing a permanganate solution. A measured amount of the standard acid is placed in a beaker and a sufficient amount of hydrogen per- oxide to reduce about 25 c.c. of the permanganate solution is added. The peroxide is titrated with the permanganate solution. The faint red color giving the end-point is removed with a drop of the peroxide. Methyl orange is added and the excess of sul- phuric acid titrated with standard alkali. The difference between the amount of acid originally added and that found by titration gives the amount of acid required for the reaction given m the preceding paragraph. It is evident from *Chem. Ind., 1885, 168; Ber., 18, 1072; Zeit. angew. Ch., 1890, 10. POTASSIUM PERMANGANATE. 315 this equation that 25 c.c. of N/5 potassium permanganate will require 15 c.c. of N/5 sulphuric acid. From this relation the strength of the potassium permanganate solution can be calculated. The methods of standardization in which a known amount of iron is used are preferred for obvious reasons when the perman- ganate solution is to be employed for the determination of iron. EXERCISE s6. Preparation and Standardization of N/s Potassium Permanganate Solution. 436. Preparation and Preservation of the Solution. — Weigh out 6.35 grams of KMnOi. Warm about half a liter of distilled water to about 50°. Add the permanganate with stirring to the warm water. Prepare two carbon filters, using asbestos which has been digested with warm concentrated aqua regia for about an hotir and then well washed with water and finally transferred to a stoppered bottle with distilled water. These filters, as well as all vessels with which the permanganate solution is to come in contact, should be freed from organic matter by digestion with cleaning mixture, made by dissolving a few crystals of potassium dichromate in concentrated sulphuric acid. Suck the asbestos down with the filter-pump, being careful to leave a fihn projecting upwards aroimd the edge. By means of rubber stoppers arrange the filters as shown in Fig. 54. Filter the permanganate solution, being careful not to apply too strong suction for fear of break- ing the hter flask. Rinse the beaker carefully and wash the asbestos until the water comes through color- less. After being washed, the asbestos in the second filter should be very nearly pure white. If much brown manganese dioxide is seen, the solution must be filtered again, the filters being interchanged, and clean asbestos put in what is now the second filter. After cooMng, the solution is diluted to the mark with distilled water. This solution should be kept in the dark and out of con- tact with organic matter. Portions taken out of the bottle should not be returned to it. Fig. 54. STANDARDIZATION. 437. First Method. By Means of Pure Iron Wire. — Clean thoroughly by rubbing with sandpaper and filter-paper some soft iron wire in which the perontage of iron has been determined. When cleaned, wind in a spiral on a lead-pencil with as httle contact with the fingers as possible. Weigh aarefuUy about 0-250 gram, place it in a 250-c.c. Florence flask, and dissolve 316 VOLUMETRIC METHODS. with gentle heat in 10 c.c. concentrated hydrochloric acid and 20 c.c. ot water. When the iron is dissolved, throw in a few crystals of potassium chlorate and boil gently a few minutes vAih. a small Bunsen-bumer flame. Carefully avoid heating with the flame the sides of the flask above the hquid. Add about 2 grams of granulated zinc which is free from iron. Close the mouth of the flask with a rubber stopper through which passes a bent glass tube dipping into a solution of sodium bicarbonate, as shown in Fig. 53. Heat nearly to boiling till the .solution is of a pure green color, free from the slightest tinge of yellow. If the zinc is all dissolved before this point is reached, add another gram and continue heating until all of the iron is reduced. Allow the flask to cool somewhat. The bicarbonate solution will rise in the glass tube and enter the flask only to be expelled by the carbon dioxide evolved when it reaches the acid solution. Add a mixture of 10 c.c. of concentrated sulphuric acid and 20 c.c. of water and again heat until no particles of undissolved zinc remain in the solution or on the sides of the flask. Allow the solution to cool and then add cold but recently boiled distilled water until the flask is about two-thirds full. Titrate immediately with the permanganate solution until a rose-tint is produced which is permanent for two minutes. Repeat the determination imtil duplicates are obtained agreeing ■within less than 0.1 c.c. 438. Second Method. By Means of Mohr's Salt. — Weigh out 19.62 grams of pure recrystalUzed ferrous ammonium sulphate (FeS04.(NHj)2S04.6H20). Transfer to a beaker.^ dissolve in about 100 c.c. of water, and transfer to a 250-o.c. flask. Add 40 c.c. dilute sulphuric acid, shake, cool, and dilute to the mark. This solution wiU be exactly N/5 normal. Titrate imme- diately 25-c.c. portions with the permanganate solution mitil dupUcates are obtained and calculate the strength of the permanganate solution in erms of normal. 439. Third Method. By Means of Sodium Oxalate. — ^ Weigh out between 0.3 and 0.35 gram of pure sodium oxalate and transfer to a beaker. Add about 50 c.c. of water and a few cubic centimeters of dilute sulphuric acid and dissolve by gently heating the solution to 50° or 60°. Be very careful not to heat any higher. Add the permanganate solution slowly with constant stirring. If the solution does not clear readily, warm gently again. When nearly all of the oxahc acid has been oxidized, the solution, wiU decolorize the permanganate almost instantly. At this point add the permanganate cautiously until a faint permanent pink is obtained. If a permanent brown coloration or precipitate is obtained, too much perman- ganate has been added. Repeat until duplicates have been obtained One part of the oxalate equals 0.4717 part KMn04. 440. Fourth Method. Determination of Oxygen Evolved with Hydrogen Peroxide. — Set up the Lunge nitrometer as shown in Fig. 55. The box is designed to prevent loss of mercury if any is accidentally spilled. Suspend a thermometer near the nitrometer. Fill with nieroury and POTASSIUM PERMANGANATE. 317 test the joints for leaks by lowering the level-tube A so as to produce a vacuum in B and noticing if the height of the mercury column in B remains constant for ten minutes. Now raise the level-tube and pro- duce pressure in the apparatus, and again wait ten minutes to see if the mercury column moves. Measure out into a beaker 15 c.c. of the perman- ganate solution, and, after acidifying with 10 c.c. of dilute sulphuric acid, add from a burette hydrogen peroxide quite rapidly with stirring until the Fig. 55. solution is clear. Repeat the determination, measuring out the hydrogen peroxide solution first. Measure out into the bottle C the amount of hydrogen peroxide solution found to be equivalent to 15 c.c. of the permanganate solution, adding a slight excess. Add about 10 c.c. dilute sulphuric acid. Measure into the inner tube 15 c.c. of the permanganate solution. Insert the stopper firmly and bring the mercury in the two tubes to the same level. Turn the stop- cock of the nitrometer through 180° and bring the mercury in B just up to the key of the stop-cock, thus expelling the air through the funnel-tube. 318 VOLUMETRIC METHODS. After clamping the level-tube, reverse the key of the stop-cock and observe if the mercury falls or rises on connecting the tube B with the bottle. If it falls a little, air should be drawn out of C by lowering the level-tube. The stop-cook is then reversed, and the air expelled as before. This operation must be repeated until the mercury column in B neither falls nor rises when the key of the stop-cock is reversed. If the mercury in B rises, air must of course be admitted through the funnel-tube. The barometer and the thermometer suspended near the gas burette should now be read. The apparatus should then be allowed to stand undis- turbed for five minutes to see if the air in the bottle is at the room tem- perature. Any change in temperature will be indicated by movement of the mercury in B up or down, in which case the amount of air in C must be again adjusted. The bottle is now tilted slightly and gently shaken, so that the two liquids come in contact. The level-tube is lowered so that no considerable pressure is developed in B. The shaking is continued until all of the permanganate is decomposed. The mercury columns are brought to the same level and the apparatus allowed to stand for five minutes. If a change in the volume of the gas has taken place, the bottle is again shaken and the test repeated. When no more gas is evolved, the final adjustment of pressure is made. The mercury surfaces are brought as nearly on a level as possible with the eye. A drop of water is introduced into the funnel-tube and the stop-cock slowly turned. If the pressure within the tube B is not identical with the atmospheric pressure, the drop of water will start to move, the direction indicating in which tube the pressure is greater. By raising or lowering the level-tube slightly, the inequahty can generally be overcome in a few minutes. When atmospheric pressure has been secured within the tube B, the volume of the gas is carefully read, and after five minutes read again. If no change has occurred, the thermometer and barometer are again read. As the entire determination need not consume more than one hour, neither the temperature nor the pressure usually change an appreciable amomit. Compute the volume, at 0° and 760 mm. pressure, of the oxygen evolved. In this computation, the pressure of water vapor must be taken into account by deducting from the observed barometric pressure the vapor tension of water at the observed temperature. 1 c.c. of dry oxygen at 0° and 760 mm. pressure weighs .0014298 gram. 1 gram of oxygen equals 1.977 grams KMnO^. Repeat the determination until duplicates are obtained. DETERMINATION OF IRON IN IRON ORES. 441. Dissolving the Ore. — By digesting the powdered ore on the hot-plate with a mixture of equal parts of water and con- centrated hydrochloric acid the iron in some ores is completely dissolved; but most iron ores contain a small amount of ferrous DETERMINATION OF IRON 319 silicate or titanate which is not decomposed by hydrochloric acid. After digesting the ore until the residue is white or unaffected by further digestion with hydrochloric acid, it is filtered off and washed free from iron. After burning the paper the silica is volatilized by treatment with hj^drofluoric and sulphuric acids. The iron remains in the crucible and may be dissolved by digestion with hydrochloric acid, or it may be fused with acid potassium sulphate. The insoluble material may also be decomposed by fusion with sodium carbonate. If the ore contains organic mater- ial a little potassium chlorate must be added to the hydrochloric acid. 442. Reduction of the Iron by Zinc. — The solution of the ore may be deoxidized by means of zinc as given for the standardiza- tion of the permanganate solution. This method introduces an error in the determination if titanium is present, as it is partially reduced by the zinc. The reduction by ammonium bisulphite or hydrogen sulphide does not affect the titanium. It is carried out as directed in Exercise 57. 443. The Reduction of Iron by Means of Stannous Chloride is very rapid and with the addition of phosphoric acid to the solu- tion to remove the color of the ferric chloride the titration is very exact. The hydrochloric acid solution of the iron is heated nearly to boihng and stannous chloride solution added cautiously until the yellow color of the ferric chloride is replaced by the light green of ferrous chloride. A slight excess of stannous chloride is added and then all at once with vigorous shaking of the flask 60 c.c. of mercuric chloride are added. 60 c.c. of an acid manganous sulphate solution containing phosphoric acid is added and after diluting the solution with cold water it is titrated immediately with the permanganate solution. The manganous sulphate is introduced to prevent reduction of the permanganate by the hydrochloric acid. The stannous chloride solution is made by dissolving 30 grams of tin in 200 to 300 c.c. of strong hydrochloric acid and diluting to 1 liter after filtering through asbestos. The mercuric chloride solution is made by dissolving 50 grams of mercuric chloride in 1 liter of water. One Uter of the manganous sulphate solution should contain 66| 320 VOLUMETRIC METHODS. grams of crystallized manganous sulphate, 333^ c.c. of phosphoric acid (sp. gr. 1.3), and 133 c.c. of concentrated sulphuric acid. 444- Reduction of Iron by Hydrogen Sulphide. — The iron may also be reduced with hydrogen sulphide. For this purpose the solution is placed in a flask closed with a two-holed rubber stopper, through which passes a glass tube extending nearly to the bottom of the flask. The solution is heated nearly to boiling and hydrogen sulphide passed through the long glass tube until the iron is com- pletely reduced. The stream of hydrogen sulphide is replaced by a stream of carbon dioxide and the hydrogen sulphide expelled by boiling the solution for a few minutes. The flask is cooled while the stream of carbon dioxide is passing and the iron tit- rated immediately with the permanganate solution. The sulphur is not oxidized by permanganate in the cold.* EXERCISE 57. Determination of Ferrous and Total Iron in an Iron Ore. 445. Solution of the Ore. — Weigh out 0.5 gram of the finely powdered ore, transfer to a 100-c.c. beaker, and add 10 c.c. of concentrated hydrochloric acid and 10 c.c. of water. Cover the beaker with a watch-crystal and digest on the hot-plate until the residue is colorless or unaffected by further digestion with the acid. The solution must not be allowed to go to dry- ness. If necessary add a httle dilute hydrochloric acid as the solution evaporates. Decant the solution into a 250-c.c. flask, transfer the insoluble material to a small filter-paper, and wash with small portions of hot water. Throw the moist paper into a platinum crucible and burn the paper. Add 20 to 30 drops of concentrated sulphuric acid and as much hydrofluoric acid and warm gently. If the material dissolves, continue heating the crucible until the hydrofluoric acid is entirely expelled. If solution is not com- plete add, after expulsion of the hydrofluoric acid, about ^ gram of acid potassium sulphate and heat the covered crucible sufficiently to fuse the sulphate. When the material is entirely dissolved, let the crucible cool and dissolve the contents in hot water, if necessary adding a little hydro- chloric acid. 446. Reduction of the Iron by Means of Sulphurous Acid. — In the mean- time the main portion of the ore is reduced by means of ammonium bisul- phite. Ammonia is added until a small permanent precipitate of ferric hydrate is produced. 5 c.c. of a solution of ammonium bisulphite, made by * Jour. Am. Chem. Soc, Vol. XVII, p. 78, 1895; Chem. News, Vol. LXXIII, p. 123, 1896. ANALYSIS OF IRON ORES. 321 saturating strong ammonia with sulphur dioxide, is added and, after shaldng vigorously, the solution is gently heated until it is colorless. The solution of the material insoluble in hydrochloric acid is now added, as well as a mix- ture of 10 c.c. of concentrated sulphuric acid and 20 c.c. of water. The solution is then boiled until the steam no longer smells of sulphur dioxide. The flask is filled with boiled distilled water without mixing with the solu- tion of iron. The flask may be covered with a small watch-crystal and cooled by placing it in cold water. The solution may then be poured into a large beaker or porcelain dish and titrated with the standard permanganate solution. Calculate the percentage of metallic iron present in the ore. DETERMINATION OF FEREOUS IRON. 447. Solution of the Ore. — Weigh out i gram of the ore and transfer to the flask A, which should have a capacity of about 250 c.c. The simplest method of weighing the pre is to use a glass tube about 10 cm. long, having a diameter of about ^ cm. and sealed at one end. Weigh the tube roughly, intro- duce about i gram of the ore and weigh carefully. Insert the tube in the neck of the flask, and by gently tapping the tube, deposit the ore on the bottom of the flask. Withdraw the tube and weigh again. The difference between the last two weights is the weight of the ore in the flask. Insert the stopper with the tubes, as shown in Fig. 56. Place a solution of sodium bicarbonate in the beaker and pass a stream of carbon dioxide from a Kipp generator through the tube B for about ten minutes. Disconnect the tube B at the joint C and attach a funnel. Pour in 30 c.c. of dilute hydrochloric acid, raising the tube D out of the bicarbonate solution. Remove the funnel and again connect the tube B at C and continue passing the current of carbon dioxide. Dissolve the ore in the acid by means of gentle heat from the Bunsen burner. 448. Titration. — When no black particles remain in the insoluble por- tion remove the burner, replace the beaker of bicarbonate solution with a beaker of boiled and cooled distilled water, disconnect the tube B and close C with a clamp. Dissolve 3 grams of zinc in 10 c.c. of concentrated sulphuric acid and 20 c.c. of water. When the solution of the iron is cold, add the acid solution of zinc and titrate at once. Fig. 56. 322 VOLUMETRIC METHODS. 449. Calculation. — Compute the percentage of iron present as ferrous oxide. 1 c.c. X/5 potassium permanganate is equal to .01438 gram of fer- rous oxide. Compute the percentage of ferric oxide present. For this purpose subtract the number of cubic centimeters of permanganate solu- tion reduced by the ferrous iron in ^ gram of the ore from the number of cubic centimeters required to titrate the total amount of iron present. The difference gives tlie number of cubic centimeters of permanganate solution required to titrate the iron present as ferric oxide in \ gram of ore. 1 c.c. of N/5 permanganate solution is equal to .01598 gram of ferric oxide. ANALYSIS OF MANGANESE ORES. The only ores of manganese which are of commercial importance are pyeoltjsite and several related ores which consist essentially of manganese dioxide combined with small amounts of the lower oxides of manganese together with more or' less iron, silica, and other impurities. Formerly the manganese ore was of value chiefly on account of the oxygen present by which free chlorine, bleaching powder, etc., could be produced. Recently an ex- tended use for manganese has been found in connection with_ the production of iron and steel, so that nine-tenths of all man- ganese ores are said to be used for this purpose. The valuation of the manganese ore may therefore require the determination of the available oxygen or the percentage of manganese present. 450. Determination of Available Oxygen. — As the oxygen present is used to evolve chlorine, the most rational method of determination would seem to be to treat the ore with hydrochloric acid and determine the amount of chlorine evolved. This may be done with great accuracy, and the process will be described in the chapter on iodometric methods. This method is not so largely used at present as those in which the ore is dissolved in acid m the presence of a known amount of a reducing agent, and the excess of the latter determined by means of a standard solution of an oxidizing agent. Oxalic acid and ferrous iron have been most largely used as reducing agents with potassium per- manganate and potassium dichromate respectively as the standard oxidizing solutions. The finely powdered material is dried by spreading it out on a watch-crystal and heating for from four to six hours on a water- 'DETERMIA'ATION OF MANGAhESE. 323 bath, or in an air-bath maintained at 100°. The dried material must be weighed quickly, as it is hygroscopic. If oxalic acid is used the ore may be dissolved in an open beaker, as this reducing agent is not affected by the oxygen of the air. If ferrous iron is used the air must be excluded by one of the methods used during the reduction of iron with zinc. It is advisable, both when using oxalic acid and ferrous iron, which is most con\'eniently introduced as ferrous sulphate or ferrous ammonium sulphate, to weigh out the dry salts instead of making a standard solution of these re- ducing agents. The reducing solutions, especially of the iron, would rquire standardization every day, while one titration of the dry salt if preserved in a well-stoppered bottle would be suffi- cient. The errors of measuring out portions of a solution are also considerably greater than the errors of weighing a salt. DETERMINATION OF MANGANESE BY VOLHARD'S METHOD. Manganese may be determined very quickly and accurately by the method of Volhard. The manganese must be in neutral solution as a manganous salt. On adding potassium perman- ganate to such a solution all of the manganese present, including that which was added as permanganate, will be precipitated as the dioxide. The precipitate has a fairly strong acid character, so that varying amounts of the bases present in the solution will be carried down with the precipitate. The manganous manga- nese present is apt to be carried down in this manner, forming salts of the general character (Mn02)^MnO. Less permanganate solution will be used in this case than theory requires. This difficulty is almost entirely overcome by having a considerable amount of zinc sulphate present in the solution. The zinc will then be carried down with the manganese dioxide in preference to the manganous manganese. 451. Determination of Amount of Error. — The extent of the error from this source may be ascertained by a blank determination as follows: A given volume of the permanganate solution, 30 c.c. for instance, is carefully measured out into a beaker and reduced bv acidifying with sulphuric acid and adding hydrogen peroxide 324 VOLUMETRIC METHODS. in small portions, carefully avoiding excess. The solution is then transferred to a liter flask and neutralized by the addition of zinc oxide. If several grams of zinc have not been used some neutral zinc sulphate should be added. The solution is then diluted to 400 c.c. and warmed to 80°. 19 c.c. of the perman- ganate solution are then added rapidly and, after vigorous shak- ing of the flask, the permanganate is added in 0.1-c.c. portions, until a slight excess is present as shown by the pink color of the solution. The solution is vigorously shaken after each addition of permanganate. On standing a few minutes the precipitate settles so that the color of the solution can be seen on holding the flask up to the light. 452. Calculation. — If the reaction proceeded wholly according to the equation 2KMnO,+2H20 +3MnS0, =2H2S04 +K2SO, +5Mn02, exactly 20 c.c. of the permanganate solution would be required to titrate the manganese obtained from the reduction of 30 c.c. of the same solution. Generally, a little less than 20 c.c. will be used, the deficiency averaging about 1% of the total volume. The volume of the permanganate solution used in a given titration or the value of 1 c.c. in manganese must be increased in the pro- portion found by the blank determination. 453. Method of Adding the Permanganate Solution. — ^If the per- manganate solution is added in small portions to the solution of the m.anganous salt the error will be greater than when the total amount of permanganate is added all at once. If the amount of manganese present is entirely unknown one or more preliminary titrations must be made. In the first the permanganate solution is added several cubic centimeters at a time until the end-point has been reached or passed. In the second nearly the total amount of permanganate is added all at once and the end-point reached by the addition of amounts less than a cubic centimeter. In the third titration the total amount of the permanganate solution less a few tenths of a cubic centimeter is added at once and the end-point reached by adding portions of a tenth of a cubic centimeter or less. 454. Interfering Metals. — Of the other metals which occur in manganese ores iron is most common, but it does not interfere DETERMINATION OF MANOANESE. 325 with the titration of the manganese since it is precipitated on neutralizing the solution with zinc oxide. Copper is also precipi- tated by means of the zinc oxide. Nickel and lead, if present, produce high results. The lead may be removed by treating the solution with sulphuric acid or may be removed with cobalt and nickel as sulphides. The solution is neutralized with ammonia and ammonium sulphide added. It is then acidified with hydro- chloric acid and filtered from the sulphides of cobalt, nickel, and lead. The hydrogen sulphide is expelled by boUing and the manganese titrated after neutralization of the solution with zinc oxide. Chromium as well as cobalt interferes if present even in small amounts. Manganese may be separated from chromium by precipitation as dioxide from a strong nitric acid solution by means of potassium chlorate. It is filtered off, washed, and, after solution in hydrochloric acid and neutralization with zinc oxide, titrated with the permanganate solution. 455. Strength of Solution. — As the Volhard reaction may be represented as follows, MnO + = MnOj, one liter of a fifth-normal permanganate solution would be equal to 1/10 of the molecular weight of MnOz in grams if the oxygen were evolved from the potassium permanganate in the usual manner. As in the Volhard reaction it decomposes as follows, 2KMnO, =K20 + 2Mn02+ 30, a fifth-normal solution has only three-fifths of its usual strength. One liter of the fifth-normal solution is therefore equal to 3/50 of the molecular weight of MnOj, i.e., 1 c.c. is equal to .00522 gram of manganese dioxide. EXERCISE 58. Determination of Available Oxygen in Pyrolusite. (a) By Oxalic Acid. The ore is finely powdered, spread out on a watch-crystal, and dried four to six hours on the water-bath or in an air-bath kept at 100°. Portions weighing 0.870 gram are quicldy weighed out and transferred to beakers. Portions of pure recrystallized oxalic acid weighing 1.2605 grams are added to the beakers containing the ore. About 50 c.c. of water and a few cubic centimeters of dilute sulphuric acid are added to each beaker and the solu- tion heated gently until the ore is dissolved. The temperature must not 326 VOLUMETRIC METHODS. be allowed to rise above 60°. Without filtering off the insoluble residue, which is generally white, the warm solution is titrated with standard per- manganate solution. More dilute sulphuric acid is added if the permanganate color fades slowly or manganese dioxide is precipitated. The percentage of MnOj present is found by subtracting the number of cubic centimeters of N/5 potassium permanganate used from 100. If pure recrystallized oxalic acid is not at hand, the C.P. article may be used. It is tested as follows : One-gram portions of the oxalic acid are weighed out and titrated against the permanganate solution. The amount of oxalic acid which will reduce exactly 100 c.c. of the permanganate solution is then calculated. This amount is then weighed out instead of 1.2605 grams of the pure acid. It is advisable to test even the most carefully prepared article by titration against the permanganate solution used. (b) By a Ferrous Salt. Determine the percentage of MnOj in the same or another sample ot pyrolusite as follows : 0.870 gram of the finely powdered and dried material is weighed out and transferred to a 250-c.c. flask. Add a httle distilled water and some dilute sulphuric acid and then a gram or two of sodium bicarbonate. Insert a rubber stopper fitted with a bent glais tube dipping into a solution of sodium bicarbonate exactly as used in the standardization of the permanganate solution with iron wire. An amount of ferrous sul- phate or of ferrous ammonium sulphate which has been found by titration to be exactly equal to 100 c.c. of the permanganate solution is now added. The solution is warmed gently until no more black particles of the pyrolusite remain undissolved. The excess of ferrous sulphate is titrated with the permanganate solution, a red color permanent for one-half minute being taken as the end-point, both in this titration and in testing the purity of the ferrous salt. The number of cubic centimeters of N/5 permanganate used, subtracted from 100 gives the percentage of MnOj present. EXERCISE 59. Determination of Manganese in Pyrolusite. 456. Solution of the Ore. — 1.5 grams of the dried ore are weighed out and digested with 30 c.c. of concentrated hydrochloric acid until no black particles remain or until no further solvent action is noticed. The insoluble residue is filtered off on a small paper, the solution being allowed to flow into a 250-c.c. flask. After washing the paper free from manganese, the residue is fused with sodiiun and potassium carbonates unless the absence of manganese has been shown by a previous test. The moist paper is thrown into a platinum crucible and burned. Four to six times as much sodium and potassium carbonates are added and the contents of the crucible DETERMINATION OF TIN. 327 fused. The presence of manganese is generally indicated by the charac- teristic gi-een color. Whether the fusion is green or not the fused material is dissolved in water, the solution acidified vnth. hydrochloric acid and added to the main solution of the ore. 457. Neutralization of the Solution. — The solution ii diluted to the mark and 50-c.c. portions withdrawn with a calibrated pipette. These portions are placed in Florence flasks of 750- to 1000-c.c. capacity. Zinc oxide is added in small portions mth vigorous shaking of the flask until the solution is neutral. At this point the iron which is almost invariably present begins to be precipitated as the hydroxide. The precipitation of the iron is hastened by warming the solution. The zinc oxide should be addsd cautiously until the solution is colorless, indicating the complete precipitation of the iron. A large excess of the zinc oxide should be avoided. If too much has been added, so that the solution is turbid or the precipitate wliitish instead of red, it may be removed by the cautious addition of hydro :hloric acid. A shght excess of zinc oxide is not disadvantageous. 458. Titration. — The solution is diluted to about 400 c.c, warmed to 80°, and titrated with the permanganate solution, adding 4- or 5-c.c. por- tions at once and shaking vigorously. After aUo^^dng the precipitate to settle a little, the color of the solution can easily be observed. When the characteristic permanganate color has been given to the solution, the titra- tion is interrupted, and to the second flask, after neutrahzing, diluting to 400 c.c, and warming, is added the total amount of permanganate solution which was added to the first flask less the final portion which gave the end- point. After thorough shaking, 1-c.c. portions of the permanganate are added until the end-point js reached. To the third flask the total amount of permanganate solution within 1 c.c. is added, and after shaking, the end- point is reached by adding 0. 1-c.c. portions of the permanganate solution. The fourth and fifth portions of the solutions are also titrated. Calculate the per cent of manganese in the ore. 459. Determination of Tin. — ^Tin may readily be determined voiumetrically by titrating the ferrous iron produced by the reduction of ferric chloride by stannous chloride or metallic tin. If the tin is obtained as the stannous salt, it is treated with an excess of ferric chloride, when the following reaction takes place: 2FeCl3 + SnCl^ = 2FeCl2 + SnCl4. By determining the amount of ferrous iron produced, the amount of tin present in the stannous condition is found. This determina- tion is of value because the tin salt is frequently used as a reducing agent. Cupric chloride may be used in place of the ferric chloride. If the tin is in the metallic condition it may be dissoIve4 328 VOLUMETRIC METHODS. in hydrochloric acid with exclusion of air and the solution of stannous chloride treated with ferric chloride. The metallic tin may also be treated directly with ferric uhloride solution, in which it dissolves readily, especially if it is finel)'^ divided. No hydrogen is evolved, but the tin dissolves according to the following equa- tion: Sn + 4FeCl3 = 4FeCl2 + SnCl,. If iron is present in the tin the solution in hydrochloric acid must be allowed to stand twelve hours after the addition of metal- lic zinc. The tin is completely precipitated and may be removed from the zinc with a brush, washed, dissolved in ferric chloride and the ferrous iron determined. The tin in stannic compounds may be determined in the same manner, the solution of the stannic compound in hydrochloric acid being treated with zinc as directed above. EXERCISE 60. Analysis of Stannous Chloride. Make a solution of ferric chloride by dissolving 60 grams of the com- mercial salt in a liter of water. One gram of the stannous chloride is weighed out, placed in a beaker, and 50 c.c. of the ferric chloride solution added together with a few cubic centimeters of dilute hydrochloric acid. 50 c.c. of the ferric chloride solution is placed in a similar beaker and the same amount of hydrochloric acid added. When the stannous chloride is dissolved, add a few grams of sodium sulphate and at least 200 c.c. of boiled and cooled distilled water to each beaker and titrate the tin solution with standard permanganate solution until a faint rose color permanent for one-half minute is obtained. Add permanganate to the ferric chloride solution drop by drop until the same color is obtained. The amount of permanganate added to this solution must be subtracted from the amount added to the solution of the tin. Calculate the amount of crystallized stannous chloride, SnCl2.2HjO, present. 1 c.c. N/5 permanganate equals .02259 gram of this salt. DETERMINATION OF CALCIUM. 460. Titration of Calcium Oxalate. — Calcium may be deter- mined very rapidly and accurately by titrating the oxalic acid in the calcium oxalate precipitate with standard permanganate solution. The calcium may be preci]iitated in the usual manner by means of ammonium oxalate and digesting the precipitate DETERMINATION OF CALCIUM. 329 until the solution is clear. The precipitate is thoroughly washed with hot water. If it is large it is advisable to wash entirely by decantation. The filter-paper through which the wash-water is passed is treated with a little warm dilute sulphuric acid and then washed with water, allowing the liquid to flow into the beaker containing the bulk of the precipitate. When the precipitate on the filter-paper has been entirely dissolved and washed out, the portion in the beaker is dissolved by gently warming the solution and adding more dilute sulphuric acid if necessary. The oxalic acid is then titrated Ln the usual manner. 461. Titration of the Excess of Oxalic Acid. — The washing and dissolving of the precipitate may be obviated, if the calcium is precipitated in a measuring-flask of convenient size, by neutral- izing with ammonia and adding a measured excess of oxaJic acid. For this purpose a standard solution may be used, or the pure crystals may be weighed out. After digesting the precipitate until the solution is clear it is diluted to the mark on the flask with water and after thorough shaking is allowed to settle a few minutes. As the precipitate occupies a small volume & few drops of water should be added after the solution is up to the mark. The exact amount to be added may be foimd by dividing the weight of the precipitate by its specific gravity. The solution is filtered through a dry paper and the oxalic acid in a measured volrune of the filtrate titrated with standard permanganate solution after acidifying with sulphuric acid. The solution need not be cooled to the room temperature if the portion of the filtrate to be titrated is measured out as soon as filtered. A very convenient method of procedure is to precipitate the calcium in a 500-c.c. flask and after digestion allow to cool somewhat, dilute, and shake. After standing a few minutes the solution is filtered through a dry paper into a 250-c.c. measuring- flask. The solution is allowed to flow into this flask until it is filled to a little above the mark. The excess is taken out with a glass tube used like a pipette. The solution is then poured into a beaker, the flask rinsed out, and the oxalic acid titrated. When a large number of calcium determinations must be made the volumetric method is convenient and rapid. If only a few need be made it is doubtful if time would be saved. 330 YOLUMETRIC METHODS. POTASSIUM DICHROMATE SOLUTION. 462. Conditions of Use. — The use of this substance as an oxi- dizing solution is limited by the fact that it does not give up its oxygen to many substances which can be oxidized by potassium permanganate or iodine. The titration must always be so con- ducted that the end-point is obtained by titrating ferrous iron, using potassium ferricyanide as the outside indicator. On the other hand the solution has the advantage that it is very stable, being unaffected by dust or the organic matter which might come in contact with it. It may be used in the Mohr burette, which is closed with a rubber tube and a pinch-cock. Dilute hydrochloric acid is not oxidized as it is by potassium permanganate. The presence of this acid in the solution does not interfere with the titration when carried out cold. Potassium dichromate may also be easily obtained very pure, so that the solution may be made by weighing out the pure salt. Its oxidizing power is consider- ably increased when used in hot concentrated solution. It differs from potassium permanganate in that under these conditions it does not lose an appreciable amount of oxygen in the free con- dition. 463. Pure Potassium Dichromate may generally be obtained commercially. Sulphates and chlorides are the common impuri- ties. If the pure salt is not at hand, it may be made by recrys- tallizing the impure substance. The salt contains no water of crystallization and is not hygroscopic. It may be fused without loss of oxygen, and this method of drying is frequently used. If dust or organic matter is present, the fused salt is reduced. If heated higher than necessary to fuse it, it may lose oxygen. As the salt dried at 100° loses only an inappreciable amount when fused, this method of drying seems to be unnecessary. 9.817 grams of the pure salt are weighed out, dissolved in water and diluted to a liter to make a fifth-normal solution. 464. Standardization by Means of Iron Wire. — The strength of the solution may be verified by standardization with a known amount of iron as directed for the standardization of potassium POTASSIUM DICHROMATE. 331 permanganate solution. The oxidation of the organic matter in the hydrochloric acid solution of the iron wire by means of potas- sium chlorate is unnecessary, as the dichromate is not reduced by the hydrocarbons present. The iron wire may therefore be dis- solved in dilute sulphuric acid, the air being excluded by means of carbon dioxide by the methods already given. If the standard- ized solution of ferric chloride is used, it is not advisable to reduce it by means of zinc, as the zinc salts react with the ferricyanide used as the indicator, forming the white zinc salt which obscures the end-point. The reduction by means of staimous chloride solution is well adapted for use when the iron is to be titrated with dichromate solution. The stannous chloride is added cautiously to the hot ferric chloride solution untU the characteristic color of the ferric chloride has disappeared. The slight excess of stan- nous chloride is removed by adding a considerable amount of mercuric chloride solution. After diluting the solution somewhat and adding dilute sulphuric acid the ferrous iron is immediately titrated with the dichromate solution. The reduction with ammo- nium bisulphite and hydrogen sulphide is carried out exactly as directed for titration with the permanganate solution. 465. Standardization by Means of Pure Ferrous Ammonium Sulphate is carried out in the same manner as when a^permanga- nate solution is standardized with this salt, the end-point of the titration being ascertained by means of potassium ferricyanide. The standardization by means of oxalic acid or oxalates is not available for dichromate solutions, nor is the decomposition by means of hydrogen peroxide and measuring the gas evolved. A potassium permanganate solution which has been carefully stand- ardized by any of these methods may be used as a standard for a dichromate solution, using a ferrous salt to make the comparison. In this case the errors of three titrations would be included in the final result. 466. Analysis of Iron Ores. — When a dichromate solution is used for the determination of the iron in an iron ore, it is decom- posed in exactly the same manner as when the iron is titrated with a permanganate solution. Zinc is not used for the reduction, and when stannous chloride is employed, the addition of the phosphoric acid and the man^anous sulphate is unnecessary. As 332 VOLUMETRIC METHODS hydrochloric acid is not readily oxidized by potassium dichromate, the solution of iron need not be so thoroughly cooled as when permanganate is used. EXERCISE 6i. Preparation of Standard K/5 Potassium Dichromate Solution. Weigh out 9.817 grams of pure dry potassium dichromate. Dissolve in warm water, transfer the solution to a liter fiask, cool, and dilute to the mark. 467. Standardization by Ferrous Ammonium Sulphate. — 1.962 grams of the pure salt are weighed out and dissolved in water, dilute sulphuric acid is added, and the solution titrated immediately with the dichromate solution. 468. Indicator. — The potassium ferricyanide solution to be used as the indicator must be made immediately before it is required for use. A crystal about the size of a pea is placed in a small beaker and 75 to 100 c.c. of dis- tilled water added. The solution is most readily withdrawn by means of a short glass tube. A row of single drops are placed by means of this glass tube on a white porcelain plate. On touching one of these drops with a rod dipped in a solution of ferric chloride only a light-brown color should be produced. If any blue color develops, the ferric chloride should be boiled after the addi- tion of a drop or two of nitric acid and the test repeated. If a blue color is still developed, the ferricyanide is contaminated with ferrocyanide and cannot be used. When an indicator has been obtained which is satisfactory, the progress' of the titration is ascertained by touching drops of the indicator with drops of the solution taken out on the stirring-rod. 469. Titration. — To avoid loss of ferrous iron, nearly all of the calcu- lated amount of the dichromate solution should be added to the iron solu- tion before a test is made. Two or three drops of the dichromate solution are then added at a time and the test made until only a faint blue color is developed. A drop at a time is then added until the spot test shows no blue color. No change in the color of the spot test can then be noticed on adding' more dichromate to the iron solution. 470. The Standardization by Means of Soft-iron Wire in which the per- centage of iron has been determined is carried out as follows: Prepare a Bunsen valve as directed on p. 311 and a 250-c.c. Florence flask. Clean some soft-iron wire, in which the percentage of iron has been determined, by rubbing with sandpaper and filter-paper. Make it into a spiral by wind- ing on a lead-pencil. After cleaning do not touch the wire with the fingers. Weigh from 0.25 to 0.3 gram and transfer to the flask. Add about 15 c.c. of dilute sulphuric acid and a gram or two of sodium bicarbonate. Close the flask with the rubber stopper and the Bunsen valve and warm gently until the iron is dissolved. Add cold boiled water until the flask is about half full, DETERMINATION OP CHROMIUM. 333 then throw in a gram or two of sodium bicarbonate, and if the solution is not strongly acid, add a few cubic centimeters of dilute sulphuric acid and titrate immediately with the dichromate solution. EXERCISE 62. Determination of Iron in an Iron Ore. To determine the iron in an iron ore, ^ gram of the powdered ore is weighed and dissolved exactly as directed in Exercise 57, page 320. The solution of the iron is reduced by means of stannous chloride. It is heated nearly to boiling and the stannous chloride solution added drop by drop until the color of the solution is a pure green and a slight excess of the stannous chloride is present. Cool the solution by allowing cold water to flow over the flask. 50 c.c. of mercuric-chloride solution is added all at once and the solution vigorously agitated by rotating the flask with the hand. The iron is immediately titrated with the dichromate solution. 1 c.c. of N/5 dichromate solution is equal to .01118 gram of iron, or .01598 gram of ferric oxide. Compute the percentage of metallic iron and also of ferric oxide in the ore. 471. Determination of Chromium in Chrome Iron Ores. — These ores are most readily decomposed by fusion mth an alkaline oxidizing mixture by which the chromium is converted into chro- mate, which on treatment with water dissolves, leaving the iron as oxide. After destroying the excess of the oxidizing agent, the chromate may be determined by means of a known amount of ferrous iron. The best fusion mixture for this purpose seems to be 4 to 5 parts of sodium hydroxide and 3 to 4 parts of sodium peroxide, or 8 to 10 parts of sodium peroxide alone. The fusion must be conducted in a silver, nickel, or copper crucible, as plati- num is strongly attacked by this mixture. EXERCISE 63. Determination of Chromium in Chrome Iron Ore. 472. Decomposition of the Ore. — Grind the ore in an agate mortar until it is reduced to the finest powder. Weigh out ^ gram and transfer to a nickel or copper crucible. Add 4 to 5 grams of sodium peroxide, being care- ful to select the yellow material. The white crust on top is sodium carbonate and oxide resulting from decomposition of the peroxide. Mix the materia/ 334 VOLUMETRIC METHODS. well with a platium wire or glass rod. This can be more efficiently done it the peroxide is placed in the crucible first and the ore on top. Heat at first with a very small flame. After about ten minutes the material should be entirely hquid. It should be kept in this condition for ten or fifteen minutes. Allow to cool until a crust forms on top, add 1 gram of sodium peroxide, and fuse again for about five minutes. After cooling, transfer the crucible to a porcelain dish and dissolve the fused mass in hot water. Take out and rinse the crucible. If the solution is purple, add a little more sodium peroxide. Boil the solution for ten minutes to decom- pose the sodium peroxide. Filter off and wash thoroughly the insoluble material. Any portion of this material which is not soluble in hydrochloric acid must be fused with a little more sodium peroxide, as it may be unde- composed portions of the ore. 473. Titration of the Chromic Acid. — Acidify the filtrate ■with dilute sul- phuric acid and add an excess of about 15 c.c. Add a weighed amount of ferrous ammonium sulphate which is more than sufficient to reduce the chromium present. The presence of an excess of the ferrous salt may be ascertained by touchin"; a drop of the ferricyanide indicator with a glass rod wluch has bee;i dipped into the solution. A sufficient amount of the , ferrous salt should be placed in a weighing-bottle and carefully weighed. After adding an excess to the chromate solution the residue is weighed. The amount added is found by difference. 5 grams of the ferrous ammo- nium sulphate will generally be found to be a sufficient amount. The excess of ferrous iron h titrated with standard dichromate solution. The value of tlia ferrous salt is ascertained by titrating 2-gram portions with the dichromate solution. Calculate the volume of the dichromate solution equivalent to 5 grams of the ferrous salt or the amount added to the solution of the chrome ore. The difference between this number and the number of cubic centimeters used in titrating the excess of ferrous iron gives the amount of cliromate present in the solution of the ore in cubic centimeters of N/5 dichromate solution. Calculate the percentage of CrjOj in the ore, 1 c.c. of the N/5 dichromate solution being equal to .005073 gram of CrjOj. PROBLEMS. Problem 11. The calcium in i gram of a sample of dolomite was pre- cipitated with 1 gram of crystallized oxalic acid and the excess titrated with 34 c.c. of N/5 potassium permanganate solution. Calculate the per cent of calcium oxide in the dolomite. How much dolomite and oxalic acid must be taken so that the per cent of calcium oxide will be given by the number of cubic centimeters of N/5 potassium permanganate? Problem 12. In the analysis of a sample of stannous chloride, a half- gram portion was treated with ferric chloride solution and titrated with N/5 potassium permanganate solution requiring 20 c.c. The tin in another half-gram portion was weighed as stannic oxide, .330 grams being obtained. Calculate the per cent of SnCl8.2H20 and SnCl^ present. CHAPTER XXV. lODOMETKIC METHODS. 474. Conditions of Use of Iodine Solutions. — lodometric meth- ods are largely used, and are very accurate on account of the great delicacy of the indicator available, starch solution giving a distinct blue color with a very small amount of free iodine. Iodine acts as an oxidizing agent in acid and neutral solutions. The caustic alkalies, as well as solutions of the alkali carbonates, combine with the iodine, rendering titration in such solutions impossible. Sodium or potassium bicarbonate is not acted on by iodine, and as solutions of these salts are alkaline toward all acids stronger than carbonic acid, by working with solutions of these salts, iodine may be used to oxidize many substances which require an alkaline solution. It is therefore possible to use iodine to oxidize a great variety of chemical substances. The solubility of iodine in water solutions of potassium iodide, as well as in alcohol, chloroform, glacial acetic acid, and other organic solvents, still further enlarges its use, since both the iodine and the substance acted on must be in solution. The use of iodine is somewhat restricted because its solutions are not stable, while many of the methods of standardization are Indirect. The volatility of the iodine also introduces difficulties in its use. Iodine solutions must not come in contact with rubber, cork, or other organic material. Paraffine is unacted on. 475. Solvents of Iodine. — The most commonly used solution of iodine is made by dissolving the element in a water solution of potassium iodide. An unstable compound of the formula KI3 is formed under these circumstances. A considerable excess of potassium iodide must be present in order to dissolve the iodine. Alcoholic solutions of iodine have been largely used for the deter- mination of the so-called iodine number of oils. For this purpose 33.5 336 VOLUMETRIC METHODS. a solution in glacial acetic acid has recently superseded the alcoholic solutions formerly used. 476. Reducing Solutions. — As many of the methods of stand- ardizing iodine solutions consist in first standardizing a reducing solution, which is then compared with the iodine solution, the reducing solution must always be at hand in iodometric work. The solution most commonly used for this purpose is sodium THiosuLPHATE. The iodine is oxidized according to the following equation: 2Na2S203+l2 =Na2S40, +2NaI. A solution of sodium arsenite is sometimes used, especially for the titration of bleaching-powder. It reacts with the iodine as follows : Na^HAsOg + 12 +2NaHC03 = Na^HAsO, + 2NaI + H^O + 2CO2. Neither of these solutions is absolutely stable, though more so than the iodine solutions. The most active decomposing influence on the sodium thiosulphate solution seems to be the action of carbon dioxide in the presence of oxygen and sunlight.* The solution should therefore be made by dissolving the thiosulphate in water from which the carbon dioxide has been expelled by boil- ing. The carbon dioxide may then be excluded from the bottle by means of a soda-lime tube or the solution may be protected from the air by pouring a layer of petroleum-oil over it and siphoning out the solution. Made and protected in one of these ways, the solution will remain unchanged for months. 477. Preparation of Solutions. — Although the commercial re- sublimed iodine may frequently be obtained free from all impuri- ties except moisture, which may be removed by drying in a desicca- tor over sulphuric acid, a given sample cannot be relied on until tested. For this purpose a portion is sublimed according to the direction given in Chapter II, p. 34. A solution made from this resublimed iodine may be titrated against a thiosulphate solu- tion, which is then titrated against a solution of a weighed portion of the unpurified iodine dissolved in potassium iodide solution. The sodium thiosulphate cannot be weighed out exactly because of uncertain hydration, unless it has been recrystallized and care- * Topf. Zeit. anal. Chem., 26, 150. STANDARDIZATION. 337 fully dried. A salt may sometimes be obtained in this manner which gives solutions of exact strength. STANDARDIZATION. Neither the iodine nor the sodium thiosulphate solution should be relied on without standardization, although if both solutions are made from carefully purified material, and on comparison by titration are found to be of the strength calculated from the weight of material used, they may be employed without further verifica- tion, except in the most careful work. 478. Resublimed Iodine. — The thiosulphate solution may be standardized by means of small weighed portions of resublimed iodine. As the iodine is hygroscopic, it should be resublimed immediately before use. The loss by volatilization may also be considerable unless the weighed portions are immediately dis- solved in potassium iodide and titrated without delay. 479. Potassium Dichromate, as well as various other oxidizing substances, have been recommended for use in standardizing the thiosulphate solutions. Potassium dichromate can readily be obtained pure. It will liberate a definite amount of iodine from a potassium iodide solution acidified with hydrochloric acid. The reaction takes place according to the following equation: K2Cr207+ 6KI+ 14HC1 =8KCH-2CrCl3+3l2+7H20. A sufficient amount of potassium iodide must be present to dis- solve the iodine liberated, which is then titrated with the thio- sulphate solution. The decomposition of the potassium dichro- mate is not complete unless the potassium iodide, as well as the hydrochloric acid, is present in considerable excess. If a suffi- cient amount of hydrochloric acid is not present, the blue starch iodide color will reappear on adding more hydrochloric acid after titrating the iodine with the thiosulphate, solution. The fol- lowing proportions will generally be found satisfactory: To 10 c.c. of fifth-normal potassiuin-dichromate solution, 10 c.c. of a 10 per cent potassium-iodide solution and 5 c.c. of strorg hydrochloric acid is added. On account of_ the volatility of the iodine, it is advisable to carry out the operation in a glass-stoppered bottle. 480. Potassium Permanganate reacts more readily with the hydriodic acid. If a standard solution is at hand, a measured 338 VOLUMETRIC METHODS. volume may be added to a solution of potassium iodide acidified with hydrochloric acid and the iodine liberated titrated at once. The decomposition takes place according to the following equation: KMn04+5KIf8HCl=6KCl+MnCl2+4H20-h5I. 481. The lodates of Potassium and Sodium have also been recommended for use in a similar manner. They are decomposed by hydriodic acid instantly and completely. The acid potassium iodate is very readily obtained in a high state of purity. This salt has the disadvantage of being sparingly soluble in water, and on account of the large percentage of available oxygen the quanti- ties to be weighed out are small, necessitating very careful weigh- ing. If a large amount is weighed out, the solution must be m.ade up to a definite volume and portions measured out. The decom- position takes place according to the following equation: KH(IO3)2+10KI-|-llHCl = llKCl+6H2O+6l2. 482. Barium Thiosulphate. — The iodine solution is standardized by titration with a sodium thiosulphate solution which has been standardized by one of the methods already given. It may be directly standardized by the, titration of a weighed amount of pure bariiun thiosulphate. This salt may be made hj mixing together a warm solution of 50 grams of sodium thiosulphate in 300 c.c. of water and 40 grams of barium chloride dissolved in the same amount of water. The solution should be stirred vigorously while the salt crystallizes out. It is filtered off, washed with water until free from chlorides, and dried in the air. It contains one molecule of water of crystallization, its formula being BaSzOa.HzO. It is very sparingly soluble in water, a solution made by shaking an excess of the salt with water at 17.5° for fifteen minutes being exactly N/100 in strength. Because of its high molecular weight careful weighing is imnecessary. Quantities weighed out for the standardization of iodine solutions are treated with water in a beaker. Because of the slight solubility of the salt, the iodine solution is added to the beaker while undissolved thiosulphate still remains. When all or nearly all of the salt is dissolved, starch solution is added and the addition of the iodine solution continued until a permanent blue color is produced. IODINE SOLUTION. 3.39 483. Pure Arsenious Oxide may also be used for the standard- ization of iodine solutions. Weighed portions are dissolved in sodium bicarbonate solution, and after the addition of starch solution the iodine is added until the end-point is reached. 484. The Starch Solution is made by placing a gram or two of starch in a porcelain mortar, adding a little water, and grind- ing to a smooth paste. 150 or 200 c.c. of boiling water are then poured into the mortar while stirring continually with the pestle. The solution is then poured into a beaker containing an equal volume of cold distilled water. After the solution has been allowed to stand and settle for some time, the clear liquid is decanted for use. This starch solution is sensitive only when fresh and must be prepared each day. If saturated with sodium chloride, or a few drops of chloroform added, it keeps much longer. Various other methods of making a permanent starch solution have been suggested, but the solution made fresh each day seems to be the most satisfactory. If the starch is added to a solution containing a considerable amount of free iodine, a compound of starch and iodine is pro- duced which is decomposed very slowly by the thiosulphate. The indicator is therefore added only Avhen nearly all of the free iodine has been removed by the sodium thiosulphate. Potassmm iodide as well as free iodine must be present to produce the blue color with starch. As solutions of potassium iodide are not stable, this salt must be kept in the solid form and only freshly made solutions used. EXERCISE 64. Preparation and Standardization of N/io Iodine and Sodium Thiosulphate Solutions. 485. Sodium Thiosulphate Solution. — If recrystallized and carefully dried sodium thiosulphate is at hand, weigh out carefully 24.83 grams, dissolve in water, and dilute the solution to a liter. If the ordinary C.P. salt is used, weigh out about 25 grams for a liter of solution. In either case use distilled water which has been thoroughly boiled to expel the carbon dioxide and place the solution in a bottle, through the stopper of which a siphon passes for withdrawing the solution. Pour enough naphtha over the solution to form a layer about 1 cm. deep, or put a straight calcium chloride tube in the second hole of the stopper. Fill the tube with fresh soda-lime. 340 VOLUMETRIC METHODS. 486. Iodine Solution. — Weigh 12.697 grams of recently sublimed iodine in a weighing-bottle. Keep the stopper tightly closed, except when putting in or taking out iodine. Only glass, porcelain, or platinum should come in contact with the iodine. Be careful not to spill crystals on the balance-pan or in the case. About 18 grams of pure potassium iodide are weighed out and dissolved in about 150 c.c. of water. The potassium iodide should be tested by dissolving about 1 gram in 10 c.c. of water and acidifying with dilute hydrochloric acid. The solution must be colorless and remain so on the addition of a few cubic centimeters of starch solution. The iodine is transferred to a liter flask by removing the stopper of the weighing-bottle and inserting the bottle into the neck of the flask, which for this purpose is held in a horizontal position. On raising the flask to an upright position and tapping the weighing-bottle, the bulk of the iodine will be transferred to the flask. The few remaining crystals are removed by adding a little of the potassium iodide solution and shaking the stoppered weighing-bottle until dissolved. The solution is poured into the flask and the weighing-bottle and stopper are thoroughly rinsed with the potassium iodide solution. The remainder of this solution is added to the liter flask, which is stoppered, and shaken at intervals until all of the iodine is dissolved. If the last of the iodine crystals dissolve slowly, a few crystals of potassium iodide may be dropped into the flask and the concentrated solution formed left imdisturbed in contact with the iodine crystals. The solution may also be very gently warmed on the water-bath. 487. Comparison of Iodine and Sodium Thiosulphate Solutions. — When all of the iodine has been dissolved, the solution is diluted to the mark and transferred to a glass-stoppered bottle, which is best kept in a cool, dark place. The solution is immediately compared with the sodium thiosulphate solu- tion. 25 c.c. of iodine solution is measured out into a beaker and the thiosul- phate solution added until the color has been almost entirely removed, leaving the solution a light yellow. A few cubic centimeters of the starch solution are then added and the addition of the thiosulphate solution continued until the intense-blue color is just removed. Until experience is gained, it is advisable to add iodine solution drop by drop until the blue color is restored before reading the burette. There is some danger of over- stepping the mark when the end-point is obtained by removing the blue color with the thiosulphate solution. STANDARDIZATION OF THE SOLUTIONS. 488. First Method. Barium Thiosulphate. — Weigh out several portions of pure barium thiosulphate of about J gram. If exactly 0.6688-gram por- tions are weighed, each will be equal to 25 c.c. of tenth-normal iodine solu- tion. Transfer the portions to beakers and add from 50 to 100 c.c. of water. Without waiting for the salt to dissolve completely, add the iodine solution IODINE SOLUTION. 341 slowly with constarxt stirring until nearly all of the thiosulphate is dissolved. Starch solution is then added and then iodine solution until a slight perma- nent blue color is produced. 489. Second Method. Arsenious Oxide. — Weigh out 4.95 grams of pure arsenious oxide. Place in a beaker and add about 13 grams of caustic soda. Dissolve in 50 c.c. of distiUed water, transfer to a liter flask, and saturate the solution with carbon dioxide. At first the delivery-tube must not be immersed in the hquid, as the latter might be sucked up the tube by the complete absorption of the gas. When the gas-bubbles pass through the Hquid undiminished in size, the latter is saturated. Rinse the dehvery- tube and dilute the solution to the mark. This solution will be exactly decinormal. Titrate 25-c.c. portions with the iodine solution. 490. Third Method. Acid Potassium lodate. — Weigh out 0.3251 gram of acid potassium iodate. Dissolve in 50 c.c. of water, heating gently if necessary. Transfer the solution to a 100-c.c. flask, rinsing the beaker care- fully, and dilute to the mark. This solution will be exactly decinormal. Measure out 25 c.c. into a beaker, add about 1 gram of potassium iodide dissolved in a little water and a few cubic centimeters of dilute hydro- chloric acid. Titrate immediately mth the thiosulphate solution. 491. Fourth Method. Resublimed Iodine- — Weigh out about 0.3 gram of recently sublimed iodine, using a tightly stoppered weighing-bottle. Add 10 c.c. of a 10% solution of potassium iodide and shake until the iodine is completely dissolved. Transfer the solution to a beaker and rinse out the weighing-bottle with water and, if necessary, a little potassium iodide solu- tion. Titrate immediately with the thiosulphate solution. 492. Calculation of the Results. — The iodine solution, having been made from pure iodine, must be considered exactly tenth-normal when made, unless some iodine was known to be lost. The titration of the tliiosulphate solution immediately after the iodine solution was made is considered as a standardization of the former and is so calculated. The average of the titra- tions which agreed mthin Isss than 0.2 or at most 0.3% is taken. If duplicate titrations agreeing within this limit are not readily obtained, some error is being made in the work, which must be discovered by varying the method of procedure. If, for instance, 24.05 c.c. of the thiosulphate solution is found to be equal to 25 c.c. of the iodine solution, the strength of the solu- tions being inversely as the number of cubic centimeters of each used, the strength of the thiosulphate solution wiU be given by the follo\ving pro- portion: 24.05 : 25.00 : : O.IN : x. a; = .l0395N. By the first method of standardization the iodine solution is compared with a weighed amount of barium thiosulphate, which is exactly equal to 25 c.c. of tenth-normal solution. By the same proportion as given above, the strength of the iodine solution may be computed. If, for instance, an average of 24.95 c.c. of iodine solution were used, the iodine solution is .1002N by the proportion 24.95 : 25.00 : : O.IN : x. As this value differs 342 VOLUMETRIC METHODS. from tenth-normal by only 0.2%, the iodine solution is considered tenth- normal, since the errors of titration may amount to 0.2%. The results of the second method of standardization are calculated in the same manner. The third as well as the fourth methods being direct standardizations of the thiosulphate solution, its strength is calculated from the results. If several days have elapsed between the standardization of the thiosul- phate solution and its comparison with the iodine solution this comparison must be repeated. The average of the values found for the thiosulphate solution is taken and used in the proportion when calculating the strength of the iodine solution. If, for instance, 25.18 c.c. of the iodine solution was found to be equal to 24.12 c.c. of the thiosulphate solution, and the average strength of this solution to be .1039N, the proportion will be 25.18 : 24.12 : : .1039 : x, from which x, or the strength of the iodine solution, will be .09952N. 493. Analysis of Reducing Agents. — All substances which may be analyzed by iodometric methods may be divided into two classes, according to whether they act as reducing agents towards iodine or as oxidizing agents towards hydriodic acid. The reducing agents are analyzed either by titrating directly with standard iodine solution, or by adding an excess of the iodine solution and titrating back with standard thiosulphate solution. The method of titrating a thiosulphate or arsenious oxide has already been given in the methods of standardization. Aniimonons oxide may be titrated in exactly the same manner as arsenious oxide, tartaric acid or cream of tartar being used to dissolve the anti- monous oxide. The titration of a sulphite will be given in Exercise 66, page 346. The sulphides, which may be completely decomposed with liberation of hydrogen sulphide, may be analyzed in a similar manner, the hydrogen sulphide being passed through excess of standard iodine solution. The follow- ing reaction takes place: H2S-l-l2="2HI + S. The amount of unreduced iodine is then determined. Easily decomposed sul- phides, such as those of cadmium and zinc," the alkali .sulphides, and hydrogen sulphide water, are completely decomposed by adding dilute acid and iodine solution. Other reducing sub- stances which can be determined iodometrically are antimony OXIDE, alkali cyanides, STANNOUS CHLORIDE, and OILS and FATS, as will be given in the chapter on this subject. lODOMETRIC METHODS. 343 494. Oxidizing Substances, on the contrary, are treated with an excess of potassium iodide, and the solution acidified. The hydriodic acid formed is oxidized to iodine, which combines with the excess of potassium iodide and is then titrated with standard thiosulphate solution. The standardization of the thiosulphate solution by means of acid potassium iodate is an illustration of this method. In this manner free chlorine, bromine, hypo- chlorites, BROMATES, FERRIC CHLORIDE, NITROUS ACID, HYDROGEN PEROXIDE, and pbrsulphuric acid may be analyzed. Oxidizing substances decomposed with difficulty, such as chlorates, chro- MATES, and minerals containing manganese dioxide, may be analyzed by treatment with concentrated hydrochloric acid and distillation of the chlorine which is absorbed in potassium iodide solution. The iodine liberated is then titrated with standard thiosulphate solution. 495. Antimony may be determined by first oxidizing with potas- sium chlorate (1 gram) in strong hydrochloric-acid solution. The excess of chlorine is boiled out, the solution cooled, and excess of potassium iodide added (1 gram). The antimony is reduced from SbzClj to SbjClj with liberation of iodine which is titrated with thiosulphate solution. Copper and iron must be absent from the solution. 496. Copper may be determined in a similar manner. The nitric-acid solution of the copper is neutralized with ammonia or sodium carbonate and then acidified with acetic acid. Ten c.c. of a 10 per cent solution of potassium iodide is added to the copper solution, which should not exceed 15 c.c. in volume. Iodine is liberated according to the following reaction : CuCC^HgO^)^ + 2KI = Cul + 1 + 2KC2H3O2. The iodine liberated is titrated with sodium-thiosulphate solu- tion. Lead, bismuth, arsenic, antimony, and iron should be absent from the solution. 497. Acids may also be determined iodometrically by using a solution of potassium iodate and potassium iodide in the propor- tion of 1 atom of the former to 5 atoms of the latter. On acidi- fying such a solution, the following reaction takes place : KIO3 +5KI + 6HC1 = 6KC1 +61 H-SH^O. 344 VOLUMETRIC METHODS. By titrating the iodine liberated, the amount of acid added may be computed, since 1 atom of iodine is liberated by 1 molecule of a monobasic acid. Bases may also be determined by adding an excess of acid and titrating the iodine liberated after addi.tion of the iodate and iodide solution. In this manner the extremely sensitive indicator of iodometric titrations may be used in acid titrations. Although in this manner a much sharper end-point may be obtained than by the use of the ordinary indicators for acids and bases, the accuracy of a given titration will be no greater unless the measurement of volumes and the purification of reagents is carried out mth the same high degree .of accuracy. 498. Bunsen's Distillation Apparatus. — For canying on the distillation of the chlorine from oxidizing agents, such as manga- nese dioxide, various forms of apparatus have been devised. One of the best is that of Bimsen, shown in Fig. 57. The oxidizing Fig. 57. substance to be analyzed is placed in the small bulb a, ^vhich is connected with the tube b by means of a ground-glass joint over which a rubber tube of suitable size is stretched. This joint must be tight, so as to prevent loss of chlorine, which must not, however, be allowed to come in contact with the rubber. The receiver c is half filled with 10% potassium iodide solution, which is decom- j)osed by the chlorine with liberation of iodine. The chlorine is driven out of the flask a and the tube h with the excess of hydro- chloric acid and the steam produced by boiling the solution. A ANALYSIS OF MANGANESE DIOXIDE. 345 slow stream of carbon dioxide may be produced by placing a few pieces of magnesite in the flask a. This is not necessary, however. 499. Mohr's Distillation Apparatus.— Another form of apparatus which was devised by Mohr is shown in Fig. 58. In this apparatus the joints are made by cork stoppers coated with parafhne, which Fig. 58. is not acted on by chlorine. The receiver is a tube about 30 cm. long and 3 cm. wide, which is immersed in a cylinder filled with cold water. The distillation-flask b should have a capacity of 50 to 75 c.c. EXERCISE 65. Determination of Available Oxygen in Pyrolusite. Weigh out between 0.2 and 0.25 gram of the finely powdered pjrrolusite and introduce into the distillation-flask of either of the forms of apparatus already described. Use a small glass tube for introducing the ore into the flask, obtaining the weight of the material by weighing the tube before and after introducing the ore. About 1.5 grams of potassium iodide are dissolved in 50 c.c. of distilled water and placed in the receiver, which must be kept cool by immersion in ice-water. 20 to 30 c.c. of concentrated hydrochloric acid are introduced into the distillation-flask, which is immediately con- nected with the delivery-tube and the latter immersed in the potassium 346 VOLUMETRIC METHODS. iodide solution. The parafRne-coated cork at a is loosely inserted into the receiver. The distillation-flask 6 is now gently heated until the manganese dioxide is completely dissolved, leaving a white residue which is not affected by further digestion. The solution is finally brought to a boil, and while the Eunsen burner is still under the distillation-flask, the delivery-tube is with- drawn from the receiver and rinsed with distilled water. The iodine solu- tion is transferred to an Erlenmeyer flask and the receiver washed -with water. A few cubic centimeters of potassium iodide solution are placed in the receiver, into which the delivery-tube is again passed, and the contents of the distillation-flask again boiled for a few minutes. If no more iodine is liberated in the potassium iodide solution, the decomposition of the manganese dioxide is complete, othermse the distillation must be continued for some time, the contents of the receiver being finally added to the main solution. The iodine liberated is titrated with standard thiosulphate solution. 1 c.c. of N/10 sodium thiosulphate is equal to .004350 gram of MnOj or .003545 gram of chlorine. Compute the percentage of MnOa as well as the percentage of chlorine hberated. EXERCISE 66. Determination of Sulphur Dioxide in Sodium Sulphite. Weigh out 1 gram of the sulphite and place in an Erlenmeyer flask. Add 100 c.c. of N/lt) iodine solution all at once. It is advisable to measure out the iodine solution with a pipette or a burette into a small beaker. The solution may then be poured from the beaker into the flask. Shake the flask until the salt is dissolved. Rinse the remainder of the iodine solu- tion in the beaker into the flask and titrate the excess wth standard thio- sulphate solution. 1 c.c. of N/10 iodine solution is equal to .003203 gram of SO2 or .01261 gram of NajSOs.yHjO. Compute the percentage of sulphur dioxide and also of the crystalhzed salt. 500. The Active Substance in Bleaching-powder is calcium oxychloride Cl-Ca-0-Cl. On adding water, this compound is decomposed into calcium chloride and calcium hypochlorite, Ca(C10)2. The value of the bleaching-powder depends on the amount of hypochlorite present. The result of an analysis is usually expressed in percentage by weight of available chlorine, although in France it is customary to report the analysis in terms of the numiber of liters of chlorine gas measured at 0° and 760 mm., which may be obtained from 1 kilogram of the bleaching- powder. ANALYSIS OF BLEACHING-POWDER. 347 501. Sample.— The analysis may be carried out by a large number of methods, of which two of the most largely used will be given. In order to secure a fair sample, a considerable amount is weighed out and dissolved in a definite volume of water. Meas- ured portions of this solution are taken for analysis. While calcium hypochlorite is quite soluble, bleach ing-powder contains a con- siderable amount of insoluble material. Even when the bleaching- powder is ground to a smooth paste with water, the insoluble material always contains some bleaching material. The insoluble material is, therefore, not filtered off, but is shaken up when por- tions are measured out. A pipette should be used for this pur- pose, so as to obtain a fair proportion of the insoluble material. 502. lodometric Titration. — The amount of available chlorine may be estimated by adding potassium iodide to a portion of the solution of the bleaching-powder and acidifying. The follow- ing reaction takes place: Ca(C10)3+4KI+4HCl = CaCl34-4KCl+2l2+2H20. As some calcium chlorate is usually present, the solution is not made strongly acid, as chlorine would in that case be produced by the decomposition of this salt. For this reason, if hydrochloric acid is employed, it is added in very slight excess. The difficulty is entirely obviated by the use of acetic acid. The iodine liber- ated is then titrated by means of sodium thiosulphate solution. 503. Titration with Sodium Arsenite. — The hypochlorite may also be titrated directly with sodium arsenite solution. To ascer- tain the end-point drops of the solution are touched to starch- iodide paper. When the hypochlorite is entirely decomposed, the starch-iodide paper is no longer made blue. EXERCISE 67. Determination of Available Chlorine in Bleaching-powder. 504. Solution. — Weigh out 7.091 grams of bleaching-powder. As the material is hygroscopic, and also loses chlorine, a weighing-bottle must be used. Transfer the material to a clean porcelain mortar and grind to a smooth paste with the addition of a little water. Add about 75 c.c. of water, mix thoroughly with the pestle, allow to settle a few minutes, and 348 VOLUMETRIC METHODS. decant into a liter flask, leaving the bulk of the bleaching-powder in the mortar. Repeat this operation until all of the material has been trans- ferred to the liter flask, which is then filled to the mark with distilled water. 505. lodometric Titration. — The solution is thoroughly shaken and 50-c.c. portions of the milky liquid are measured out with a pipette and transferred to beakers. About 1 gram of potassium iodide dissolved in a little water is added to one of the beakers. The solution is acidified ■\\'ith acetic acid and the iodine liberated titrated with N/10 sodium thiosulphate solution. Duplicate determinations with the other portions are made in the same manner. 1 c.c. of the tenth-normal thiosulphate solution is equal to .003545 gram of chlorine. If exactly 7.091 * grams of the bleaching- powder was weighed out, the number of cubic centimeters of N/10 thiosul- phate solution used will be equal to the percentage of available chlorine in the bleaching-powder. 506. Titration with Sodium Arsenite. — 50-c.c. portions of the solution of bleaching-powder may also be titrated with the sodium arsenite solution made for standardizing the iodine solution (page 341). Starch-iodide PAPER may be made as follows : One gram of starch is ground to a paste with a httle water and 100 c.c. of boihng water added with stirring. The solution is filtered and to the filtrate 0.1 gram of potassium iodide is added. Strips of filter-paper are dipped in this solution and dried at 40° to 50°. Strips of this paper are moistened Avith distilled water and laid on a porcelain or glass plate. Nearly all the sodium arsenite solution thought necessary is added to the solution of the bleaching-powder. After stirring thoroughly, a drop is taken out and touched to the test paper. The depth of the color indicates the amount of arsenite solution still to be added. When consid- erable hypochlorite is still present, the spot becomes almost black and then colorless again as the iodine is converted into the trichloride ICI3. Several cubic centimeters of the arsenite solution should then be added. When the spot test is only light blue, the arsenite solution should be added by drops until the blue color has entirely disappeared. The calculation is made in the same maimer as with the iodine solution. PROBLEM. Problem 13. A mixture of crystallized sodium thiosulphate, sodium sul- phite, and sodium sulphate was analyzed in the following manner : A gram portion was dissolved in water and titrated with N/10 iodine solution, 49 c.c. being required. The hydriodic acid in this solution was then titrated with N/5 soduim-hychoxide solution, 9.5 c.c. being required. The total amount of sulphur in a half -gram portion was then determined as barium sulphate, of which .821 grams was obtained. Calculate the amount of Na2S203.5H20, NajSOs-yHjO, and NajSOi.lOH^O present in the mixture. * Note for student: Show how this figure is obtained. PRECIPITATION METHODS. CHAPTER XXVI. DETERMINATION OF CHLORIDES, CYANIDES, AND SILVER. 507. Theory of Indicators. — In the volumetric methods already discussed the products of the reactions have been soluble, and the end-point has been obtained by the action of a slight excess of one of the reagents on the indicator. In ACiDiMETRY and alkalimetry the indicator is either an acid or a base, but possessing such a weak chemical affinity that it cannot remain combined until an excess of the appropriate reagent (base or acid) is present. Methyl orange, for instance, is a weak acid having a bright-red color, while its salts are light yellow. If methyl orange is added to a solution containing a free base, the salt of methyl orange and the base will be formed and the yellow color of the salt will be given to the solution. If carbon dioxide in excess is added to this solution, the color remains yellow because the carbonic acid formed is weaker than the methyl-orange acid, and is therefore unable to displace the latter from its combination with the base. If a strong acid like hydrochloric is added to the solution, it remains yellow as long as there is more base present than can combine with the hydrochloric acid added. When a slight excess of the strong acid has been added, the base combined with the weakly acidic methjd orange is taken by the strong acid. The free methyl orange then gives the reddish color to the solution. In precipitation methods, on the other hand, advantage is most frequently taken of the relative solubilities of the salts of a metal to obtain the end-point. The colored salt which indicates the end of a reaction must be a soluble compound, while the ele- ment to be determined must form an insoluble compound. 349 350 VOLUMETRIC METHODS. 508. Silver Chromate as Indicator. — In the determination of chlorides by means of silver nitrate solution, a little neutral potas- sium chromate is added to the solution. On adding the silver nitrate solution the silver is precipitated as silver chloride. When enough silver nitrate has been added to precipitate all of the chlorine, the red silver chromate is formed, which indicates the end of the reaction. As long as chlorides exist in the solution, any silver chromate formed will be decomposed, since if any of this salt is present, it will tend to keep the solution saturated, and any silver which goes into solution as chromate will be imme- diately precipitated as chloride because of the insolubility of the latter salt. Only when chlorides are absent from the solution can silver chromate be permanently formed. As the silver chro- mate is decomposed by even a small amount of free acid, the solution must be neutral or slightly alkaline with sodium car- bonate. 509. Ferric Sulphocyanate as Indicator. — Silver also forms an insoluble salt with sulphocyanic acid. This acid produces an in- tensely red color with solutions of ferric iron. The titration of silver is conducted by adding a standard solution of the sulphocyanate to the silver solution containing some ferric sulphate or nitrate. As long as silver is present in the solution the sulphocyanic acid is precipitated, so that the iron compound which is soluble cannot exist. When all of the silver has been precipitated, the next drop of the sulphocyanate solution forms a little of the red ferric com- pound which indicates the end of the reaction. As the silver sulphocyanate is insoluble in dilute acids, and the ferric sulpho- cyanate is readily formed under these conditions, the titration may be carried on in acid solutions. The titration of chlorides in acid solution may also be carried out by this method which is known as Volhard's. A measured amount of silver nitrate solution which is more than sufficient to precipitate the chloride present is added to the solution. After the addition of some ferric iron as indicator, the excess of silver is titrated back with standard sul- phocyanate solution. Because of the solubility of silver chloride in solutions of the thiocyanates, this method does not give abso- lutely correct results unless the silver chloride is filtered off and the excess of silver determined in the filtrate. TITRATION OF CYANIDES. 351 510. Silver Chloride as Indicator. — Chlorides may also be titrated with silver nitrate solution without the addition of an indicator. After each addition of silver nitrate the solution is thoroughly shaken. The silver chloride readily collects so as to leave a clear solution in which the formation of a white precipitate on the addition of another drop of silver nitrate may readily be observed. The silver chloride does not collect readily unless a considerable amount of chloride is present and the volume of the solution is not too large. 511. Titration of Cyanides. — The alkali cyanides may be titrated with a silver nitrate solution in the same manner as chlo- rides, the end-point being obtained when no more precipitate is formed. The reaction takes place in two stages as represented by the following equations: 2KCN + AgNO, = KCN.AgCN + KNO3; KCN.AgCN + AgN03= 2AgCN + KNO3. Until one-half of the cyanide has been converted into the silver salt, any precipitate of silver cyanide which forms is dissolved by the undecomposed cyanide on stirring the solution. When the reaction represented by the first equation is completed, further addition of silver nitrate forms a permanent precipitate of sih^er cyanide according to the second equation. The precipitate readily collects in flocks on vigorously stirring the solution, leaving a clear liquid in which the absence of a precipitate on the addition of a drop of silver nitrate may be readily observed and serves to indicate the end-point. This titration may also be conducted so that the reaction represented by the first equation is carried out. The silver solu- tion is added until a slight permanent precipitate is formed. If CHLORIDES are present in the alkali cyanides, as is frequently the case, silver chloride will be precipitated in preference to the silver cyanide because of the greater insolubility of the former. The silver chloride reacts with the alkaU cyanide as follows: AgCl -t- 2KCN = KCN. AgCN + KCl. 352 VOLUMETRIC METHODS. When all of the cyanide has been converted into the double cyanide of silver and potassium, the next drop of silver nitrate produces a permanent white precipitate. If chlorides are present, this precipi- tate will be silver chloride, thus forming a more delicate indicator than the more soluble silver cyanide. The amount of cyanide present will also be correctly determined, as is evident from the equation given. If the titration is carried to the second stage, the presence of the chloride introduces a;i error, as all the chloride combines with silver before the end-point is reached. The amount of chloride present may be determined by noting the volume of silver nitrate solution required for the completion of each stage of the titration. The excess of the silver nitrate required for the second part of the titration is that necessary to precipitate the chloride. 512. Strength of Solutions. — The standard solutions used in precipitation methods may be made according to the normal system, a gram molecular weight of silver nitrate, sodium chloride, etc., being dissolved in a liter of a normal solution. Such solutions are convenient for use when several substances must be titrated with the same standard solution. When a solution is to be used exclusively for the titration of a given substance, it is more con- venient to make it of such strength that 1 c.c. is equal to an amount of the given substance which can be represented by a simple number. If a silver nitrate solution is made of such a strength that 1 c.c. is equal to 1 mg. of chlorine, the number of cubic centi- meters of silver nitrate solution used to titrate the chlorine m 1 gram of a given substance will give the per cent of chlorine if divided by 10. 513. Standard Silver Nitrate Solutions are made by dissolving weighed amounts of the crystals and diluting to the required volume. The salt may be purchased in very pure condition, seldom requiring drying even. As pure metallic silver may also be readily obtained the standard solutions are sometimes made by weighing out the pure metal and dissolving in nitric acid. If a neutral solution is required, the excess of nitric acid may be removed by evaporation to drjmess on the water-bath. The last traces of acid are removed by moistening with water several times and evaporating to dryness on the water-bath. If protected from TITRATION OF CHLORIDES. 353 the sunlight and contact with organic matter, solutions of silver nitrate retain their strength for a considerable time unless very dilute. 514- Standardization. — Solutions of silver nitrate may be standardized by titration against weighed amounts of pure sodium chloride prepared as directed in Exercise 6. A saturated solution of potassium chromate is used as the indicator. 4 to 5 drops of the chromate solution are added to the solution to be titrated, the volume of which should not be large. The detection of the first faint tinge of red in the yellow chromate solution offers some difficulties. It is more readily done by gaslight than by sun- light because of the yellow tinge of the former. By daylight the red tinge is more readily detected by a comparison of the color of the solution being titrated, with the color of a solution previously titrated from which the red color has been removed by the addi- tion of a pinch of sodium chloride. The red color is also made more pronounced by looking at the solution through a layer of water tinted with chromate to the same color as the solution to be titrated. It is also advantageous to carry out the titration in a porcelain dish, thus giving a white surface for comparison. 515. Indirect Determinations by Means of Standard Silver Nitrate. — Besides the determination of chlorides, this method may be used to determine indirectly all substances which can be con- verted into neutral chlorides by treatment with hydrochloric acid and evaporation to dryness to expel the excess of hydrochloric acid. The carbonates of the alkalies and the alkaline earths may be treated in this manner to determine the base, and in cases where only the normal carbonate can be present, carbon dioxide may be determined in this manner. The nitrates also of these metals may be converted into chlorides by a similar treatment and determined. 516. Determination of Sodium and Potassium in the Mixed Chlorides. — The amount of potassium and sodium present in a mixture of the chlorides of these metals may be ascertained by a determination of the amount of chlorine present, the weight of the mixed chlorides having first been found. An indirect deter- mination of this kuid gives good results only when there is a con- siderable difference in the atomic weights of the elements, and a large precipitate is weighed. The method is not accurate for a 354 VOLUMETRIC METHODS. small amount of one element in the presence of a large amount of the other. The weight of sodium chloride in the mixture is calculated by multiplying the amount of chlorine found by 2.1032, /Molecular weight KCl \ Atomic weight CI deducting from the product the combined weight of the chlorides and multiplying the remainder by 3.6426: Mol. wt. NaCl Mol. wt. KCl -Mol. wt. NaCl In multiplying the weight of chlorine found by 2.1032, the amount of potassium chloride is found which would be obtained if this salt alone were present. If this product is the same as the weight of the mixed chlorides, only potassium can be present. EXERCISE 68. Determination of Cyanogen in Potassium Cyanide. In order to secure a fair sample of the material, weigh out 5 grams of the potassium cyanide selected from various parts of the stock to be exam- ined. Dissolve in water and dilute to 500 c.c. Withdraw 50-c.c. portions for titration. If a pipette is used for this purpose, a loose plug of cotton moistened with silver nitrate solution should be placed in the upper part of the stem, and great care should be taken not to draw any of the solution into the mouth. The silver nitrate solution is made by weighing out 17 grams of the pure dry crystals, dissolving in water, and diluting to 1 liter. It is advisable to test the distilled water for chloride before adding it to the silver nitrate. Pour the solution into a glass-stoppered bottle and keep in the dark. The 50-c.c. portions of the cyanide solution are diluted with an equal volume of water and the silver-nitrate solution added with constant stirring until a slight permanent opalescence is obtained. 1 c.c. N/10 silver nitrate solu- tion is equal to .01304 gram of potassium cyanide. As commerical potas- sium cyanide is frequently contaminated vfith sodium cyanide, the percentage of potassium cyanide found may be greater than 100%. Tliis wiU not occur if the result is given in percentage of cyanogen. 1 c.c. of N/10 silver nitrate solution is equal to .005208 gram of cyanogen. DETERMINATION OF SILVER. '"^SB 517, Datarmination of Silver in Alloys. — The silver in alloj'S can very readily be titrated by the Volhard method, since the presence of a considerable amount ef free nitric acid does not dis- solve the silver sulphocyanate. Nitrous acid must be absent, however, and that produced by dissolving the alloy must be entirely expelled by boiling the solution. The amount of copper present must not exceed 70%. Alloys in which the percentage of copper is higher, maj- be analyzed by adding an accurately measured volume of standard silver solution or an accurately weighed amount of the pure metal, so that the ratio of copper to silver shall not exceed 7 to 3. Mercury must be removed by ignition. Arsenic, antimony, cadmium, lead, bismuth, tin, and zinc do not interfere with the titration. Cobalt and nickel interfere because of the color of their nitrates. EXERCISE 69. Determination of Silver in a Coin by Volhard's Method, Weigh out 7.5 to 8 grams of ammonium sulphocyanate, dissolve in water, and dilute to 1 hter. Standardize the solution by titration against the tenth-normal silver solution prepared for Exercise 68. Measure out into a beaker 30 to 40 c.c. of the standard silver solution, dilute with 100 to 200 c.c. of water, and add 5 c.c. of a cold saturated solution of ferric alum. The ferric iron solution must be free from clilorine, and if turbid, a few cubic centimeters of dilute nitric acid may be added. Place the beaker containing the silver on a white surface and add the sulpho- cyanate solution with constant stirring until a faint permanent reddish tinge is given to the solution. Repeat the titration until duphcates are obtained. As the sulphocyanate solution is permanent, it is advisable to dilute it to exact strength. When it is used to determine silver, it is mo.^t convenient to dilute it to such a strength that 1 c.c. w.ll be equal to exactly .010 gram of silver. Having found, for instance, that 30 c.c. of the sulpho- cyanate solution is equal to 32 c.c. of the N/10 silver nitrate solution, the amount of silver equivalent to the 30 c.c. of the sulphocyanate solution is foimd by multiplying 32 by .010793, which gives .3454 gram of silver. As it is desired to dilute the solution so that 30 c.c. shall be equal to .300 gram, the volume to be diluted to 1000 c.c. is found by the proportion .300 : .3454 : : x : 1000. 356 VOLUMETRIC METHODS. As a; = 868.58, this number of cubic centimeters is diluted to 1 liter. The strength of the solution after dilution is verified by again titrating against the standard silver solution. Thoroughly clean a dime or other silver coin by rubbing with alcohol ammonia, etc. Weigh it carefully and dissolve in 20 c.c. of a mixture of equal parts of nitric acid and water. For this purpose place the coin in a beaker, cover with a watch-crystal, and warm on the water-bath until it is entirely dissolved. Remove the watch-crystal after rinsing it with water and continue warming the solution until the nitrous fumes are entirely expeUed. Transfer the solution to a 250-c.c. flask, rinsing the beaker thor- oughly. Dilute the solution to the mark and shake thoroughly. Take out 50-c.c. portions with a pipette, dilute with 100 to 200 c.c. of water, add 5 c.c. of the ferric iron solution, and titrate with the sulphocyanate. Repeat the titration until duplicates are obtained. Calculate the per cent of silver in the coin. PROBLEMS. Problem 14. The combined weight of sodium and potassium chlorides was 2.5685 grams. 1.432 grams of chlorine was pre.sent. Calculate the percentage of potassium and sodium chloride present. Problem 15. The combined weight of the sodium and potassium chlorides was 2.032 grams. When converted into sulphates, the weight was 2.385 grams. Calculate the percentage of sodium and potassium present. CHAPTER XXVII. DETERMINATION OF PHOSPHORIC ACID. A GOOD many methods of determining phosphoric acid volu- metrically have been devised. None of these methods give as reUable and accurate results as the gravimetric method. The great amount of labor necessary and the great length of time which must elapse from the beginning to the end of the gravimetric determination render the use of the volumetric methods very desirable. Three of the best of these methods will be described. THE URANIUM METHOD. This method is applicable to the titration of the phosphoric acid in the alkali and alkaline earth phosphates. Iron, and espe- cially ALUMINIUM, if present in considerable amount, interfere seriously with the titration. The method is applicable to most of the material used for fertilizers, such as bones, bone-ash, soluble phosphates, phosphorite, etc. The method is based on the fact that when uranium nitrate or acetate is added to a neutral solution of a phosphate all of the phosphoric acid is thrown down as yellow uranyl phosphate, UOzHPO^. Free mineral acid must be absent, but the solution may be moderately acid with acetic acid. A very delicate indicator for the excess of the uranium is found in the reaction with potassium ferrocyanide, which gives a decided brown tinge with very small amounts of uranium. Unfor- tunately, the method is not an absolute one, but gives results which vary with the volume of the solution, the salts present, etc., so that the standardization of the solution must be made under as nearly as possible the same conditions as the determination to be carried out. The standardization and determination should be 357 358 VOLUMETRIC METHODS. carried out by the same individual. To complete the precipitar tion of the phosphoric acid, the solution must be boiled after each addition of the standard solution. Some calcium phosphate is carried down with the uranium phosphate, so that if a solution containing calcium is to be titrated, the uranium solution must hfe standardized by calcium phosphate, as the alkali phosphates are not carried down in this manner. Several titrations must also be made, in the first adding the uranium solution in small amounts until the end-point is reached. In the second titration nearly all of the uranium solution is added at once. 518. Standard Solution. — The uranium solution is made by dissolving about 35 grams of uranium nitrate or acetate in water and diluting to a liter. 3% grams of ammonium or sodium acetate are added to the nitrate solution to neutralize any free acid present. The addition of 50 c.c. of pure glacial acetic acid to either the nitrate or the acetate solution renders it more stable on exposure to light. The nitrate may usually be obtained in a purer condi- tion than the acetate, while the addition of an alkali acetate to the latter is unnecessary. The solution is allowed to stand for several days and is then filtered, or the clear liquid drawn off. 519. Standardization. — For standardizing the solution a pure salt or a solution of a phosphate in which the phosphoric acid has been accurately determined must be available. Disodium phos- phate and MiCROCOSMic salt may be recrystallized until absolutely pure, although some Uncertainty always exists as to the amount of water present in the dried crystals. Dihydric potassium PHOSPHATE has the advantage that it contains no water of crystal- lization. If these pure salts are not at hand, a solution may be made containing approximately 0.1 gram P2O5 in 50 c.c. 100-c.c. portions of this solution may be measured out and the phosphoric acid present determined by precipitation with magnesia mixture and weighing as magnesium pyrophosphate. 520. Titration. — 50 c.c. of the standard phosphate solution is measured out into a beaker and heated to 90° or 100°. The uranium solution is added until a drop taken out and placed on a white porcelain surface gives a brown ring when the centre of the drop is touched with a glass rod which has lieen dipped in a freshly made solution of potassium ferrocyanide. This solution DETERMINATION OF PHOSPHORIC ACID. 359 is made by dissolving about ^ gram of the salt in 20 c.c. of water. After each addition of uranium the solution is brought to a boil. The color of the spot test increases slowly on standing. The worker must therefore note the depth of color after a fixed interval of time. Uniformity in this respect is most readily secured by always observing the color after an additional drop has been added to the solution being titrated and the solution brought to a boil. When the end-point is reached, this drop will be in excess, but a cor- rection may easily be made to the reading after noting the change in the reading of the burette produced by letting one drop fall from the burette. A second and third titration are made, adding nearly the entire amount of uranium solution at once. The ura- nium solution is then diluted so that 1 c.c. is equal to .005 gram of P2O5. 521. Standardization with Calcium Phosphate. — ^When the ura- niimi solution is to be used for the titration of phosphoric acid in the presence of calcium, it must be standardized by means of a solution of tricalcium phosphate. About 5.5 grams of the pure material is weighed out and dissolved in a little dilute nitric or hydrochloric acid. It is precipitated by adding a slight excess of ammonia and redissolved in a moderate excess of acetic acid The solution is then dUuted to a liter. The phosphoric acid in 50- c.c. portions of this solution is determined by precipitation as molybdate, dissolving in ammonia, and reprecipitating with mag- nesia mixture and weighing as magnesium pyrophosphate. The titration of 50-c.c. portions of this solution with the uranium solution is conducted as already given. 522. Titration of Phosphates. — ^The analysis of phosphates which are free from more than traces of iron and aluminium is made by dissolving weighed portions, so that 50 c.c. shall con- tain about 0.1 gram P2O5. If the material is dissolved in the strong mineral acids, the solution must be neutralized with anmionia and then acidified with acetic acid. 50-c.c. portions are then titrated exactly as directed for the standardization. When considerable iron or aluminium is present, the phosphoric acid must be separated from these elements by precipitation as phosphomolybdate, which is dissolved in ammonia, and the phos- phoric acid again precipitated with magnesia mixture. The mag- •360 VOLUMETRIC METHODS. nesia precipitate is then dissolved in a little nitric or hydrochloric acid, the solution neutralized with ammonia and then acidified with acetic acid. The phosphoric acid is then titrated with the uranium solution. TITRATION OF THE PHOSPHOMOLYBDATE. Two volumetric methods of determining phosphoric acid have been based on the properties of the phosphomolybdate precipi- tate. By one of these methods, that of Pemberton,* the molybdic acid in this precipitate is titrated by means of standard caustic soda or potash with phenolphthalein as the indicator according to the following equation : 2(NH,)3P04.12Mo03+46KOH= 2(NH4)2HP04+ (NH4)2Mo04+ 23K2M0O4+ 22H2O. By the other methodf the molybdic acid is reduced by means of zinc to molybdic oxide according to the equation: 2M0O3+ 6H = MoA+ SH^O. The molybdic oxide is then oxidized by means of standard potas- sium permanganate solution. 523. Pemberton's Alkalimetric Method. — This method is appli- cable to the titration of the phosphates of the alkalies and the alka- line earths, but the results are not so good when considerable quantities of iron or aluminium are present. In the presence of large amounts of these metals, it is advisable to redissolve and reprecipitate the phosphomolybdate. As sulphuric acid may be carried down with the precipitate, it should be absent from the solution. If present it may be removed by means of barium chloride. 524. Standard Solutions. — ^A standard solution of nitric acid and one of caustic soda or potash free from carbon dioxide are required. The solutions are made of such a strength that 1 c.c. is equal to .001 gram of P2O5. Since 46 molecules of caustic soda or potash are equal to 1 molecule of phosphorus pentoxide, a liter of "" * Jour. Am. Chem. Soc, 1894, '278. ' ~ t Blair, The Chemical Analysis of Iron, 1901, p. 104. DETERMINATION' OF PHOSPHORIC ACID. 361 normal solution will be equal to l/46th of the molecular weight of PjOi, or 3.087 grams. A solution of convenient strength will be .3239 normal, 1 liter of such a solution being equal to 1 gram of PjOj. The nitric acid may be made by diluting 21 c.c. strong nitric acid (sp. gr. 1.40) to 1 liter. The caustic soda solution may be made by adding to 100 grams of pure caustic soda an amount of water just insufficient to dissolve it. After allowing the solution to settle in a tall covered vessel, 17 c.c. of the clear liquid is with- drawn with a pipette and diluted to 1 liter. Most of the sodium carbonate remains undissolved. If caustic potash is used, about 20 grams are weighed out and diluted to a liter. The carbon dioxide present must be precipitated by means of barium hydroxide or chloride. Both the caustic soda and caustic potash solutions must be protected from the carbon dioxide of the air. 525. Standardization. — The solutions may be standardized by any of the methods given in Chapters XXI and XXII. The acid must be compared with the alkali, using phenolphtha'ein as the indicator. A much better method of standardization consists in precipitating a known amount of phosphoric acid by means of the molybdate solution and titrating the precipitate in exactly the same manner as in the analysis of an unknown phosphate. 526. Precipitation of the Phosphoric Acid. — The molybdate solution is made according to the directions given on p. 507. For washing the precipitate, a dilute solution of nitric acid is used, a solution made by adding about 15 c.c. concentrated acid to 1 liter of water being suitable. The acid is washed out by means of a potassium nitrate solution made by dissolving 1 gram of potassium nitrate in 1 liter of water. The phosphate must be free from organic matter. If present, it is destroyed by ignition of the material in a platinum dish or crucible. If the percentage of phosphoric acid is large, a solution should be made by dissolving a weighed amount in nitric acid and diluting to a known volume. Portions are then measured out containing not more than .050 gram P2O5. The solution is nearly neutralized with ammonia, warmed to about 40°, and 40 c.c. of the molybdate solution added. The precipitation is best made in an Erlenmeyer flask of about 250 c.c. The solution is allowed to digest for ten to fifteen minutes, the temperature of about 40° 362 VOLUMETRIC METHODS. being maintained by warming on the water-bath. By closing the flask with a rubber stopper and shaking about five minutes, com- plete precipitation will be insured. Allow the precipitate to settle for a few minutes and decant the solution through a small filter- paper. Transfer the precipitate and wash with dilute nitric acid until free from molybdenum as indicated by the absence of a pre- cipitate with ammonium sulphide. The precipitate is then washed with the potassium nitrate solution until free from. acid. The washing must be thorough, but as little wash-water as possible must be used, because of the slight solubility of the precipitate. If properly done, 200 c.c. will be found sufficient. 527. Titration. — The paper and the precipitate are placed in a beaker and a measured amount of the standard caustic soda or potash solution sufficient to dissolve the precipitate is added. 10- c.c. portions may be added, and if on agitation the precipitate does not dissolve, another 10-c.c. portion is added. Continue the addi- tion of 10-c.c. portions until the precipitate is completely dissolved. Add a few drops of phenolphthalein and titrate the excess of alkali with the standard nitric acid. The amount of phosphoric acid present is calculated from the number of cubic centimeters of caustic soda used. REDUCTION OF THE MOLYBDATE PRECIPITATE WITH ZINC AND TITRATION WITH PERMANGANATE SO- LUTION. 528. Standard Solution. — A standard solution of potassium permanganate will be required for this method. As 1 molecule of PjOj is combined with 24 molecules of M0O3, which will require 36 atoms of oxygen to reoxidize after being reduced to M02O3, we have the relation Mol. wt. P205(142) =Mol. wt. 36 0(576). A liter of a normal solution of potassium permanganate will therefore be equal to 1.9722 gram of P2O5. From this figure the value of a fifth- or tenth-normal solution may be computed. If standardized by means of iron, the value of the solution in phosphorus pentoxide may be calculated directly from the factor .03528, which is the equivalent of 1 gram of iron in phosphorus pentoxide. The best method consists in reducing and titrating the molybdate precipitate from a known amount of phosphoric acid. DETERMINATION OF PHOSPHORIC ACID. 363 S29. Titration. — The precipitation of the phosphate with the molybdate solution is carried out exactly as given for the preceding method. Instead of washing with dilute nitric acid and potassium nitrate, a solution of acid ammonium sulphate is used. This is made by adding to 1 liter of water 15 c.c. of strong ammonia (sp. gr. 0.90) and 25 c.c. of concentrated sulphuric acid (sp. gr. 1.84). After washing the precipitate until it is free from the molybdenum solu- tion, as indicated by the test with ammonium sulphide, it is dis- solved in dilute ammonia. During the washing, the phospho- molybdate need not be entirely removed from the flask in which it was precipitated. The ammonia solution of the precipitate is allowed to flow into the flask together with the water used to wash the paper. 5 grams of pulverized zinc and 15 c.c. of concentrated sulphuric acid are now added. The flask is closed with a stopper through which passes a glass tube bent at a slightly acute angle and dipping into a saturated solution of sodium bicarbonate. When all of the zinc is dissolved the solution is titrated with the standard potassium permanganate solution. The solution before titration should be green in color. If it is brown the determi- nation should be rejected. A blank should be made in which the same amount of ammonia, zinc, and sulphuric acid is added to enough distilled water to make the volume of the solution the same as that in the titration of the molybdate precipitate. After the zinc is dissolved, the solution is titrated with the standard permanganate and the amoimt added is subtracted from the vol- imie used in the determination of phosphorus. TECHNICAL ANALYSIS. CHAPTER XXVIII. ANALYSIS OF IRON AND STEEL. Iron and steel as manufactured to-day are quite complex substances, containing almost invariably, besides iron, the ele- ments CARBON, SILICON, SULPHUR, PHOSPHORUS, and MANGANESE, while special varieties may contain nickel, chromium, manganese, etc. 530. The Metals are undoubtedly present in the conditions in which they usually exist in alloys. The analysis for the determina- tion of the amoimt of the metals present other than iron is carried out by the methods already given for the analysis of alloys. 531. The Acid Elements are for the most part combined directly with the iron as carbide, silicide, sulphide, and phosphide. As these elements are usually met with in combination with oxygen, the ordinary methods of analysis are not always suitable for their determination in iron and steel. The modification usually consists in adopting a method of dissolving the metal by which the acid element is certain to be thoroughly oxidized. Solution of iron or steel in a non-oxidizing acid like hydrochloric results in the evo- lution of more or less of the acid elements present as hydrogen compounds, such as hydrogen sulphide, phosphine, and various hy- drocarbons. Silicon even is liable to be volatilized as silicon tetra- chloride. When solution is effected with nitric acid, silicon, phos- phorus, and sulphur are converted into non-volatile oxygen com- pounds, while more or less of the carbon is lost as carbon dioxide. For the determination of carbon, the iron is dissolved in solutions of cupric salts or the carbon dioxide evolved on dissolving the metal in an oxidizing solution is collected and determined. The determination of these elements must be carried out with unusual care, since a very slight variation in the percentage 364 SILICON. 365 produces a very considerable variation in the properties of the metal. 532. To Obtain a Sample for Analysis a perfectly clean drill is taken, and by means of a drill-press or an ordinary brace, holes are bored in various parts of the casting or bar of iron or steel. Loose dirt, sand, etc., must be removed from the surface so as not to fall into the drillings. If the sm-face cannot be cleaned perfectly the first drniings should be discarded. The remainder of the drillings may then be collected on a piece of clean glazed paper. The iron may also be separated from sand and dirt by means of a magnet. Oil may be removed by washing a few times with ether. The drill- ings are usually fine enough for most purposes. If not, they may be crushed in an agate or hardened-steel mortar. In no case may the fine portions be taken and the large pieces discarded. A fair sample of both portions, must be taken and further pulverized if necessary. DETERMINATION OF SILICON. 533. Solution of the Iron in Nitric Acid. — One to five grams of the drillings are placed in a beaker of 200- to 300-c.c. ca- pacity and 40 c.c. of a mixture of 20 c.c. concentrated nitric acid and 20 c.c. water are added. The action is generally quite violent accompanied by frothing, so that to prevent loss it should be moderated by placing the beaker in cold water. If action does not begin as soon as the acid is added, the beaker may be gently warmed and cooled when the action becomes violent. It may also be moderated by adding the acid in several portions. When the action has ceased, the beaker is warmed on the hot-plate to dissolve the few remaining portions of metal. Solution may be hastened by the addition of a little hydrochloric acid. The solu- tion is evaporated to dryness and the residue heated in the air-bath xmtil nitrous fumes are no longer evolved. The residue is then dissolved in 30 c.c. hydrochloric acid by warming gently. The solution is diluted somewhat, warmed, and the silica filtered off on an ashless paper and washed. Dilute hydrochloric acid is first used and then water. 534. Volatilization of Silica. — The moist paper is transferred to a weighed platinum crucible, the paper burned, and the precipi- 366 ANALYSIS OF IRON AND STEEL. tate ignited, finally over the blast-lamp, and weighed. ■ Even if the silica is perfectly white, it is necessary to test its purity by add- ing a few drops of concentrated sulphuric acid and then enough pure hydrofluoric acid to dissolve the silica. The solution is then evaporated to dryness and the residue ignited and weighed. The weight of the residue is subtracted from the weight of the impure silica and the amount of silicon calculated. 535. Drown's Method. — A more rapid method which is very much used for pig iron and which gives excellent results is based on the fact that the silicic acid may be dehydrated by heating with concentrated sulphuric acid. The drillings are dissolved in nitric acid as already described: 20 c.c. of sulphuric acid consisting of equal parts of concentrated acid and water are then added for each gram of drillings taken. The solution is heated until copious fumes of sulphuric acid are given oS. The lumps of ferric sulphate must be broken up with a rod and the solution stirred to prevent bumping and spattering of the material. The concentrated acid is allowed to cool somewhat, cautiously diluted with water, and the solution warmed gently until the ferric sulphate is completely dissolved. The hot solution is filtered immediately, as the silica dehydrated in this manner gradually goes into solution again. The silica is washed first with water and then with dilute hydro- chloric acid and finally with water. It is ignited and weighed, and then volatilized as directed in the first method. This method is used when silica alone is to be determined. 536. Solution of the Iron in Hydrochloric Acid. — The silica may also be determined after solution of the iron in hydrochloric acid and potassium chlorate. The evaporation of the solution to dry- ness may then be accomplished more rapidly, but a slight loss of silicon as SiCl4 is almost inevitable. 537. Fusion of Silica Precipitate. — Instead of ascertaining the amount of impurity in the silica by treatment with hydrofluoric and sulphuric acids, it may be fused with four or five times its weight of sodium carbonate. The fused mass is dissolved in water, acidified with hydrochloric acid, evaporated to dryness, redissolved in hydrochloric acid and water, and the silica filtered off, washed, ignited, and weighed. The purification of the silica by volatiliza- SULPHUR. 367 tion with hydrofluoric acid is more convenient if a pure acid is at hand. DETERMINATION OF SULPHUR. 538. Solution of the Iron in Nitric Acid. — Sulphur may be deter- mined in the filtrate from the silica by precipitation with barium chloride and weighing as barium sulphate. The iron must be dis- solved by the first method given, but concentrated nitric acid must be used in place of the acid of sp. gr. 1.2 in order to insure com- plete oxidation of the sulphur. If the metal dissolves slowly, a few drops of hydrochloric acid may be added. As the percentage of sulphur is usually small, the large amount (5 grams) should bo used. After solution has been effected a little sodium carbonate is added to prevent loss of sulphur trioxide, and the solution is evaporated to dryness. The residue is dissolved in hydrochloric acid, and again evaporated to dryness to completely remove the nitric acid and dehydrate the silicic acid. The residue is again dissolved in hj^drochloric acid, and the excess expelled by evapora- tion. The solution must not be allowed to go to dryness, as a crust of insoluble basic ferric chloride would be produced. The solution is diluted and the silica and graphite filtered off and washed. The filtrate is heated to boiling, and 10 c.c. of a 10% solution of barium chloride added slowly and with constant stirring. The solution is allowed to stand overnight in the cold. The precipitate is then filtered off and washed with very dilute hydrochloric acid until free from iron, and then with water. The barium sulphate is ignited and weighed in the usual manner. 539. Errors. — This method tends to give a low result on account of the solubility of barium sulphate in hydrochloric acid and ferric chloride. The result is apt to be high on account of the great diffi- culty of obtaining a precipitate of barium sulphate which is free from iron. The iron carried down seems to be present as a basic ferric sulphate which loses SO3 on ignition and leaves the barium sulphate colored red by the ferric oxide. Fusing the precipitate with sodium carbonate, dissolving in water, and filtering effects a separation from the iron, but 'the barium sulphate obtained from the filtrate will not represent the sulphur present in the iron because 368 ANALYSIS OF IRON AND STEEL. of loss of SO3 during the first ignition. The presence of considerable amounts of hydrochloric acid in the iron solution tends to prevent the precipitation of the iron with the barium sulphate, but another error is introduced because of the marked solubility of the latter in the dilute hydrochloric acid and the ferric chloride. The large excess of barium chloride used tends to produce complete precipi- tation of the sulphuric acid. The success of the method therefore "depends on regulating the amount of hydrochloric acid so as to prevent the precipitation of iron, but not that of the barium sul- phate. The so-called evolution methods are therefore generally preferred when applicable. In steel the sulphur is generally so combined that the evolution method may be used, but the evolution of the sulphur is not complete in all pig and cast irons. 540. For Pig Irons the following method is used: After solu- tion of the pig iron in strong nitric acid, 10 grams of sodium car- bonate are added, and the solution transferred to a platinum dish, evaporated to dryness, and the residue ignited. It is then digested with water, a little sodium carbonate added, and the solution filtered. The insoluble material is washed with water containing sodium carbonate. The filtrate is acidified with hydro- chloric acid and evaporated to dryness to dehydrate the silica, which is filtered off and washed. The sulphuric acid in the filtrate is precipitat 3d with barium chloride and weighed as barium sul- phate. 541. Determination of Sulphur in the Reagents. — All of the reagents used in sulphur determinations must be free from sulphur. Their purity is most easily assured by making a blank determina- tion as follows: The total amount of nitric and hydrochloric acid used is measured out into a beaker, a little sodium carbonate added, and the solution evaporated to drjmess. The residue is dissolved in a little water, a few drops of hydrochloric acid added, and after heating to boiling a few cubic centimeters of barium chloride solution are added. Any precipitate formed is filtered off, washed, ignited, and weighed, and the amount of barium sulphate obtained deducted from the amount obtained from the solution of iron. SULPHUR. 369 EVOLUTION METHODS FOR SULPHUR. The evolution methods are now more largely used than the oxidation methods. In the evolution methods advantage is taken of the fact that when iron or steel is dissolved in hydrochloric acid the sulphur is evolved as hydrogen sulphide. Air as well as all oxidizing agents must be excluded from the evolution-flask. 542. For the Absorption of the Hydrogen Sulphide a large number of solutions have been suggested. Formerly a hydrochloric- acid solution of BROMINE was largely used. The sulphur was oxi- dized by this solution to sulphuric acid, which was precipitated and weighed as barium sulphate. The evolution of the disagreeable bromine fumes may be obviated by the substitution of an ammo- niacal solution of hydrogen peroxide or a 3% solution of sodium peroxide. These reagents are, however, almost never free from sulphur, necessitating a separate determination of the amount present. Another class of absorption liquids consists of solutions of metallic salts. The hydrogen sulphide is precipitated as a metallic sulphide, which in some cases is filtered off, washed, and weighed, and in other cases is oxidized by means of an appropriate solution to sulphuric acid which is precipitated as barium sulphate. The use of solutions of silver and copper for this purpose seems to be objectionable because of the reduction of the former by the hydro- carbons evolved, and the formation of an insoluble phosphorus compound in the latter. Alkaline solutions of lead, zinc, and cadmium seem to be very efficient absorbents. Very good results are obtained by washing the cadmium sulphide, drying at 100°, and weighing. 543. lodometric Determination of the Sulphur. — The sulphur in the cadmium and zinc sulphides may also be determined by treatment with a measured excess of standard iodine solution, acidulating, and titrating the excess with standard sodium thio- sulphate. The following reactions then take place: CdS+2HCl=CdCl2+HA HjS+l2=2HI + S. 370 ANALYSIS OF IRON AND STEEL. The hydrogen sulphide may also be absorbed by passing the gas directly through a measured amount of standard iodine solu- tion and titrating the excess. The volatility of the iodine intro- duces an error in this determination. This may be obviated by absorbing the hydrogen sulphide in a solution of caustic soda and determining the sulphur with iodine solution in the same manner as with cadmium sulphide. 544. All of the Sulphur is not Evolved as Hydrogen Sulphide when a metal of high carbon content is analyzed. Organic sul- phides are formed among which methyl sulphide (0113)28 has been identified.* These sulphur compounds are expelled from the evolution-flask with difficulty, and are not taken up so readily by the absorption solution. By passing the gases through a red- hot glass tube these compounds may be decomposed so that the sulphur is absorbed. 545. A Convenient Form of Apparatus for carrying out the deter- mination is shown in Fig. 59. The Kipp generator is charged Fig. 59. with zinc and acid for the generation of hydrogen. The wash- bottle A contains a solution suitable for the absorption of hydro- gen sulphide. A solution of lead nitrate in caustic potash is very efficient. It is made by dissolving one part of caustic potash in two parts of water and pouring in a solution of lead nitrate with constant stirring until a permanent precipitate is produced. After being * Philips, Stahl und Eisen, 16, 633. SULPHUR. 371 allowed to settle, the clear liquid is siphoned off for use. It should be kept in a glass-stoppered bottle. The stopper should be coated with a little parafhne to prevent its sticking. This solution may also be used in the absorption-flasks C and E. 20 to 30 c.c. in each flask will be sufficient, water being added until the flasks are about half-full. The flask B should have a capacity of about | liter. 546. Solution of the Iron. — Ten grams of fllings are placed in the flask B. If a portion weighing 13.732 grams is used, the number of mUhgrams of barium sulphate obtained will be equal to the number of one-thousandths per cent of sulphur. If, for instance, .0156 gram of barium sulphate is obtained, the percentage of sulphur present will be .0156. The air in the apparatus is displaced with a stream of hydrogen. The glass tube D is then heated to redness, the dropping-funnel fllled with dilute hydrochloric acid, and connection made immediately with the Kipp generator. About 20 c.c. of dilute acid should be used for each gram of iron. Some workers use for each gram of iron 10 c.c. of a dilute hydrochloric acid composed of equal volumes of concentrated acid and water, while others recommend 18 c.c. of a mixture of two-thirds dilute hydrochloric acid and one-third dilute sulphuric acid. When the glass tube has been heated to dull redness the acid is allowed to flow into the flask B by turning the stop-cock of the dropping-funnel. The flask is heated with a Bunsen burner until the solution boils, a rapid stream of gas being allowed to pass through the apparatus. After boiling for about flfteen minutes the iron is usually dissolved. 547. Absorption of the Sulphur by Lead Solution. — ^The Bun- sen burner is removed and the stream of hydrogen allowed to flow for about fifteen minutes more to sweep all of the hydrogen sul- phide out of the apparatus. The precipitate in the flask C is filtered off on a small paper. If a precipitate has formed in the flask D, it is filtered off on the same paper and washed once or twice with hot water. The moist paper with the precipitate is thrown into a beaker, a little potassium chlorate added, and 5 to 20 c.c. of con- centrated hydrochloric acid added, the amount used depending on the amount of lead sulphide present. Allow the beaker to stand in a warm place until chlorine is no longer evolved, dilute the strong 372 ANALYSIS OF IRON AND STEEL. acid and filter. After washing the paper with hot water the fil- trate is neutralized with ammonia and then acidified with 2 or 3 drops of hydrochloric acid. The sulphuric acid is then precipi- tated with barium chloride and weighed as barium sulphate. 548. Absorption of the Sulphur by a Cadmium Solution. — Instead of using an alkaline solution of lead nitrate in the absorp- tion-flasks C and D, an alkaline solution of cadmium may be used as proposed by T. T. Morrel.* The cadmium sulphide is filtered off on a Gooch crucible and washed with water containing a little ammonia. The precipitate is dried at 100° and weighed. The cadmium sulphide together with the paper may also be placed in a beaker, 200 or 300 c.c. of cold water added, and then dilute hydrochloric acid until the precipitate is dissolved. A few cubic centimeters of freshly prepared starch solution are then intro- duced, and standard iodine solution added from a burette until a permanent blue color is produced. DETERMINATION OF PHOSPHORUS. The determination of phosphorus in iron or steel is invariably preceded by solution of the iron and oxidation of the phosphorus to phosphoric acid, which is then separated from the iron by pre- cipitation as ammonium phospho-molybdate. The estimation of the amount of phosphoric acid may then be carried out gravi- metrically or volumetrically by any of the methods given in Chapters IX and XXVII. 549. Oxidation with Nitric Acid and Separation of Silica. — From 1 to 5 grams of the iron or steel, according to the amount of phosphorus present, are dissolved in nitric acid, sp. gr. 1.2, about 15 c.c. being used for each gram of iron. Evaporate to dryness, and heat in the air-bath at 200° for one hour. At this tempera- ture the oxidation of the phosphorus is completed. The residue is treated with 30 c.c. concentrated hydrochloric acid, and warmed until the oxide of iron is dissolved. The solution is evaporated to dryness again to dehydrate the silica. Dissolve once more in * Chem. News. XXVIII, 229. PHOSPHORUS. ;573 30 c.c. hydrochloric acid. Evaporate the solution to a sirupy consistency, dilute, and filter. The insoluble residue may be used for the determination of silicon, if that is desired. 550. Separation of Titanium. — If titanium is present it will form an insoluble phosphate which will remain with the silica and cannot be dissolved by means of hydrochloric acid. The paper containing the washed silica and graphite is transferred to a plati- num crucible and burned. The silica is volatilized by treatment with hydrofluoric and a few drops of concentrated sulphuric acid and heating until the sulphuric acid is volatilized. The residue, consisting of TiOj and P2O5, is fused for one half hour with 2 or 3 grams of sodium carbonate, the melt dissolved in water, and the solution containing the P2O5 as sodium phosphate is filtered. The filtrate from the insoluble sodium titanate is acidified with nitric acid and added to the first filtrate from the impure silica. The hydrochloric acid is then expelled by addition of 30 c.c. of concen- trated nitric acid and evaporating nearly to dryness. 551. The Phosphoric Acid may be Precipitated, after nearly neutralizing the nitric acid with ammonia, by means of molybdate solution and determined gravimetrically as directed in Chapter IX, or volimietrically as directed in Chapter XXVII. 552. Oxidation with Potassium Permanganate. — The destruc- tion of the organic matter as well as the complete oxidation of the phosphorus in the solution of the iron in nitric acid may be advan- tageously accomplished by means of potassium permanganate as well as by heating the dry nitrate at 200°. After dissolving the iron in nitric acid, 10 c.c. of a 1.5% solution of potassium perman- ganate is added and the solution boiled until the pink color is destroyed and the manganese is reduced to the dioxidej This precipitate is dissolved by the addition of small amounts of sul- phurous acid, ferrous sulphate, or sodium thiosulphate. 'V^Tien the solution has cooled to about 40°, it is neutralized with ammonia, the precipitate dissolved with a little nitric acid, and the phosphoriu acid precipitated by the addition of molybdate solution. 553. The Molybdate Precipitate may be directly weighed as directed in Chapter IX. It may also be titrated with standard caustic potash or soda as directed in Chapter XXVII, or with 374 ANALYSIS OF IRON AND STEEL. standard potassium permanganate solution after reduction with zinc as directed in the same chapter. DETERMINATION OF CARBON. Carbon exists in iron and steel in several conditions, and the qualities of the metal are considerably influenced by the condition in which the carbon exists. Analytical methods are in use for detei'mining the amount of carbon existing in two conditions. The so-called free or graphitic carbon as well as the total amount of carbon are determined directly. The combined carbon is usually determined by difference. TOTAL CARBON. a. Solution of Iron and Oxidation of Carbon by Chromic Acid. This is a very accurate as well as a rapid and largely used method of determining carbon in all grades of iron and steel, with the exception of ferric chrome and irons high in silicon. The carbon is oxidized by the chromic acid to carbon dioxide, which may be absorbed in an appropriate apparatus and weighed or the volume of the gas may be measured and the weight calculated. 554. The Following Solutions are prepared for the combustion of the carbon: 1. Two hundred grams of copper sulphate are dissolved in t liter of water. 2. A saturated solution of chromic acid. 3. A solution made by taking 35 c.c. of solution 2, 190 c.c. 01 water, 750 c.c. of concentrated sulphuric acid, and 340 c.c. of phosphoric acid, sp. gr. 1.4. After adding a few cubic centimeters of sulphuric acid, solution 2 should be heated to boiling to destroy any organic matter that might be present. Solution 3 should also be heated to boiling before use. 555. The Apparatus is set up as shown in Fig. 60. The round- bottomed flask A should have a capacity of about 250 c.c. The short piece of combustion tubing B is filled wi'th lumps of copper oxide which are kept in place by short coils of copper gauze. The TOTAL CARBON. 375 tube is placed in a short tube furnace or is heated by several Bunsen burners. This tube is heated to redness during the com- bustion of the carbon in order to decompose the hydrocarbons evolved. If it is omitted, the result must be increased 2% for malleable iron and steel and 3^% for pig iron. The presence of copper sulphate in the combustion-flask reduces the amount of the hydrocarbons evolved. The U-tube C is filled with calcium chloride for drying the carbon dioxide which is absorbed in the caustic potash bulb D. If this bulb is not fitted with a small calcium chloride tube, the U-tube E must be weighed as well as Fig. 60. the caustic potash bulb D. Another calcium chloride tube must then be inserted between the U-tube E and the aspirator. The U-tube attached to the dropping-funnel is filled with soda-lime and serves to remove carbon dioxide from the air drawn through the apparatus. The apparatus is set up and tested for leaks as already directed for the determination of carbon dioxide in Chap- ters IX and XV. 556. Process. — From 1 to 5 grams of the iron, depending on the amount of carbon present, is weighed out and placed in the flask A. 1.5 c.c. of copper sulphate solution is introduced through the drop- ping-funnel. After one or two minutes introduce 15 c.c. of the chromic acid solution and 135 c.c. of solution 3. Allow a slow stream of air to pass through the apparatus and heat the solution gently with the Bunsen burner, bringing it to a boil in fifteen to twenty minutes. Continue the boiling for about two hours. 376 ANALYSIS OF IRON AND STEEL. Aspirate about 2 liters of air througii the apparatus, disconnect the absorption-tubes, and weigh after allowing them to remain in the balance-case for about fifteen minutes. b. Weighing of Carbon After Separation of the Iron by Means of Potassium Cuprig Chloride Solution. 557. Action of Copper Solutions, — On treating iron or steel with a solution of copper, the copper is displaced by the iron which unites with the acid, forming a ferrous salt. The combined carbon as well as the graphitic carbon remains undissolved, no hydrogen or hydrocarbons being evolved. Formerly a neutral solution of copper sulphate was employed. This has largely been replaced by a solution of cuprig ammonium chloride, by which the iron is dissolved much more rapidly, the precipitated copper also going into solution. On account of the difficulty of obtaining the double ammonium salt free from organic material, the use of the double chloride of potassium and copper has arisen. It has also been recently shown by the American members of the International Steel Standards Committee that more accurate results are obtained by the use of a copper solution containing about 10 per cent of hydrochloric acid. The reactions taking place may be represented as follows: Fe+CuCl,=FeCl2+Cu; Cu+CuCl2 = 2CuCl. The cuprous chloride produced in this manner may be recon- verted into cupric chloride by passing chlorine through the solu- tion until it smells strongly of chlorine. Blair * states that a solu- tion regenerated in this manner is more active than a fresh solution. The same solution may be used repeatedly until the concentration of the iron salts is too great. Blair has used the same solution eleven times with good results. 558. The Solution of the Iron is effected by treating 1 gram of pig iron, Spiegel, or ferromanganese with 100 c.c. of a saturated solution of the potassium cupric chloride and 7.5 c.c. of concen- trated hydrochloric acid. 3 grams of steel or puddled iron are * Chemical Analysis of Irom, 1901, p. 165. TOTAL CARBON. 377 taken and treated with 200 c.c. of the copper solution and 15 c.c. strong hydrochloric acid. The iron is dissolved very much more rapidly if the solution is stirred. After digestion for several minutes at the ordinary temperature the solution may be warmed gently, but not higher than 60° or 70°. If many determinations are to be made, a mechanical device for stirring should be used. 559. Direct "Weighing of Carbon. — ^When the iron and the precipitated copper are completely dissolved* the carbon is fil- tered off and washed with hot water and a little hydrochloric acid to dissolve the copper and iron salts. If the iron or steel contains only combined carbon and is free from graphite the carbon may be weighed directly, after being filtered off on a Gooch crucible and dried at 100°. The carbon is then burned off and the residue weighed. The difference is carbonaceous matter which contains about 70% of carbon. c. Combustion of Separated Carbon in a Stream op Oxygen. 560. Filtration of Carbon on a Platinum Boat. — A more common and generally applicable method consists in burning the separated carbon in oxygen or chromic acid and collecting and weighing or measuring the carbon dioxide produced. If the carbon is to be burned in a stream of oxygen, the most suitable apparatus on which to filter it off and wash it is a platinum boat with perforated bottom which is of su6h a size that it may be placed in the glass or porcelain tube used for the combustion. A suitable holder is provided for the boat so that it may be used in the same manner as a Gooch crucible. Asbestos which has been ignited so as to free it from organic matter is made into a pulp by mixing with water, floated on the bottom of the boat and sucked into a smooth compact layer with the filter-pump. After filtering off the carbon on the boat and wash- ing it, the whole is dried at 100° and transferred to the combus tion-tube. Carbonaceous matter adhering to the sides of the beaker in which the iron was dissolved is wiped off with a wad of ignited asbestos held by a pair of forceps. The asbestos is then trans- ferred to the boat. * The metallic copper may easily be seen because of its bright-red color. 378 ANALYSIS OF IRON AND STEEL. 561. Filtration of Carbon on a Carbon Filter. — Instead of the perforated boat and holder a small platinum tube fitting into the combustion-tube may be used for filtering off the asbestos. The asbestos is held on a perforated platinum disk resting on a shoulder in the tube. After the carbon has been filtered off and washed, the disk is loosened by pushing it further into the tube by means of a glass or platinum rod. This is done to provide a passageway for the oxygen used in the combustion. After drying, the tube con- taining the asbestos and the carbon is inserted into the combustion- tube. An ordinary glass carbon filter may also be used in which the asbestos rests on a perforated platinum or procelain disk or a spiral of platinum wire. After filtering off the carbon, the disk with the asbestos and the carbon is pushed out of the tube and transferred to a platinum boat. The particles of asbestos and carbon still adhering to the glass tube are wiped out by wads of ignited asbestos held in a pair of forceps. The boat and contents are then dried at 100° and transferred to the combustion-tube. 562. A Platinum, Glass, or Porcelain Combustion- tube may be used. It should be long enough to extend several centimeters beyond the furnace, so that the rubber stoppers are not softened by the heat. The forward half of the portion of the tube which is in the furnace is filled with granulated copper oxide which is held in place by short rolls of copper gauze. Another roll of copper gauze, to which a hook made of copper wire is fastened, is pro-- vided for insertion into the tube after the boat has been placed in the empty rear end. 563. A Stream of Oxygen is obtained from a steel tank which has been filled under pressure or from an ordinary gasometer which has been filled with oxygen by heating gently in a retort an intimate mixture of 1 part of powdered manganese dioxide and 20 parts of potassium chlorate. When the oxygen begins to be given off from this mixture the heating with the Bunsen burner must be discon- tinued or the evolution of gas will be too violent. The oxygen is purified by passing through a large U-tube or tower filled with soda-lime and pieces of fused caustic potash. A similar apparatus should be provided for purifying air. 564. Absorption Apparatus. — The carbon obtained from iron and steel is almost never free from sulphur. Some chlorides are TOTAL CARBON. 379 also generally present which cannot be washed out. Sulphur dioxide, hydrochloric acid, and chlorine are therefore given off on burning the carbon, and these gases must be absorbed before the carbon dioxide is allowed to pass into the caustic potash. A spiral of silver foil should be placed in the end of the combustion- tube for the absorption of the chlorine. A U-tube filled with pieces of pumice-stone which have been saturated with copper- sulphate solution and thoroughly dried at a little over 200° should be connected with the combustion-tube for the absorption of hydrochloric acid. For the absorption of sulphur dioxide a layer of lead peroxide is placed in the bottom of this tube and separated from the copper sulphate by means of wads of asbestos or cotton wool. A U-tube filled with calcium chloride, a caustic potash bulb for the absorption of carbon dioxide, and a guard tube containing soda-lime and calcium chloride are attached in the order mentioned. After the apparatus has been tested for leaks in the usual manner, the combustion-tube is heated to redness and oxygen and air passed through to burn out any organic matter which may be present. 565. The Combustion. — The caustic potash bulb is weighed and replaced in the series of U-tubes. The boat containing the carbon is inserted in the combustion-tube and pushed in until it is close to the layer of copper oxide. The roll of copper gauze is inserted in the end of the combustion-tube, which is closed with the rubber stopper connecting with the supply of oxygen. The apparatus is tested for leaks by sucking out a little air so that the liquid rises in one limb of the caustic potash bulb. If this column of liquid remains at a constant height for at least five minutes the apparatus is tight. A slow stream of oxygen is now admitted, and two burners are lit under the forward end and one under the copper gauze in the rear end of the combustion-tube. When these portions of the tube are red-hot, the third burner from the forward end is lighted and then the other burners in order with a few minutes ' interval. If a glass tube is used, the combustion may be observed by the glowing of the carbon. The combustion is usually complete fifteen minutes after the tube has become red- hot. The asbestos in the boat then appears nearly pure white.' 380 ANALYSIS OP IRON AND STEEL. The burners are then turned down and, after a few minutes, extin- guished so as to cool the tube gradually. The stream of oxygen is replaced by a stream of air of which from 1 to 2 liters should be passed through the apparatus to sweep out the carbon dioxide. Neither the oxygen nor the air should pass faster than three bub- bles per second. The absorption apparatus is then disconnected, wiped perfectly clean with a linen handkerchief, and after stand- ing in the balance-case for not more than one-half hour is weighed. Until experience has been attained in conducting the combustion, it is advisable to insert the absorption apparatus and repeat the combustion to ascertain if any of the carbon remains unburned. 566. If Several Determinations are to be made a duplicate absorption apparatus should be prepared and weighed while the first combustion is in progress. When the first absorption appara- tus is disconnected, the second is inserted, and a boat contain- ing the carbon from another sample of iron is inserted in place of the first boat. During the progress of the second combustion the absorption apparatus of the first is weighed and may then be used for the third determination. When work is discontinued the entrance of air or other gases into the apparatus is prevented by closing all openings with glass plugs. The carbon from which the iron has been separated may also be burned with chromic and sulphuric acids. d. Combustion of the Separated Carbon with Chromic Acid. 567. An Apparatus similar to that shown in Fig. 61 is used. The flask A has a capacity of about 250 c.c. The U-tube B con- tains a sufficient amount of a solution of sih-er sulphate in con- centrated sulphuric acid to fill the curved part of the tube. An empty U-tube C is placed next and then a U-tube D containing pumice-stone which has been saturated with copper sulphate solu- tion and dried. The usual U-tube filled with calcium chloride for drying the carbon dioxide, the caustic potash bulbs, and the guard tube are then connected in the order mentioned. 568. The Combustion. — The asbestos containing the carbon is transferred to the flask and, after inserting the stopper, 10 c.c. TOTAL CARBON 381 of a saturated solution of chromic acid is introduced through the funnel, followed by 100 c.c. of concentrated sulphuric acid which has been brought to the boiling-point after the addition of a little chromic acid. A slow current of air is passed through the apparatus and the flask is gently heated with the Bunsen burner and gradu- ally brought nearly to the boiling-point. The heating is con- Hh Fig. 61. tinued until no more gas is evolved and the oxidation is complete. The current of air is continued until from 1 to 2 liters have passed thr'ough the apparatus, when the absorption-bulbs are discon- nected and weighed. e. Determination of Carbon After Volatilization of the Iron in a Stream of Chlorine. The carbon in irons containing chromium and a large percentage of silicon can be determined only after volatilization of the iron in a stream of chlorine. 569. Separations Effected. — The iron is converted into ferric chloride, which can be completely expelled by heating the tube. The silica is also volatilized as silicon tetrachloride, SiCl^, and may be absorbed in water which will also contain any titanium present in the iron, being volatilized as titanium tetrachloride, TiCl4. These elements may be separated by adding to the solu- tion concentrated sulphuric acid and evaporating to fumes. The silicic acid is dehydrated and may be filtered off and weighed while the titanium remains in solution. Any metals present 382 ANALYSIS OF IRON AND STEEL. which form volatile chlorides will also be present in this solution. In the boat in which the iron was placed will remain all of the carbon, most of the manganese, chromic chloride, and the slag which was present in the iron. 570. Process. — The chlorine is generated in a large flask of about 1\ liters capacity, charged with 190 grams of manganese (lio:dde intimately mixed with 280 grams of sodium chloride. The chlorine is evolved on adding from a dropping-funncl -350 c.c. of a mixture of equal parts of concentrated sulphuric acid and water. The chlorine is washed by passing through water and dried by passing through concentrated sulphuric acid. The iron is spread over the bottom of a porcelain boat which is placed in a glass combustion-tube which has been drawn out and bent at right angles where it emerges from the furnace so that it can dip under water. All connections must be made of glass as far as possible, the inner surface of rubber stoppers being coated with parafhne. The air is displaced from the apparatus by the stream of chlorine and the combustion-tube heated just sufhciently to expel the ferric chloride from the boat. This is usually accomplished by heating the tube to dull redness. The chlorides are dissolved out with water and the carbon filtered off on ignited asbestos, which is re- turned to the boat and dried. The carbon may then be burned in a combustion-tube in a stream of oxj^gen, or it may be burned with chromic and sulphuric acids as already directed. DETERMINATION OF GRAPHITIC CARBON. 571. On dissolving iron and steel in acid, part of the carbon remains undissolved while part goes into solution or is evolved in combination with hydrogen. As graphite is insoluble in dilute acid, all of the carbon present in this form remains undis- solved. If dilute hydrochloric acid is used a part also of the combined carbon remams undissolved. If nitric acid of specific gravity 1.2 is used, only the graphitic carbon remains undis- solved. For this determination 1 gram of pig iron is taken, while as much as 10 grams of steel containing a very small per cent of graphitic carbon is taken. 15 c.c. of the nitric acid is taken for each gram of iron. The first violent action of the acid COMBINED CARBON. 383 is moderated by cooling the beaker with water. "When all of the iron is dissolved, dilute the acid, filter off the carbon on ignited asbestos, and wash with water until free from iron. Treat the residue on the filter with a hot solution of potassium hydroxide of sp. gr. 1.1 to remove siUca, and wash with water, then with dilute hydrochloric acid, and finally with water. Determine tlie amount of carbon present by burning in oxygen or in chromic acid as already directed. DETERMINATION OF COMBINED CARBON. The amount of combined carbon may be found indirectly by subtracting the amount of graphitic carbon from the total amount of carbon found. 572. Colorometric Method of Eggertz. — This method is founded on the fact that when steel is dissolved in nitric acid of 1.2 specific gravity a brown color is given to the solution which is proportional to the amount of combined carbon present. Two specimens of steel which give the same depth of color contain the same percentage of combined carbon. If the colors differ, the darker solution may be diluted until the depth of color is identi- cal. The ratio of the percentage of combined carbon in the two samples of steel will then be the same as the ratio of the volumes of the two solutions. If the percentage of combined carbon in one of the steels is known, the amount present in the other may be calculated. A so-caUed standard steel must therefore be se- lected and the percentage of combined carbon determined by one of the methods already given. 573. Standards. — It has been found, however, that two samples of steel containing exactly the same percentage of combined carbon give different colors if one sample has been treated differently, so far as tempering, hammering, etc., or if it has been made by a different process of manufacture. This peculiarity is believed to be due to the fact that the carbon exists combined in two different conditions, and when existing in one condition does not impart the same depth of color to the acid as when it exists in the other condi- tion. Standards should therefore be at hand for each kind of steel to be tested, such as Bessemer, crucible, open-hearth, etc., 384 ANALYSIS OF IRON AND STEEL. and the rolling, hammering, tempering, etc., of the standard should as far as possible be the same as that of the sample to be analyzed. The percentage of carbon should also be as nearly as possible the same in the standard as in the steel to be analyzed. 574- Solution of the Steel. — A sufficient number of test-tubes 6 inches long and f inch in diameter are taken and 2 grams of each sample, including the standard, are weighed out and placed in a test-tube. The test-tubes are carefully numbered to correspond to the numbers of the samples, and are placed in beakers or bottles containing cold water. Nitric acid of 1.2 specific gravity is then added to the test-tubes. Acid of this strength is made by diluting concentrated nitric acid (sp. gr. 1.4) with an equal volume of water. The acid is added from a burette, 3 c.c. being used for steels containing .3% of carbon; 4 c.c. for .3 to .5% of carbon; 5 c.c. for .5 to .8%; 6 c.c. for .8 to 1%; and 7 c.c. for over 1%. If the amount of carbon present is unknown, 3 c.c. of the acid should be added at first, and then successive portions of 1 c.c. as the increase in the depth of color or the presence of undissolved carbonaceous matter indicates the necessity for more acid. The nitric acid used must be absolutely free from chlorine. A small funnel or a glass bulb of suitable size should be placed in the mouth of each test-tube. The test-tubes are now immersed in the water of a water-bath while held upright by some suitable device, such as a metallic disk in which holes of the proper size to admit the tubes have been cut. The water is heated to boiling and the tubes shaken at intervals until the carbonaceous matter is completely dissolved, which will require from fifteen to forty-five minutes. As soon as solution is complete each tube is removed from the water- bath and placed in cold water, the light being excluded as much as possible. Direct sunlight bleaches the color very quickly. 575. For Comparing the Color of the Solutions Eggertz used graduated tubes. These tubes are of uniform bore, made of clear, colorless glass, and are carefully graduated from the bottom up- wards, having a capacity of 30, or more, cubic centimeters. Trans- fer the solution having the lightest color to one of the tubes, using the same amount of water to rinse the test-tube as the volume of the solution. Mix the resulting solution thoroughly. In every case the acid solution of the steel must be diluted with at least as MANGANESE. 385 much water as the amount of acid used to dissolve the steel in order to destroy the color due to the ferric nitrate. The solution of the steel to be compared with the one first diluted is transferred to another tube in the same manner, and the color of the two solutions is compared by holding the tubes vertically and side by side in front of a piece of white paper between the observer and the light. The darker solution is diluted by the addition of water until the depth of color in the two tubes is iden- tical. The percentage of combined carbon known to be present in the standard steel is divided by the volume of the solution of this steel. The quotient gives the amount of carbon represented by each cubic centimeter of the solution. This number multiplied by the number of cubic centimeters to which the solution of the steel to be analyzed was diluted gives the percentage of combined carbon present. DETERMINATION OF MANGANESE. The difficulties attending the determination of manganese in iron and steel arise mainly from the presence of the large amounts of iron. The iron may be precipitated as basic acetate, but the precipitation must be repeated several times, and if the percentage of manganese is small, so that a large amount of iron must be taken for the analysis, the basic acetate precipitate becomes very large and unwieldy. The separation of the iron by solution of the ferric chloride in ether is a more rapid and convenient method, which has already been described in Chapter XIII, p. 160. a. Ford's Method, Equally as efficient is a method known in this country as Ford's method.* Advantage is taken of the fact that potassium chlorate converts the managnese present in strong nitric acid solu- tion into manganese dioxide, which can be filtered off from the iron which remains in solution. The separated manganese may * Trans. Inst. Min. Engineers, IX, 397, X, 100. 386 ANALYSIS OF IRON AND STEEL. then be tletermined gravimetrically or volumetrically. The reac- tions taking place may be represented as follows: KCIO3 + HNO3 = KNO3 + HCIO3 ; 2HCIO3 + Mn (NO3) 2 = MnO., + 2CIO2 + 2HNO3. 576. Solution of the Steel. — 5 to 10 grams of steel or pig iron are taken, while only 1 gram of Spiegel or ferro-manganese contain- ing 20 to 40% of manganese, and ^ gram if the per cent of manga- nese is still higher. If less than 0.2% of silicon is present, the weighed portion is dissolved in 60 c.c. of nitric acid of 1.2 specific gravity and the solution evaporated to small bulk. A large beaker should be used and the acid added cautiously. If the percentage of silicon is high, the steel should be dissolved in hydrochloric acid and the solution evaporated to dryness to dehydrate the silica, which otherwise forms a gelatinous mass and interferes with the subsequent filtration. Pig iron is also dissolved in hydro- chloric acid, the solution filtered, and the filtrate evaporated to dryness. The hydrochloric acid is removed in either case by evaporation with 50 c.c. of strong nitric acid to a small bulk. The silica may also be removed by dissolving the iron in nitric acid to which a few drops of hydrofluoric acid have been added and evaporating to a small bulk. 577. Precipitation of the Manganese. — This solution, as well as that obtained from the steel low in silica, is treated with 100 c.c. of concentrated nitric acid and 5 grams of potassium chlorate added. The solution is boiled for fifteen minutes, 50 c.c. strong nitric acid are added, and 5 grams of potassium chlorate, and the solution again boiled for fifteen minutes. When no more chlorine fumes are evolved the solution is rapidly cooled by standing the beaker in cold water and the precipitate filtered off on asbestos. It is washed with strong nitric acid and then with water. It is advisable to keep the strong nitric acid filtrate separate from the wash-water and to test it for manganese by adding a little more potassium chlorate and again boiling. 578. If the Manganese is to be Determined Gravimetrically the manganese dioxide is dissolved in sulphurous acid or hydrogen peroxide and hydrochloric acid. The excess of the sulphur dioxide MANGANESE. 387 or hydrogen peroxide is expelled by boiling the solution. The small amount of iron present is oxidized by the addition of a little bromine water the excess of ^vhich must be boiled out. It is precipi- tated by adding an excess of ammonia to the hot solution. The precipitate is filtered off, washed once or twice, dissolved in hydro- chloric acid, and the solution boiled to expel a trace of chlorine from the manganese dioxide and reprecipitated with ammonia, fil- tered off, and washed thoroughly. The combined filtrates con- tain all of the manganese and a little cobalt if that element was present in the iron. The cobalt may be precipitated by passing hydrogen sulphide through the solution after acidifying with acetic acid and heating to boiling. After filtering off the precipitate and boiling the solution to expel the excess of hydrogen sulphide, the manganese may be determined either as sulphide or as phos- phate as directed in Chapters VI and VII. 579. The Manganese may also be Determined Volumetrically by dissolving the manganese dioxide in excess of oxalic acid or a pure ferrous salt and titrating the excess with a standard solution of potassium permanganate (see pp. 322, '325). As the per cent of oxygen in the precipitate may not be exactly that correspond- ing to the formula MnOa, it is advisable to standardize the potas- sium permanganate solution by using the precipitate obtained from a sample of steel or iron in which the manganese has been determined gravimetrically. The manganese dioxide obtained from a weighed amount of this iron is dissolved in excess of a reducing agent and the excess determined by means of the per- manganate solution. h. Volhard's Method. 580. Solution of the Metal. — This very accurate method of determining manganese may be used for determining this element in all samples of iron and steel except those containing very minute amounts of manganese. Weigh out 5 grams of ferro- manganese and from 1 to 2.5 grams of spiegeleisen, white pig iron, steel, etc. Dissolve * in a mixture of 275 c.c. of water 125 c.c. of * Reis, Zeit. f. angew. Ch., 1892, pp. 604, 672. 388 ANALYSIS OP IRON AND STEEL. concentrated nitric acid (sp. gr. 1.4) and 100 c.c. concentrated sulphuric acid. Use from 25 to 50 c.c. of this acid mixture, accord- ing to the amount of iron taken. After the iron is dissolved, evapo- rate the solution until fumes of sulphuric acid are evolved. After cooling, add 100 c.c. of water and 10 c.c. of the acid mixture and warm until solution is complete. Transfer to a flask of 750- to 1000-c.c. capacity and add 3 grams of barium peroxide and 5 c.c. concentrated nitric acid to oxidize the hydrocarbons present. Boil several minutes to decompose the excess of barium peroxide. 581. Titration. — Dilute the solution until the flask is about one- half full, add zinc oxide suspended in water, shaking the solution after each addition until the solution is neutral and the iron is pre- cipitated, leaving it colorless or slightly milky with the excess of zinc oxide. Warm the solution to about 90°, but not to boiling, and titrate with standard potassium permanganate solution as directed in Chapter XXIV, page 323. 582. If the Zinc Oxide contains manganese it may be purified as follows: Dissolve the impure oxide in hydrochloric acid, finally adding an excess of the zinc oxide so that part of it remains undis- solved. Add a little bromine and heat the solution to boiling and filter. Precipitate the zinc oxide by means of ammonia, carefully avoiding an excess. Wash the oxide thoroughly and then transfer to a bottle with water. Shake the bottle well before withdrawing any of the oxide for use. c. Deshat's Method.* This method is based on the fact that manganese nitrate when boiled with nitric acid and lead peroxide is converted into permanganic acid, which may then be titrated with a standard solution of sodium arsenite. The method is especially applicable to iron and steel containing less than 2% of manganese. It caimot be used when chromium is present, as chromic acid is produced. 583. Process. — 0.5 gram of steel or pig iron is dissolved in a small beaker or test-tube in 30 c.c. nitric acid of 1.2 specific gravity. The solution is boiled until the nitrous fumes are completely * Bull. Soc. Chim. de Paris, June 20, 1878. MANGANESE. 389 expelled. From 1 to 3 grams of lead peroxide or red lead free from manganese are then added and the solution diluted with hot water to about 60 c.c. It is boiled for three or four minutes and after allowing the peroxide of lead to settle, the solution is decanted and the residue again boiled after the addition of 50 c.c. of dilute nitric acid. If the solution again becomes colored with permanganic acid, it must be decanted and the boiling repeated after the addition of dilute nitric acid. The solution is filtered through asbestos which is free from organic miatter. The filtrate is titrated with a standard solution of sodium arsenite. 584. The Sodium Arsenite Solution may be made by dissolving 4.5 grams of pure arsenious oxide in water to which 25 grams of sodium carbonate has been added and diluting the solution to 2 liters. This solution should then be standardized by comparison with a standard potassium permanganate solution, or a still better plan is to standardize it with a sample of steel in which the man- ganese has been determined gravimetrically. This is done by weighing out J gram of the steel, dissolving in nitric acid, and oxidizing with peroxide of lead. The permanganic acid produced in this manner is titrated with the arsenite solution to be standard- ized. The amount of manganese corresponding to 1 c.c. of the arsenite solution may then be calculated from the known amount of manganese in the steel used. 585. Separation of Lead Peroxide by a Centrifugal Machine. — Instead of filtering the permanganic acid from the lead peroxide, the separation may be effected by means of a centrifugal machine. In order to secure complete oxidation of the manganese by boiling once, a smaller amount of the steel is taken, from .05 to .10 gram being used. 10 c.c. of the nitric acid of sp. gr. 1.2 are added to dissolve the iron and after the nitrous fumes have been expelled by boiling, 10 c.c. of hot water and ^ gram of lead peroxide are added and the solution boiled for three minutes. The solution is washed into a wide-niouthed two-punce bottle, enough water being used to make 50 c.c. The bottle is then placed in the cen- trifugal machine and rotated for two minutes. The liquid will then be perfectly clear and may be poured out without disturbing the lead peroxide, which forms a compact layer on the bottorn- of the bottle. The permanganic acid is then titrated with a solution 390 ANALYSIS OF IRON AND STI^EL. of sodium arsenite which is equal to about .0001 gram of man- ganese per cubic centimeter. PROXIMATE ANALYSIS OF COAL. The methods used in the analysis of coal do not give the amount of any definite chemical compound which is present in the coal. The determination of moisture gives very nearly the amount of water present, but as this constituent is obtained by loss in weight after drying the coal in the air, an error is introduced because sulphur and other constituents present are oxidized at the temperature used for drying the coal, and therefore the coal loses weight for a time, and then begins to increase in weight. The conditions under which this as well as other determinations are made are somewhat arbitrarily fixed so that the results obtained by various chemists may be comparable with each other and form a basis for the comparison of the coal§ analyzed. The methods given in this chapter are those adopted by a committee of the American Chemical Society, consisting of Wm. A. Noyes, W. F. Hillebrand, and C. B. Dudley.* 586. The Determination of Moisture is made by heating 1 gram for one hour at 104° to 107°. This temperature may most readily be maintained by using a double-walled air-bath in which pure toluene is boiled. By attaching a reflux condenser to the usual steam vent, loss of toluene may be obviated. As considerable moisture is lost in powdering the coal, the committee recommends the making of two determinations of moisture, one cii the coal as powdered for the determination of sulphur, ash, etc., and another on the material broken up only sufficiently fine for sampling. As it is necessary that the percentages found shall apply to the unpow- dered coal as it is actually used, a correction must be made on the percentages f')und by using the powdered material. The differ- ence in the percentages of moisture found is divided bj^ the sum of the percentages of the other constituents in the powdered coal. The quotient multiplied by 100 is the percentage hj wliich * Jour. Am. Chem. Soc, XXI, 1116. VOLATILE MATTER. 391 the percentages of the other constituents are reduced. The illus- tration given in the report mentioned is as follows : Coarsely Ground Powdered Coal. Coal. Moisture 12.07 10.39 Volatile combustible matter 34.25 The correction factor will be 12.07-10.39 1.68 = .0187. 100-10.39 89.61 The true per cent of volatile combustible matter will be 34.25 - (34.25 X. 0187) =33.61. 587. The Volatile Combustible Matter is determined by heating a fresh portion of the powdered coal in a tightly closed platinum crucible with the full flame of a Bunsen burner for se^•en minutes. The loss in weight less the moisture is the volatile combustible matter, Avhile the residue in the crucible is the coke. This determination serves to distinguish the different kinds of coal which are mined for various industrial purposes. The anthra- cite COALS give only a small per cent of volatile combustible matter, leaving a rciidue in the crucible which can hardly be distinguished from the powdered coal. The bituminous and CANNEL COALS melt ou being heated in the crucible, and the vola- tile matter in escaping produces a porous mass which is much more bulky than the original coal. This material is what is known in commerce as "coke." A good bituminous coal gives about 60% of coke. Cannel coals, on the other hand, give 60% or 70% of volatile combustible matter with a corresponding smaller per cent of coke. Between these extremes coals may be found, giving nearly all percentages of coke and volatile combustible matter. 588. The Percentage of Ash is found by burning a weighed portion of the coal in a platinum crucible. The percentage of ash subtracted from the percentage of coke gives the percentage of fixed CARBON, the volatile combustible matter being composed of carbon which has united with the hydrogen present, forming vola- tile combustible hydrocarbons. 392 ANALYSIS OF COAL. It is evident that a low percentage of ash in coal is very desir- able, as this constituent not only has no calorific value, but forms a waste product which must be disposed of. 589. For Determining Sulphur ^'arious methods have been pro- posed, and used, the difference being found in the methods of oxidizing the carbon and the sulphur, which are ultimately precipi- tated by means of barium chloride and weighed as barium sul- phate. Eschka's method, by which the coal is burned by the oxygen of the air after mixing with magnesium oxide and sodium carbonate, and the sulphur oxidized by means of bromine water, is one of the most rapid as well as a very accurate method. As the sulphuric acid is extracted from the alkaline flux with hot water, some of the silica in the coal is also dissolved, and this is subsequently carried down with the barium sulphate. More accurate results may be obtained by evaporating the solution to dryness after acidifying with hydrochloric acid in order to dehydrate and remove this silica. 590. The Sulphur Exists in Coal in at least three conditions: as a METALLIC SULPHIDE, especially pyrites, as calcium or barium SULPHATE, and as a sulphuretted hydrocarbon. In a proximate analysis of coal about one-half of the sulphur existing as pyrites and all of the organic sulphur is probably ^'olatilized with the volatile combustible matter. The remainder of the sulphur exist- ing as pyrites is expelled during the combustion of the fixed carbon, only that existing as barium or calcium sulphate remaining with the ash unless calcium or barium carbonate was present in the coal. EXERCISE 70. Proximate Analysis of Coal. The carefully selected sample is coarsely ground in a porcelain mortar and well mixed. A few grams are taken and powdered and placed in a well-stoppered bottle. The main portion of the sample is preserved in the same manner. 591. Determination of Moisture. — The temperature of an air-bath is brought to 104°-107° by adjusting the flame of the Bunsen burner. Two platinum or porcelain crucibles are cleaned and, after drying for a few minutes in the air-bath, are cooled in desiccators and weighed. One gram VOLATILE MATTER. 393 of the coarsely gi-ound coal is weighed and placed in one of the crucibles, while one gram of the powdered material is weighed out and placed in the other. The uncovered crucibles are then placed in the air-bath and heated for one hour. They are then cooled in the desiccators after replacing the covers, and weighed. If porcelain crucibles have been used, the coal is carefully brushed out and the crucibles weighed, as the latter may have lost weight during the heating. The loss in weight in each case in centigrams gives the percentage of moisture. The percentage found in the coarsely ground material is reported, while the difference between this percentage and that found in the powdered material is used to calculate the correction on the percentage of the other constituents. 592. Determination of Volatile Combustible Matter- — A 20- to 30-gram platinum crucible is cleaned, placed on a platinum or pipe-stem triangle, and ignited with the Bunsen burner. It is cooled in a desiccator and weighed. One gram of the powdered coal is weighed out and transferred to the crucible, which is placed on a large ring of the iron ring-stand, and the height of the ring so adjusted that the bottom of the crucible is 6 to 8 cm. above the top of the Bunsen burner, which must have a clear blue flame fully 20 cm. liigh when burning free. The cover of the crucible, which must fit tight, is adjusted, and the lighted burner placed under the crucible, wliich is heated seven minutes. The crucible must be protected from draughts during the heating, and the upper surface of the cover must burn clear. The hot crucible is transferred to the desiccator, and after cooling for fifteen to twenty minutes is weighed. The loss in weight less the moisture in the powdered coal is the volatile combustible matter. Compute the percentage, making the correction for difference in percentage of moisture in the coarsely ground and powdered coal. Coke. — The residue in the crucible is coke. The percentage is cor- rected as already directed. 593- Ash is determined by burning one gram of the coal placed in a platinum crucible. The portion used for the determination of moisture may be taken, or a fresh portion of the powdered material may be weighed out. The coke left after expeUing the Volatile matter burns with difficulty. The crucible is placed on its side on a pipe-stem or platinum triangle, and is heated at first with a very low flame. The burner must not be placed so that the flame passes in front of the crucible, thus preventing access of air. The combustion is accelerated by bringing fresh portions of coal to the" surface by turning the crucible when the carbon has burned out, leaving the nearly white ash on the surface. When no more black particles are visi- ble which burn on exposure to the air, the crucible is cooled in the desiccator and weighed, and the percent of ash calculated. The per cent of coke minus the per cent of ash gives the per cent of fixed carbon. 594. Determination of Sulphur by Eschka's Method. — An intimate mix- ture of 1 part of dry sodium carbonate and 2 parts of magnesiimi oxide is made. The magnesium oxide must be light and porous, not compact and •■^94 ANALYSIS OP COAL. heavy. One and one-half grams of this mixture are weighed and placed in a large platinum crucible, or, still better, in a platinum dish having a capac- ity of 75 to 100 c.c. One gram of the finely powdered coal is weighed out and placed on the magnesia mixture, and the two thoroughl)'' mixed by stirring with a platinum or glass rod. The dish is heated very cautiously with the Bunsen burner, which is at first held in the hand. The heat is raised very slowly, especially mth soft coals. When the strong glowing has ceased the heat is gradually increased, until in fifteen minutes the bottom of the dish is at a low red heat. When the carbon is completely burned the residue is transferred to a beaker, the dish is rinsed with about 50 c.c. of water, 45 c.c. of bromine- water is added, and the solution is boiled for five minutes. After allowing the insoluble matter to settle the solution is decanted through a filter. The residue is treated tmce with 30 c.c. of water, which is brought to a boil, and decanted through the filter, which is then washed until the filtrate gives only a sUght opalescence with silver nitrate and nitric acid. To the filtrate, wliich should have a volume of about 200 c.c, li c.c. of concen- trated hydrochloric acid or 3 c.c. of dilute acid are added. The solution is boiled until the bromine is expelled, and 10 c.c. of a 10 per cent solution of barium chloride is added drop by drop with constt,nt stirring, especially at first. Digest on the hot plate until the solution is clear, filter off the barium sulphate, and wash free from chlorides with hot water. Transfer the moist precipitate to a weighed platinum crucible which is heated with a small flame until the paper is burned. Finally heat to redness, cool in the desicca- tor, and weigh. A BLANK DETERMINATION should be made to ascertain if the fusion mix- ture is free from sulphur. This determination is carried out exactly as directed for the determination of sulphur in the coal. The amount of barium sulphate obtained is subtracted from that obtained from the coal. The percentage of sulphur is then calculated. If the coal contains much p3rrite or calcium sulphate, the residue of magnesium oxide must be dissolved in hydrochloric acid and barium chloride added to the hot solution. Any baTium sulphate precipitated is filtered and weighed as already directed. 595. Determination of Sulphur with the Parr Calorimeter. — If a determina^ tion of the calorific value of the coal is made by means of the Parr calorim- eter, the sulphur may be conveniently determined in the same sample. The sodium peroxide oxidizes the sulphur to sodium sulphate. The cartridge is placed in a porcelain dish and the fused mass dissolved out with hot water. After rinsing the cartridge thoroughly, the solution is acidified with hydrochloric acid and evaporated to dryness on the water-bath to dehydrate the silica. The residue is taken up with water and a little hydrochloric acid, and filtered. After washing the paper free from chlorides, the solution is heated to boiling and the sulphuric acid precipitated by means of barium chloride, which is added slowly and with constant stirring. The precipitate is filtered off on a quantitative paper, washed free from chlorides, and ignited and weighed in the usual manner. THERMAL UNITS. 395 DETERMINATION OF THE CALORIFIC VALUE OF FUELS. 596. Effect of Composition on Calorific Value. — The results of a proximate analysis of a fuel give only a general idea of its calorific or heating value by indicating the class in which the fuel belongs. Anthracite coals are characterized by a small percentage of vola- tile combustible matter (usually less than 10%), while bituminous coals contain a large amount of this constituent (usually 30-40%), and in cannel coals this is the main constituent. The proportion of this constituent is dependent on the amount of hydrogen pres- ent, as this element forms volatile compounds with the carbon. When hydrogen burns, it evolves a nmch larger amount of heat than is produced by the combustion of the same amount of carbon. Coals containing a large proportion of volatile combustible matter have therefore generally a high calorific A'alue. On the other hand the calorific value is reduced by the presence of ash which has no calorific value, and by water which absorbs heat when it is con- verted into steam. It is evident that the calorific value of a fuel could be calculated from the percentage of carbon and hydrogen present. This method is subject to the error arising from the fact that the heat d: combustion of hydrogen and car- bon varies with the nature of their chemical combination with each other. Direct methods for determining calorific values are therefore more reliable. The instruments devised for this pur- pose are called calorimeters. 597. Thermal Units. — Various units for measuring heat are in use. They are all based upon the fact that a definite amount of heat will raise the temperature of a definite amount of water a definite number of degrees. Those based on the metric system are the large and the small calorie. A small calorie is defined as the amount of heat necessary to raise 1 gram of water from 4° to 5° Centigrade. The large calorie is the amount of heat necessary to raise 1 kilogram of water from 4° to 5° C, and is equal to 1000 small calories, and is denoted by the capital C. A unit of heat equal to 100 small calories has also been used, and is designated by K. 396 CALORIFIC VALUE OF FUELS. As the specific heat * of water changes but slightly with ordinary changes of temperature, only a very slight error will be intro- duced by using other temperatures than 4° to 5° C. when meas- uring the rise in temperature. For practical purposes, there- fore, a calorie may be defined as the heat necessary to raise one gram of water one degree Centigrade. Another unit of heat in common use is the British Thermal Unit, designated by the initial letters B. T. U. This unit is the amount of heat necessary to raise the temperature of 1 pound of water from 39° to 40° Fahren- heit, or, for ordinary purposes, one degree F. at ordinary tem- peratures. The heat of combustion of a fuel may therefore be expressed in calories or in B. T. U., the relation between these units being easily calculated from the known equivalents of grams and pounds and degrees Centigrade and degrees Fahrenheit (large calorie = 3.96832 B. T. U.). However, when the heat of com- bustion of a substance is given in calories, 1 gram or 1 kilo- gram is the amount of the substance for which the heat is given, while if the heat is given in B. T. U., the number given refers to the amount of heat obtained from 1 pound of combustible substance. As 1 kilogram = 2,20462 pounds,! to convert the heat of combustion in calories to the heat of combustion in B. T. U., multiply the number of calories by 1.8 (3.96732 -.-2.20462). 598. Calorimeters embod}' three essential features : (a) A com- bustion chamber; (h) a tank of water with a delicate thermometer for indicating the amount of heat absorbed; and (c) an insulating device to prevent the external heat from reaching the water and the thermometer, and also to prevent the heat generated in the combustion chamber from escaping from the apparatus. 599. The Bomb Calorimeter of Berthelot is the most accurate cal- orimeter which has been devised. It is adapted for the combustion of any solid or liquid combustible substance. The bomb is the * Specific Heat of Water According to RowlanI) and Peruet. op on oQ on ori 5' 1.0054 9 1.0026 13 1.0009 17' 0,9993 21' 0.9977 6 1.0047 10 1.0019 14 1.0005 18 0.9988 22 0.9974 7 1.0040 11 1.0014 15 1.0000 19 0.9984 23 0.9974 8 1.00.33 12 1.0012 IC 0.9995 20 0.9979 24 0.9972 t For these and other equivalents see Van Nostrand's Cheniical Annual. THE PARR CALORIMETER. 397 characteristic part of this instrument. It consists of a brass or steel receptacle of 400 to 600 cubic centimeters capacity, which can be tightly closed and can sustain a pressure of at least 25 atmospheres. A small platinum cup is suspended from the lid of the bomb. The substance to be burned is placed hi this cup. Two insulated wires are led through the lid and extend to within a short distance of the cup. These wires are connected with a short piece of fine iron or platinum wire which can be fused by means of an electric current and thus ignite the combustible material at any desired moment. Pure oxygen under pressure is led into the bomb by means of a tube passing through the lid and containing a close-fitting valve. The oxygen is forced in until a pressure of 25 atmospheres is produced. In this atmos- phere of pure oxygen under pressure, all substances may be burned. The bomb is immersed in a tank of water in which a delicate thermometer is suspended to register the increase in temperature, the water being agitated by means of an appropriate stirring devise to insure uniform distribution of the heat. The tank of water is enclosed in a vessel made of some good insulating material. The Pahr Calorimeter. 6oo. Chemical Reactions Involved. — This instrument differs from the bomb calorimeter in that sodium peroxide or other strong oxidizing substances are mixed with the fuel to produce the com- bustion. This results in two decided advantages. In the first place the combustion chamber may be very materially reduced in size (from 300 to 600 c.c. in the bomb calorimeter to 25 c.c. in the Parr calorimeter). In the second place the combustion chamber is not required to stand high pressure because the oxygen is not present as a gas under pressure and the products of combustion (CO2, SO3, H2O, etc.) are not given off in the gaseous form. The alkali resulting from the decomposition of the sodium peroxide com- bines with the carbon dioxide, sulphur trioxide, and water, forming sodium carbonate, sodium sulphate, and sodium hydroxide. Veiy little gas pressure is therefore produced in the combustion chamber. An additional error is introduced, however, because heat is evolved by the combination of the products of combustion with the alkali. 398 CALORIFIC VALUE OF FUELS. These are mainly carbon dioxide and water. The former reacts with the sodium oxide to form sodium carbonate, while the latter forms sodium hydroxide. The chemical reactions which take place may therefore be represented in simple form by the follow- ing equations: 2Na,0, + C = Na^CO, + Na,P ; NaA+2H = 2NaOH. Fig. 62. It has been found by experience as well as by calculation from the heats of formation of the compounds given in the equations, that very nearly 73 per cent of the heat produced in the reac- tion is due to the oxidation of the carbon to carbon dioxide and of the hydrogen to water. The factor 0.73 is, therefore, used as a correction constant in the final calculation. 6oi. The Sodium Peroxide. — Another soiu'ce of error is met with in the fact that sodium peroxide readily absorbs water, forming the compound Na202.2H20. When heated this compound decomposes THE PARR CALORIMETER. 399 into sodium hydroxide, oxygen, and water, with evolution of heat. In one experhnent 10 grams of sodium peroxide absorbed 0.5 gram of water during an exposure of 1 hour to the air. The presence of this amount of water caused an increase of 0.194 degree dur- ing the combustion. A special grade of sodium peroxide is, therefore, manufactured for the Parr calorimeter. It is carefully protected from exposure to the air and is sold in hermetically sealed tin cans. When desired for use, tlie contents of the can are transferred to a glass jar pro- vided with a tightly fitting lid. Glass-stoppered bottles are not satisfactory. The "Lightening" or "Putnam" jar, which closes with a lever clamp, has been found well adapted for this purpose. The sodium peroxide must also be finely divided. It is not advis- able to attempt to grind ordinary samples of the peroxide because of the unavoidable exposure to the air. 6o2. Construction of the Instrument. — Fig. 6.3 shows the con- struction of the cartridge which is made of brass. Both top and bottom are held in place by means of the cylinders or caps F and E which are screwed to the body of the cartridge. The joints are made air-tight by means of rubber washers. An insulated wire J passes down through the stem and at / is connected to the wire H by means of a piece of soft iron wire which may be fused by means of an electric current and thus ignite the combustible mix- ture in the bottom of the cylinder. Fig. 64 shows the cartridge D in place in the instrument. It rests on a pivot at F and can be rotated by means of the pulley P. The blades attached to the center of the cartridge tend to force the water downwards within the cylinder E, and produce an upward current outside of this cylinder, thus keeping the tem- perature of the water uniform. Both the small cylinder E and the metaUic vessel A are nickel-plated and highly polished. Such surfaces neither radiate nor absorb heat readily. Loss or gain of heat is still further prevented by the two vessels C and B of indurated fiber, as well as by the air spaces between them. A double close-fitting lid of the same material is also provided. Two liters of water for the absorption of the heat are placed in the vessel A, the temperature being measured by the thermometer T. Radiation of heat cannot be wholly prevented, however, so 400 CALORIFIC VALVE OF FUELS. that if the temperature of the water is below that of the sur- rounding objects, it will gain heat, while if it is above that of the surrounding objects, it will lose heat. Error from this source is compensated for by starting a determination with the water as much below atmospheric temperature as it will rise above this Fig. 63. Fig. 64. temperature after the combustion. The rate at which the water increases or decreases in temperature may also be determined, so that the amount, of heat gained or lost may be calculated. 603. Water Equivalent of the Instrument. — All of the heat produced by the combustion of the fuel does not enter the water. The combustion cartridge, the containing vessel, as well as the THE PARR CALORIMETER. 401 thermometer, absorb and retain a portion of the heat. The amount of heat lost ui this manner majj- be easily calculated. These parts of the instrument are weighed and the water equiva- lent calculated by multiplying the total weight by the specific heat of the material used* The cartridge and containing vessel in the Parr calorimeter are made of brass, specific heat 0.092. The thermometer is made of glass and mercury and only a por- tion is immersed in the water and is affected by the changes in temperature of the water. In order to ascertain the water equivalent of the thermometer, advantage is taken of the fact that the water equivalent of equal volumes of mercury and glass is the same. This follows from the fact that the product of the specific heat and specific gravity of those substances is the same: 2.5X0.19=0.47 for glass and 13.6X0.034=0.46 for mercury. That is, one cubic centimeter of either mercury or glass is raised one degree by .46-.47 calories of heat. The volume of that por- tion of the thermometer which is immersed in the water is ascer- tained by immersing the thermometer in a graduated cylinder partially filled with water to the same point to which it is immersed when used in the calorimeter. The volume of the portion of the thermometer which is immersed is then read directly on the cylinder. 604. Accelerators. — The use of sodium peroxide alone has not been found entirely satisfactory. The addition of small amounts of substances . like potassium chlorate or a persulphate is advan- tageous because the oxygen in these compounds is more easily liberated, and this assists in starting the reaction which proceeds easily when the temperature is raised sufficiently to decompo'se the sodium ];)eroxide. The same object is attained by the addition of substances which are more readily oxidized than the coal or other substance to be burned. Tartaric acid is such a substance. Ammonium compounds are also very easily oxidized on account of the large amount of hydrogen present. Ammonium persulphate has been found to be admirably adapted for this purpose because * Specific Heats. Brass, 0.092-0.094 Nickel, 0.110 Glabs, 0.190 Silver, 0.057 Mercury, 0.034 Platinum, 0.033 402 CALORIFIC VALUE OF FUELS. it contains both the ammonium radical and also the easily liberated oxygen. The term accelerator is used to designate this class of substances. 605. Correction Factors. — A correction factor must be intro- duced into the calculation when accelerators are used. The heat of formation of potassium chlorate is —11.9 calories for the gram molecular weight when formed from KCl and IJ 0,, while the same constant for sodium peroxide is 119.8 when formed from NajO and 0. When ammonium persulphate is used, heat is evolved by the combustion of the hydrogen. The correction for J gram of potassium chlorate is — .040° C, and for J gram of a mixture of one part of ammonium persulphate and two parts of potassium persulphate, — .197° C. This persulphate mixture has been found especially advantageous for use with anthracite coal. No appreciable oxidation of nitrogen takes place in the Parr calorimeter. A correction must be introduced for sulphur and for ash. When using J gram of coal, these corrections are — .006° C. for each per cent of sulphur, and —.001° C. for each per cent of ash. A correction of —.008° C. must be made for each length of 10 mm. of fine iron wire. Another correction should also be introduced when bitu- minous coals are burned, on account of the oxygen or hydroxyl compounds present. No such correction is necessary for anthra- cite coals. The corrections may be summaiized as follows: Corrections for average anthracite coals, using J gram of coal, 10 grams of NajOa, and J gram of persulphate mixture: I gram persulphate mixture . 0.197 1 per cent of sulphur 0.006 7 " " "ash 0.007 Fuse wire .008 Total 0.218° C, or 0.392° F. Corrections for average semi-bituminous coal (not over 25 per cent volatile matter), using J gram of coal, 10 grams of NaA. and h gram KCIO3: THE PARR CALORIMETER. 404 J gram KCIO3 .040 1 per cent sulphur .006 7 " " ash 0.007 4 " " combined water. .. .013 Fuse wire .008 Total 0.074° C, or 0.133° F. Corrections for average bituminous coal (over 25 per cent volatile matter): i gram KCIO3 0.040 2 per cent sulphur 0.012 10 " " ash 0.010 10 " " combined water... 0.033 Fuse wire .008 Total 0.103° C, or 0.185° F. EXERCISE 71. Determination of the Calorific Value of Coal by Means of the Parr Calorimeter. 606. Electric Ignition. — Grind the coal fine enough to pass through a sieve of .25 mm. mesh or 100 to the square inch. Weigh out exactly J gram and dry for 1 hour in an air-bath heated to 104°-107° 0. If the coal contains less than 2^ per cent of water, the drying is uimecessary. By adding cold or hot water, bruig the temperature of two liters of water to about 2° F. below room temperature. Measure out exactly two liters of this water and place in the calorimeter. Transfer one full measure (10 grams) of sodium peroxide to the cartridge, which must be absolutely dry. Transfer the coal to the cartridge, brushing off the watch-crystal carefully. If the coal is anthracite, add exactly ^ gram of the persulphate mixture. If a bituminous coal is being tested, add exactly J gram of potassium chlorate. Mix the contents of the cartridge thoroughly by stirring with a rod which is afterwards well brushed off. The more recent instruments are supplied with a false cap with burnished face, which is fastened temporarily in place, and the contents of the cartridge mixed by shaking. After thorough mixing in this manner, the provisional cap is removed and the cap with stem and terminals substituted. In this way no chemical can come in contact with 404 CALORIFIC VALVE OF FUELS. the base of the terminals and produce a short circuit. The cartridge is then tapped Ughtly to shake all the material from the upper part of the cylinder and to settle the contents. When the electric method of ignition is used, the fine iron wire (34 American gauge) is wrapped firmly with good contact around the ends of the terminals, so as to leave a U-shaped loop extending about three-quarters of an inch below the terminals and well into the chemical mix- ture. If the loop is too long, it is not readily fused. The proper length of wire may be ascertained by a preliminary test without screwing the top on the cartridge. The current required is from 2 to 4 amperes and is readily obtained by placing in parallel four to eight 16-candle-power lamps in an ordinary lighting circuit of 110 volts. The current may also be obtained from 6 to 8 dry cells connected in series. The adjustment of the wire must be made before the sodium peroxide is placed in the cartridge, as it rapidly absorbs moisture. When the top has been pressed firmly in place and screwed down, the vanes are adjusted and the cartridge placed in the can of water. The cover is adjusted and the thermometer inserted so that the bulb is about half-way down in the water. The pulley on the stem is connected with the motor and the cartridge rotated to the right, or as the hands of a watch turn, so as to deflect the water downwards in the can. The ther- mometer is read at three-minute intervals until the readings are constant or the rise in temperature is constant. These readings must be made with the greatest care and tenths of the smallest divisions estimated as care- fully as possible. After igniting the charge by bringing the spring in contact with the revolving stem of the cartridge, the temperature rises rapidly at first, and then more slowly, until the maximum temperature is reached. The readings are made at three-minute intervals as before, until the tempera- ture is constant or it begins to fall. The cartridge is removed and the parts unscrewed and placed in a dish of hot water to dissolve the melt. It is then rinsed thoroughly with distilled water and dried, when it is ready for another determination. 607. Ignition by Hot Wire. — If the ignition is made by means of a hot wire, the method of filling the cartridge and rotating it until the tempera- ture is constant is the same as when the electric ignition is used. The short piece of soft iron wire is placed on a wire gauze and heated to redness in a Bunsen-burner flame and dropped quickly into the opening at the upper end of the valve. With the pincers the valve is now pressed completely down and released with a quick motion, so as to prevent the escape of heated air from the interior. During this operation the cartridge is kept revolving. The remainder of the operation is the same as when the electric method of ignition is used. 608. Calculation. — The various correction factors, as given on page 402, are subtracted from the total rise in temperature. The remainder is multi- plied by 3117. This factor is obtained from the following formula: CALCULATIOn. 405 Total water eq v. X 0.73 X Corrected rise _, , .^ , ,,, . ' — j-f — r— -j = Calorific value. Weight or luel taken Substituting the known values, 2135 X 0.73 Xj- 0.5 which reduces to 3117 Xr. The product will be the calorific power of the coal in B. T. U. per pound of coal. CHAPTER XXIX. WATER ANALYSIS. A CHEMICAL analysis of water is carried on by very different methods, depending on the use to which the water is to be put. These methods may conveniently be divided into two classes. In the first class are placed those methods by which the potability of a water is determined, while in the second class are included the methods by which the suitability of a water for industrial PURPOSES is determined. As the requirements of the various industries are somewhat diverse, the methods used in testing a water for industrial purposes cannot be so definitely stated as those employed when the water is to be used for domestic consumption. In the latter case no poisonous substances may be present even in minute traces. The absence of lead, copper, etc., being insured, we may conveniently classify the methods commonly used into those by which the amount of organic impurities and those by which the amount of inorganic impurities are determined. As the commonly occurring inorganic salts, such as sulphates, chlorides, and carbonates of lime, magnesia, the alkalies, iron, etc., are not objectionable unless present in excessive amounts, only the total amount of the calcium and magnesium salts present is determined and reported as hardness of the water. If the water is classed as a MINERAL WATER, and is therefore to be used because of the inor- ganic salts present, a complete analysis of these must be made, as well as determinations of the organic impurities. The most important determinations for indicating the presence of organic impurities are the estimation of the amount of nitrogen in its various forms and chlorine. SANITARY ANALYSIS. 609. Bacteriological Examination. — As has been said, the sani^- tary analysis consists in the determination of the organic and the inorganic impurities present. The most serious ill effect of 406 SANITARY ANALYSIS. 407 using impure water undoubtedly arises from taking into the system disease germs which exist in the water. It would seem, therefore, that the examination of the water should primarily be entrusted to the bacteriologist. Unfortunately, it is extremely difficult, if not impossible, in the present condition of bacterio- logical science, to detect and identify the specific disease germs even if they undoubtedly exist in a given water. This is largely due to the fact that the great majority of the germs that exist in water are either harmless or even beneficial. The continual drink- ing of a water containing relatively few disease germs by a person in whose system the germ can live, ultimately results in the ac- cumulation of the germs in sufficient quantity to produce disease. The routine bacteriological tests generally made on water consist of a count of the total number of bacteria present and a test for the presence or absence of B. coli. These bacteria are almost invariably present in the intestines of human beings and other warm-blooded animals. Their presence in water is regarded as evidence, of contamination with sewage, which almost always contains disease germs excreted from the body, thus rendering the water unwholesome and dangerous. The presence of an ex- ces,si\-f number of bacteria in water also renders it deleterious to health. 6io. Source of Impurities. — The chemical examination gives indirect evidence of the presence of disease germs by revealing the presence and the amount of organic matter in the water, as well as whether it is of animal or vegetable origin, and the extent to which it has undergone decomposition. As sewage is undoubt- edly the most common source of contamination of water with pathogenic germs, the evidence furnished by the chemist of its presence is quite sufficient to condemn a water for household use, even if the specific disease germ is not found. If the water is liable to be contaminated with sewage, it is almost certain that sooner or later it will be rendered poisonous by the presence of the germs of disease. It is evident that after organic impurities have been found in a water it is important to know their source before a deci- sion can be reached as to the potability of a water. Whenever possible, therefore, the chemical examination should he supplemented 408 WATER ANALYSIS. by an examination of the source of the water-supply; that is, whether it is a spring, lake, river, well, etc. If surface or underground drainage can mingle with the water, the entire region drained should be examined, if possible, for sources of contamination. 6ii, The Samples must be taken with the greatest care. As the impurities at best are present in very small amount, careless wash- ing of the bottle, or collecting water contaminated after leaving the source of supply, may easily introduce as much impurity as was originally present. The bottles used should be thoroughly cleaned by means of strong dichromate solution and then thor- oughly rinsed out with distilled water and drained. They should then be rinsed out with the water to be examined. If water from a city supply is to be examined, it must be allowed to flow from the pipes a considerable time before filling the bottle, unless the water is to be examined for lead or other contamination from the pipes. If a WELL-WATER is to be examined, the stagnant water in the pump must first be completely removed. A sample of spring, river, or lake water is best taken by allow- ing the bottle to fill after immersion some distance under the sur- face of the water, so as to avoid the floating surface contamination. The TIME which may elapse between the collection of a sample and the beginning of its analysis varies with the character of the sample and other conditions. The following limits are generally safe: for fairly pure surface-water, 24 to 48 hours; and for normal ground-water, 48 to 72 hours. Polluted water requires analysis within 12 hours. A bacteriological examination should in all cases be made within 6 hours. If for any reason the examination is delayed, the sample must be preserved on ice. PHYSICAL EXAMINATION. The physical examination includes observations of the tem- perature, general appearance, color, turbidity, and the odor in hot and cold samples. 6i2. The Temperature should be taken at the time of collec- tion to the nearest 0.5° Centigrade. 613. Turbidity. — The general appearance of the water should be determined by inspection in strong light after standing several . PHYSICAL EXAMINATION. 409 hours. Substances remaining in suspension are then classified as turbidity on standing, and substances settling to the bottom as sediment. Instead of expressing the turbidity as none, slight, dis- tinct, decided, etc., it is advisable to use the numerical standards recently introduced by which the turbidity is expressed in parts of silica per one million. The stock suspension of silica is prepared as follows : Diatomaceous earth, as free from amorphous silica and sponge spicules as may be obtained, is washed with water to free it from any soluble salts, and ignited to remove any organic matter. It is then treated with warm dilute hydrochloric acid, and washed with successive portions of distilled water until free from acid. The material, now composed of practically p\ire diatomaceous frustules, is ground to an impalpable powder in an agate mortar, sifted through a sieve that has 200 meshes to the inch, and dried ia a desiccator. One gram of this prepared diatomaceous earth in 1 liter of distilled water gives the stock suspension which has a turbidity of 1000, i.e., contains 1000 parts of silica per million. Standards for comparison should be prepared from this stock suspension by dilution with distilled water. For turbidity readings below 20, standards of 0, 5, 10, 15, and 20 are kept in gallon bottles made of clear white glass; for readings above 20, standards of 20, 30, 40, 50, 60, 70, 80, 90, and 100 are kept in 100- c.c. Nessler jars, approximately 20 mm. in diameter. Comparison of the water under examination with the stand- ards is made by viewing them sidewise towards the light, looking at some object and noting the distinctness with which the margins of the object can be seen. The standards are kept stoppered, and both sample and standards are thoroughly shaken before making the comparison. In order to prevent any bacterial or algae growths from appearing in the standards, a small amount of bichloride of mercury may be added to them. 614. The Color of the water is measured by comparison with the color of a solution of platinum and cobalt and is expressed in parts of platinum per million. A standard solution which has a color of 500 is made by dissolving 1.246 grams potassium platinic chloride (PtCl4.2KCl), containing 0.5 gram platinum, and 1 gram of crystallized cobaltous chloride (C0CI2.6H2O), containing 0.25 gram of cobalt in water, with 100 c.c. concentrated hydrochloric 410 WATER AKALYSIS. acid, and making up to 1 liter with distilled water. By diluting this solution standards are prepared having colors of 0, 5, 10, 15, 20, 25, 30, 35, 40, 50, 60, and 70 parts per million. These are kept in 100-c.c. Nessler jars of such diameter that the liquid has a depth between 20 and 25 cm. and is protected from dust. The color of a sample is observed by filling a similar tube with the water and comparing it with the standards. The observation is made by looking vertically downwards through the tubes upon a white surface placed at such an angle that light is reflected upwards through the column of liquid. The reading is recorded to the nearest unit. Waters that have a color darker than 70 are diluted before making the comparison, in order that no difficulties may be encountered in matching the hues. • Water containing matter in suspension is filtered until no visible turbidity remains. 615. The Odor of the water both hot and cold is observed. To obtain the odor of the cold water the sample in the large collecting- bottle, which should be about one-half or two-thirds full, is shaken vigorously, the stopper removed, and the nose immediately placed to the mouth of the bottle. To obtain the odor of the hot water it is poured into a moderately large beaker until it is about one- third full. The beaker is covered with a watch-crystal and the water is rapidly brought to just below boiling. It is immediately removed from the source of heat and allowed to cool for about five minutes. The beaker is then shaken with a rotary motion, the watch-crystal sUpped to one side, and the odor observed. The odor is expressed by such terms as the following: vegetable, aromatic, grassy, fishy, earthy, mouldy, musty, disagreeable, peaty, and sweetish. The intensity of the odor is usually indicated by the numbers 1, 2, 3, 4, and 5, 1 indicating very faint; 2, faint; 3, distinct; 4, decided; and 5, strong. CHEMICAL EXAMINATION. The first determination carried out should be that of the amount of nitrogenous organic matter present, as this material is continually undergoing changes, and the condition in which the nitrogen exists is taken as an index of the source of the contamina- tion, and the time which has elapsed since it has been introduced. NITROGEN. 411 6i6. The Nitrogen may exist as so-called albuminoid nitro- gen, AMMONIA, NITRITES, or NITRATES. The all.umiDoid nitrogen is undoubtedly present as part of a class o: complex organic com- pounds, called proteins, which may have teen derived from plant, but more likely animal, tissues which are present in the water. These tissues undergo a more or less rapid decomposition or putrefaction by which the nitrogen assumes for the most part the condition of free ammonia, which may then be oxidized to nitrites or nitrates. The presence of the unstable nitrites generally indi- cates the presence of sewage or other organic matter in a state of decomposition. While nitrates may be produced by the oxidation of organic nitrogen, this is by no means the only source, since the atmosphere and the soil contain considerable amounts of nitric acid or nitrates, which are dissolved by the rain. The amount of nitrates in a surface-water increases with the density of popula- tion in the region drained. 617. Determination of Free and Albuminoid Ammonia. — Ammonia is determined by means of the so-called Nessler solu- tion, to which a brown color is given by very small amounts of ammonia or ammonium salts. The reagent consists of a strongly alkaline solution of the double iodide of potassium and mercury, Hgl2.2KI. Ammonia produces in this solution a dark-brown col- oration or precipitate of dimercuric ammonium iodide, NHgJ.HjO, according to the following equation: 2(Hgl2.2KI) + NH3+ 3K0H = NHg J.H^O-h 7KI + 2H,0. The depth of the color produced is proportional to the amount of ammonia present, so that by comparison with solutions con- taining known amounts of ammonia, the amount of ammonia present in a given solution may be determined, the solution con- taining the known amount of ammonia being so adjusted that its color exactly niatches the solution containing the unknown amount. 618. Preparation of Nessler Solution.* — Dissolve 62.5 grams of potassium iodide in 260 c.c. of distilled water. Make a saturated solution of mercuric chloride by dissolving 35 grams of mercuric * D. D. Jackson, Tech. Quar., XIII, No. 4, Dec, 1900, p. 320. 412 WATER ANALYSIS. chloride in 500 c.c. of hot water. Allow the mercuric chloride solution to cool and to 250 c.c. of the potassium iodide solution add the cold mercury solution slowly with stirring, until a sUght permanent precipitate is formed, which will require about 400 c.c. Dissolve the red precipitate by adding the remaining 10 c.c. of the potassium iodide solution. Again add mercuric chloride solu- tion drop by drop until a slight permanent precipitate remains. Then add 150 grams of potassium hydrate dissolved in 250 c.c. of water and dilute the solution to 1 hter. Mix thoroughly and allow the solution to stand until any precipitate formed has settled, leaving a pale straw-colored solution which is siphoned or decanted off into another bottle for use. The solution improves with age. Its sensitiveness is increased by adding mercuric chloride and decreased by adding potassium iodide. The 2-c.c. portions used for each test must be measured quite carefully, as the depth of color produced with a given amount of amimonia depends to a certain extent upon the amount of Nessler solution used. The following solutions will also be required: 619. Alkaline Potassium Permanganate Solution. — Dissolve 200 grams of soUd caustic potash * and 8 grams of • potassiimi permanganate in 1250 c.c. of distilled water. Boil the solution down to 1 Uter and preserve for use. 620. Sodium Carbonate Solution. — ^Dissolve 50 grams of the pure salt in 300 c.c. of ammonia-free water. 621. Standard Anmionia Solution. — Dissolve 1.9094 grams of pure dry ammonium chloride in ammonia-free water and dilute to 500 c.c. 1 c.c. of this solution will contain 1 mg. of nitrogen as ammonia. 10 c.c. should be diluted to 1 Hter with ammonia-free water. 1 c.c. of this solution will contain .01 mg. of nitrogen and is the standard solution used. 622. Ammonia- free Water. — Ordinary distilled water gener- ally contains considerable amounts of ammonia. It may be tested by adding 2 c.c. of Nessler solution to 50 c.c. of the water placed in a Nessler tubj;. These tubes are made of colorless glass and contain 50 c.c. when filled to a mark near the top. A tube of * K caustic potash free from carbonates is used, the bumping so apt to occur while boiling an alkaline solution will be very much reduced. NITROGEN. 413 suitable proportions has a diameter of 1.7 cm. and a height of 21 cm. to the 50-c.c. mark. If after standing five minutes no trace of a yellow tint is developed in the water it may be used without purification. Generally, however, it must be freed from ammonia. This may be done by nearly filling a large glass retort or a suitable copper or tin vessel with distilled water to which some of the alka- line potassium permanganate solution has been added. A glass or block-tin condenser is attached and the water is distilled until 50-c.c. portions tested with the Nessler solution are found free from ammonia. The distillate is then collected in a clean bottle for use. The distilled water may also be freed from ammonia by adding enough bromine water to color it distinctly and then boiling until the excess of bromine is expelled. The boiling may be omitted if a drop of caustic soda is added to the water tinted with bromine, and after standing ten minutes the undecomposed hypobromite removed by adding a little potassium iodide. The preparation of the ammonia-free water and all ammonia determinations must be carried out in a room whose atmosphere is free from ammonia and ammonium salts. The ordinary labora- tory is, therefore, not suited to the purpose. DETERMINATION OF FREE AMMONIA. 623. Distillation. — ^A liter flask is connected by means of a short piece of rubber tubing or a rubber stopper with a Liebig condenser. A tubulated glass retort capable of holding a liter of solution is also very convenient for this purpose, as the reagent can be introduced through the tubiilure. The stem of the retort may be introduced into the end of the condenser-tube and a tight joint made by means of a short piece of rubber tubing. The dis- tillate is led into the Nessler tubes by means of an adapter. Although the apparatus is thoroughly cleaned before being set up, to completely remove ammonia about 200 c.c. of ammonia- free water is introduced into the retort and 10 c.c. of the sodium carbonate solution added. The solution is boiled and the first 50 c.c. of the distillate discarded while the second 50-c.c. portion is tested for ammonia by adding 2 c.c. of the Nessler solution. If 414 WATER ANALYSIS. no color develops after standing five minutes the apparatus may be considered free from ammonia and the water to be analyzed may be introduced into the retort. Unless the water contains an unusual amount of ammonia 500 c.c. will be found to be a convenient amouivc. A preliminary test may be made by adding 2 c.c. of Nessler solution to 50 c.c. of the water to be tested. The color developed is compared with the color produced by adding 0.5 c.c. of the standard ammonia solution to 50 c.c. of ammonia- free water. If the color produced in the tube containing the water to be analyzed is the darker, less than 500 c.c. of the water should be placed in the retort and the difference made up by ammonia- free water. 624. Nesslerizing. — The distillate is collected in the Nesslerizing tubes, which are placed in a jack in order after being filled to the 50-c.c. mark. The flame of the Bunsen burner should be so adjusted that 50 c.c. are distilled in about eight minutes. As soon Fig. 65. as the second tube is filled and has come to room temperature it is Nesslerized. 2 c.c. of the Nessler solution is added. Several clean Nessler tubes are taken and carefully measured amounts of the standard ammonia solution added from a burette, such as 1 c.c, 2 c.c, 3 c.c, etc. The tubes are then filled with ammonia- Iree water to the 50-c.c. mark, and 2 c.c of the Nessler solution added to each tube without stirring. The colors are compared by holding two tubes in the hand FREE AMMONIA. 415 and looking vertically downward through the length of the tubes, which are held over a white paper or piece of porcelain placed at an angle in front of a window or in a white light so as to reflect the Ught upward. If the color of a tube does not match that of any of the standards made, others must be made with the amount of standard ammonia solution judged necessary until the color of the unknown ammonia-tubes have been matched. After adding the Nessler solution, ten minutes must elapse before comparing the color, as it does not develop instantly. After being fully formed, the color is quite permanent, so that a set of standard tubes may be used during a day's work without error. A permanent set of standard tubes has been devised by D. D. Jackson.* The temperature of the water may vary between 18° and 25° without material change in depth of color. If the ammonia in the second tube corresponds to 1 c.c. or less of the standard solution, further distillation may be dispensed with. Otherwise 50-c.c. portions must be collected until this limit is reached. Usually three 50-c.c. portions are collected. If the ammonia in the second tube corresponds to 2 c.c. or less of the standard solution, the first tube maj' be Nesslerized in the same manner as the second. If the second tube contains a greater amount of ammonia, a measured portion of the first tube must be taken and diluted with ammonia-free water and Nesslerized. Or the contents of the first and third tubes may be mixed and 50 c.c. of the mixture Nesslerized. In no case should the color of a tube be read which responds to more than 10 c.c. of the standard ammo- nia solution. 625. Calculation. — The various amounts of ammonia found in the different tubes are added and the result stated in parts per million of nitrogen. If, for instance, the first tube was found to match a tube containing 9 c.c. of standard ammonia solution, the second 3 c.c, the third 1 c.c, and the fourth free from ammo- nia, then, as 1 c.c. is equal to .01 mg. of nitrogen, the total amount of ammonia as nitrogen would be .09-|-.03-f-.01 = .13 mg. As 500 c.c. of the water was taken and as 1 mg. is the millionth part by weight of a liter of water, the amount of nitrogen present as free ammonia in the water would be .25 part per million. ' * Tech. Quar., Vol. XIII, No. 4, Dec, 1900, p. 320. " 416 WATER ANALYSIS. DETERMINATION OF ALBUMINOID AMMONIA. 626. Distillation. — ^When all the free ammonia has been dis- tilled off, which is usually the case when 150 c.c. of distillate has been collected, 50 c.c. of the alkaline permanganate solution is added and the distillation continued, 50-c.c. portions being col- lected in the Nessler tubes and the ammonia determined and cal- culated exactly as given for free ammonia. The albuminoid ammonia does not always come off as readily as the free ammonia. This is undoubtedly due to the greater or less difficulty of decomposing organic matter from different sources. Organic matter derived from vegetables is generally decomposed with more difficulty than the contamination derived from partially decomposed animal refuse as found in sewage. Each 50-c.c. por- tion of the distillate for albuminoid ammonia must therefore be tested to be sure that all of the ammonia has been obtained, the solution in the retort being boiled nearly dry. As it is almost impossible to obtain all of the albuminoid ammonia from many waters, especially highly colored surface-waters, the practice is growing of collecting five 50-c.c. portions of the distillate for the albuminoid ammonia. In this way comparative results are ob- tained which are of almost as great value as if the total amount of albuminoid ammonia were obtained in each case. Much time and labor are also saved. About one-half of the total organic nitrogen will generally be obtained in the 250 c.c. of distillate. Total Organic Nitrogen. — ^The total amount of organic nitrogen may be obtained by decomposing the organic matter by evaporat- ing a pwrtion of the water with sulphuric acid, adding copper sul- phate and digesting according to the method of Kjeldahl, then neutralizing with sodium carbonate and distilling off the ammonia and Nesslerizing. DETERMINATION OF NITRITES. A colorimetric process similar to that used to determine ammonia is in use for the estimation of the amount of nitrites present. Of the several methods which have been suggested, that of Griess will be given. It depends on the fact that a red NITRITES. 417 coloration is produced when a solution of sulphanilic acid and NAPHTHYLAMiNE ACETATE are added to an acidified solution of a nitrite. This test is said to be capable of showing one part of nitrogen as nitrite in one thousand million parts of water. The following reagents must be preapred: 627. Sulphanilic Acid. — Dissolve 0.8 gram of the acid m 100 c.c. of hot water. 628. a-Naphthylamine Acetate. — ^Dissolve 0.8 gram of the salt in 100 c.c. of water and 1 c.c. cf strong hydrochloric acid. De- colorize the solution by adding a little powdered charcoal. Place the solution in a glass-stoppered bottle with the charcoal and filter as required for use. 629. Standard Solution of Sodium Nitrite. — Pure silver nitrite is prepared by mixing a warm concentrated solution of 8 parts of sodium nitrite with a warm concentrated solution of 16 parts of silver nitrate. The precipitate, when cold, is washed with cold water and quickly dried over a water-bath with as httle exposure to the fight as possible. Weigh .1091 gram of the dry nitrite and dissolve in warm water. Add a slight excess of pm-e sodium chloride and dilute to 1 liter. After allowing the silver chloride to settle, draw out 10 c.c. and dilute to 1 liter. This will give a standard solution, 1 c.c. of which wiU be equal to .0001 nag. of nitrogen as nitrite. 630. The Determination is carried out as follows: 100 c.c. of the water to be examined are placed in a cylinder capable of hold- ing 100 c.c, which is made of the same kind of colorless glass as the Nessler tube, although a shorter and wider tube is found con- venient, a suitable size being 3 cm. in diameter and 13.2 cm. from the bottom to the 100-c.c. mark. Three drops of acetic acid are added and then 2 c.c. of the sulphanilic acid solution and 2. c.c. of the solution of acetate of naphthylamine. After mixing well, the tube is tightly stoppered and allowed to stand for ten minutes until the color is fuUy developed. In the meantime the tubes are prepared containing known amounts of nitrite, 1 c.c, 2 c.c, 3 c.c, etc., of the standard nitrite solution being added to tubes containing 100 c.c, of ordinary distilled water or, still better, water redistilled as directed for ammonia-free water. After the addition of 2 c.c. of each reagent to each of the color solutions and 418 WATER ANALYSIS. a drop of acetic acid, and allowing the tubes to stand for twenty minutes, the colors of the known and the unknown solutions are compared and the amount of nitrogen present per million parts is computed. If 100 c.c. of the water is taken and the color is matched by the tube containing 3 c.c. of the standard nitrite solution, the water contains .03 part per million of nitrogen as nitrite. As this test is extremely delicate, it is advisable to make a blank determination to detect accidental contamination which at times arises from nitrous fumes in the air. DETERMINATION OF NITRATES. This determination also is carried out by a colorimetric process, the yellow ammonium salt of nitrophenol-sulphonig ACID * being produced when the water to be examined is evap- orated to dryness and treated with phbnol-sulphonic acid and ammonia. The following reactions take place: CeH4.0H.S03H+ HNO3 = CeH3.0H.S03H.N02+ H^O ; C,H3.0H.S03H.NO,+ NH,OH=CeH3.0NH,.S03H.N02+H,0. The intensity of the color developed is compared with the color produced by the same process in water containing a known amount of nitrate. The following solutions are required: 631. Phenol-sulphonic Acid. — In 370 grams of pure concen- trated sulphuric acid 30 grams of pure phenol are dissolved by heating in the water-bath for six hours. 632. Standard Potassium Nitrate Solution.— 0.722 gram of pure dry potassium nitrate is weighed out and dissolved in 1 liter of distilled water. 10 c.c. of this solution are e-\'-aporated to dryness, 2 c.c. of the phenol-sulphonic acid added, and quickly and thor- oughly mixed with the residue by means of a glass rod. The mixture is dissolved in distilled water and the solution diluted to 1 Hter. The color-tubes for comparison are prepared by measuring out portions of this solution into 100-c.c. tubes of the size used for the nitrite determination, adding 5 c.c. of strong ammonia and * Jour. Am. Chem. Soc, XXVI, 388. NITRATES. 4 19 diluting to 100 c.c. These standards are permanent for several months if the tubes are well corked with stoppers from which the coloring matter has been extracted by boihng water. 1 c.c. of the dilute nitrate solution is equal to .001 mg. of nitrogen as nitrate. 633. The Determination is carried out as follows: 25 c.c. of the water to be analyzed are taken, or, if the amount of nitrate present is small, 50 or even 100 c.c. are taken and evaporated nearly to dryness. Remove from the water-bath and allow the last few drops to evaporate at room temperature in a place protected from dust. The residue is treated with 1 c.c. of the phenol- sulphonic acid and well mixed by stirring with a glass rod. 10 c.c. of distilled water are added and then enough strong caustic potash solution to make it alkaline. The solution is washed into a 100-c.c. tube and diluted to the mark with distilled water. The depth of color is now compared with that of the standard color- tubes containing various amounts of the standard nitrate solution. When the color has been exactly matched, the parts per mil- lion of nitrogen as nitrate are calculated. If, for instance, 50 c.c. of the water were taken and the color was matched by the standard color-tube containing 3 c.c. of the dilute nitrate solution, then 50 c.c. of the water would contain .003 mg. of nitrogen as nitrate, which would be .06 part per million. If 25 c.c. of a water had been taken, the result would be .12 part per million if the color produced were matched by a color-tube containing 3 c.c. of the standard nitrate solution. DETERMINATION OF THE OXYGEN REQUIRED TO OXIDIZE ORGANIC MATTER. The methods already given have shown the amount of nitro- gen present in a given water in the four forms: ammonia, nitrous and nitric acid, and albuminoid substances. A very simple method is in use for the determination of the amount of carbon present. It consists in ascertaining the amount of per- manganate required to oxidize the organic matter present. The water is boiled for ten minutes with sulphuric acid and an excess of standard potassium permanganate solution. The excess of the permanganate is then titrated back with standard oxalic acid. 420 WATER ANALYSIS. i 634. Errors. — It is evident that by this process any substance other than carbonaceous material capable of being oxidized under the conditions will also reduce the permanganate. On the other hand, the oxidation of the organic matter present may not be complete. It has indeed been shown that the extent to which different kinds of organic matter are oxidized varies greatly. As the method, therefore, is an empirical one, the conditions of the oxidation must be rigidly adhered to. If this is done, very con- stant results are obtained and the method has shown itself of value in forming a judgment of the amount of organic matter in the water, as other oxidizable substances are rarely present. Ferrous iron and sodium chloride are the only common excep- tions. The iron is quickly precipitated as ferric oxide after a few hours' exposure of the water to the air. If a considerable amount of sodium chlobidb is present, a cor- rection must be applied as found by a blank test. Sodium chloride is added to distilled water until there is present the amount of chloride which has been foimd in the water being examined. The oxidation with the permanganate is then carried out in the same manner as when the water was tested. The amount of perman- ganate required is subtracted from that used in testing the water. 635. The Following Solutions are required and are made with water free from organic matter. This is prepared by adding a little potassium permanganate and caustic alkali to distilled water and redistilling it. The first portions of the distillate are discarded. The steam or distilled water should not come in contact with rubber or other organic matter. A suitable distilling apparatus may be made by connecting a retort with a Liebig condenser by inserting the tube of the retort into the inner tube of the condenser as shown in Fig. 65, page 414, and packing the joint with ignited asbestos instead of the rubber connection. 636. Potassium Permanganate Solution is made by dissolving .3953 gram of the pure salt in 1 liter of the redistilled water. 1 c.c. of this solution will contain 0.1 mg. of available oxygen. 637. Oxalic Acid Solution is made by dissolving .7875 gram of the pure recrystallized and carefully dried salt in 1 liter of water. If this salt as well as the permanganate are pure, these solutions will be exactly equal. They should be titrated against each other NITRATES. 421 by measuring out 10 c.c. of the oxalic acid, adding 10 c.c. of dilute sulphuric acid and 200 c.c. of the redistilled water. The whole warmed gently and the potassium permanganate solution intro- duced tuitil a faint, permanent pink coloration is produced. Neither the oxalic acid nor the permanganate solutions are permanent and must be made up fresh or standardized from time to time. Dilute sulphuric acid is made by diluting 1 part of concentrated acid with 3 parts of the redistilled water. 638. The Determination is carried out by placing 100 c.c. of the water to be tested in a flask and adding 10 c.c. of the dilute sulphuric acid. The standard potassium permanganate solution is introduced from a burette until the water is colored. It is heated to boiling and more permanganate is introduced if the color fades. If brown flakes of manganese dioxide are formed, more dilute sulphuric acid must be added. The boiling is continued for thirty minutes. An excess of oxalic acid is then introduced and titrated back with the permanganate solution. The volume of oxalic acid introduced is subtracted from the total amount of the permanganate solution used, and the number of milligrams of oxy- gen consumed is calculated. The result is stated in parts of oxygen required per million parts of water. If the amount of organic matter in the water is considerable, less than 100 c.c. must be taken for this determination. The volume should in each case be made up to 100 c.c. with the redis- tilled water. As the success of this determination is assured only by working under absolutely fixed conditions, the determination should be repeated several times until constant results are obtained. Dissolved Oxygen. — This is an important criterion of the purity of a sample of water. If a water contains readily decom- posable organic matter very little dissolved oxygen can be present as it will unite with the organic matter present. If, on the other hand, such organic matter is absent, or present only in small amount, the water will soon become saturated with oxygen if the air has free access to the water. The amount of oxygen which water can hold is a function of the temperature decreasing rapidly with increase of temperature. The amount of dissolved oxygen is reported as the percentage of the amount necessary 422 WATER ANALYSIS. for saturation at the temperature of the water when the sample was taken. Extreme care must be used when taking the sample to avoid the entrainment or absorption of any oxygen from the atmosphere. The sample is taken in a glass-stoppered bottle of about 250 c.c. capacity. The exact capacity is carefully determined. The sample should be taken so that the water enters the bottom of the bottle. This may readily be accomplished by allowing the water to enter the bottle through a funnel of suitable size. The bottle must be completely filled and tapped to remove any adhering bubbles of air. The stopper is inserted so as to leave no air bubbles underneath. The following reagents are used in the test: Manganous sulphate solution, made by dissolving 48 grams of manganous sulphate in 100 c.c. of distilled water. Sodium hydrate and potassium iodide solution, made by dis- solving 36 grams of sodium hydrate and 10 grams of potassium iodide in 100 c.c. of distilled water. Sulphuric acid, specific gravity 1.4, made by diluting 50 c.c. of concentrated acid with an equal volume of distilled water. Standard sodium thiosulphate solution, prepared by dissolving 6.2 grams of chemically pm^e sodium thiosulphate in one liter of distilled water. This gives an N/40 solution, each c.c. of which is equivalent to 0.2 mg. or 0.1395 c.c. of oxygen at 0° C. and 760 mm. pressure. The solution may be standardized by any of the methods given in chapter XXX. Starch solution, made as directed on p. 339. Immediately after being collected, proceed as follows with the sample. Remove the stopper and add 2 c.c. of the man- ganous sulphate solution and 2 c.c. of the sodium hydrate potas- sium iodide solution, delivering each of these solutions beneath the surface of the water by means of a pipette. Replace the stopper and mix the contents of the bottle by shaking. After the precipitate has been allowed to settle, remove the stopper and add 2 c.c. of the sulphuric acid by means of a pipette. Re- place the stopper and mix thoroughly. These reagents must be added immediately after taking the sample. The titration may be carried out at leisure as follows. NITRATES. 423 Rinse the contents of the bottle into an Erlenmeyer flask, and titrate with the n/40 sodium thiosulphate solution until the yellow color is nearly gone. Add a few c.c. of the starch solution and continue the titration until the blue color has dis- appeared. If nitrites are present, 2 c.c. of a strong solution of potassium acetate should be added before adding the 2 c.c. of sulphuric acid. The following reactions take place dm-ing this test: 4MnS04 + 8K0H + 02 = 2Mn203 + 4K2SO4 + 4H2O. The brown manganic oxide is dissolved by the sulphuric acid and potassimn iodide as follows: Mn203 + 2KI + 3H2S04 = MnS04 + K2S04+l2 + 3H20. The iodine is dissolved by the excess of potassium iodide, giving a reddish-yellow solution and is acted upon by the sodium thio- sulphate as follows: I2 +2Na,2S203 = 2NaI + Na2S406. The result is calculated to parts of oxygen per miUion, in c.c. per hter, or to per cent of saturation by the following formulas: 0.0002NX 1,000,000 _ 200N Oxygen in parts per million = y y- 0.1395NX1000 139.5N Oxygen in c.c. per hter = y = y 200N + 100 20,000N Oxygen in percent of saturation = — y^TQ — ^ — yQ~ ■ where N = number of c.c. of N/20 thiosulphate solution; 7= cap- acity of bottle in c.c, less the volume of the manganous sulphate solution and potassium iodide solution added (i.e., less 4 c.c). = the amount of oxygen in parts per miUion in water sat- urated at the same temperature and pressure. See table, p. 530. 421' WATER ANALYSIS. DETERMINATION OF CHLORIDES. The chlorine in natural waters exists for the most part in the form of sodium chloride, and as such is harmless unless present in very large amounts. As considerable amounts of salt are used very generally in articles of food, it is present in large amounts in urine, and therefore in sewage. For this reason the amount of chlorides present is used as an index of the amount of contamina- tion of the water. As almost all natural waters, including rain- water, contain chlorides in greater or less amount, only that amount of chlorides which is above the average of similar but pure waters can be used as an argument for condemning a water as contami- nated. 639. Silver Nitrate Solution. — The chlorine is determined volu- metrically, using a standard silver nitrate solution with potassium chromate as indicator^ as described in Chapter XXVI. The silver nitrate solution is made of such a strength that 1 c.c. is equal to 1 mg. of chlorine. For this purpose 4.7938 grams of pure silver nitrate are weighed out and dissolved in a liter of water. The potassium chromate indicator is made by dissolving 5 grams of the pure salt in 100 c.c. of distilled water. 640. The Determination is carried out by placing 100 c.c. of the water to be examined in a porcelain dish and adding 1 c.c. of the indicator. The silver nitrate solution is added until a faint per- manent red color is produced. This red color should be removed by adding a pinch of sodium chloride or a drop or two of a chloride solution and the titration repeated, using the first solution for com- parison. This solution will have a pure-yellow color with a white silver chloride precipitate equal to that obtained in the second titration. The end-point may be made still sharper by preparing a solution of distilled water containing 1 c.c. of the indicator, through which the solution being titrated may be observed. As an appreciable amount of the silver nitrate solution may be required to produce the color indicating the end-point, a blank determination should be made by adding 1 c.c. of the indicator to 100 c.c. of distilled water and then adding silver nitrate solution until the red color is observed. The volume of silver nitrate added ORGANIC MATTER. 425 should be subtracted from the amount added to the water being tested. When the amount of chlorides present is very small, or great accuracy is required, an amount of water greater than 100 c.c. should be taken and evaporated down to a bulk of about 100 c.c. after the addition of 0.1 c.c. of a saturated solution of sodium carbonate. This solution is then titrated as already given. If the water contains considerable coloring-matter, so that after concentration the color of the indicator cannot be observed, the coloring-matter may be removed before concentration by shaking the water with some freshly precipitated aluminum hydrate and filtering before measuring out the portion for evaporation. If the water is acid, it must be neutralized with sodium carbonate. 641. The Determination of Hardness in water, both temporary and permanent, has already been given in Chapter XXI, page 269. Alkalinity may also be determined by titration of the water with standard acid, using erythrosine as indicator.* This is especially advantageous when the water is turbid or highly colored. The erythrosine solution is prepared by dissolving 0.1 gram of the salt in one Uter of distilled water. One hundred c.c. of the water to be examined is measured into a 250-c.c. white glass-stoppered bottle. 2.5 c.c. of the ery- throsine solution and 5 c.c. of chloroform (neutral to erythrosine) are added. Sulphuric acid (N'/50) is added in small quantities, the bottle being shaken vigorously after each addition of the acid until the rose color of the chloroform is discharged. A white surface held back of the bottle facilitates the detection of any trace of color remaining as the end-point is reached. The number of c.c. of the N/50 sulphuric acid multiplied by ten gives the alkalinity in parts per million of calcium carbonate. Permanent hardness may also be determined by means of " soda reagent " which gives more accurate results, especially in the presence of much magnesium, f The soda reagent is prepared by combining equal volumes of N/10 caustic soda and sodium carbonate. The method is carried out as follows: * Standard Method of Water Analysis, p. 37. t Ibid., p. 41. 426 WATER ANALYSIS. Two hundred c.c. of the water to be exammed is measured into a Jena Erlenmeyer flask and boiled for ten minutes to expel the free carbonic acid. 25 c.c. of N/10 soda reagent is then added and the water boiled to a volume of 100 c.c. After cooling, it is rinsed into a 200-c.c. graduated flask and made up to 200 c.c. with boiled distilled water. After shaking, the water is filtered, the first 50 c.c. rejected, and 100 c.c. of the filtrate titrated for excess of soda reagent with N/20 sulphuric acid, using erythrosine as the indicator. If S= c.c. of N/20 sulphuric acid, equivalent to the soda reagent used and N= c.c. of N/20 sulphuric acid re- quired to titrate the excess, then the permanent hardness in parts per milhon of calcium carbonate is equal to 12.5 (S — 2N). 642. Determination of Total Solids. — ^This determination is conducted by evaporating a measured volume of the water, 100 to 500 c.c. being usually taken, to dryness in a weighed platinum dish. Various temperatures from 105° to 180° have been used for drying the residue before weighing, 103° having now been fixed as the standard temperature. After weighing, the organic matter is volatilized by gently igniting the residue. When the carbon has been burned off, the alkaline-earth metals are recon- verted to carbonates by treatment with carbonic-acid water, and after evaporating and gently igniting, the residue is again weighed. The loss in weight was foTmerly reported as organic matter. It is generally recognized that the results of this determination are at best but roughly approximate. During the evaporation considerable organic matter may be lost, still more being volatilized if the residue is dried at 180°; while if the residue is dried at 105°, much water will be retained if hygroscopic salts, such as cal- cium chloride, are present. During the ignition this water will be expelled together with considerable amounts of sodium chloride, if this commonly occurring salt is present. The loss on ignition is therefore considered of very little value, the chief reason for igniting the residue being to observe the amount of carbonization which takes place. The blackening of the residue by the forma- tion of carbon, which subsequently burns off, is taken as an indica- tion of the presence of organic matter, even if the loss in weight is inconsiderable; while the absence of a charred residue indicates the absence of organic matter, even if considerable loss in weight TOTAL SOLIDS. 427 occurs. The presence of nitrates is indicated by sparking of the carbon on ignition. On account of the difference in practice, the temjperature at which the total solids were dried should be stated in reporting the analysis. BACTERIOLOGICAL EXAMINATION OF WATER. 643. Bacteria Present in Water. — ^All natural waters contain varying numbers of bacteria which are microscopic single-celled organisms generally living on organic matter. The bacteria which feed on living organic matter, and are therefore parasitic, con- stitute the class of disease germs. This class is far less numer- ous than the bacteria which live on dead matter and frequently cannot exist for any length of time out of their living host. Only a few of this class can multiply or even live in water, the most common being B. typhi, which produces typhoid fever when it secures a foothold in the human system. Dysentery and similar intestinal diseases are produced by bacteria which are carried by water, but which have not been identified and studied. Another bacterium, B. coli, which is almost always present in the intestines of healthy human beings and other warm-blooded animals, can also live in water. The presence of this bacterium in water is therefore taken as evidence of sewage contamination. Such water is always regarded as unwholesome, since disease germs from the human system may at any time be present in the water. B. typhi will almost certainly be present at times in such water if sewage is the source of contamination, because cases of typhoid fever may always be found in any large community. 644. Test for B. Coli. — Fortunately a very simple method of testing for B. coli has been devised. During its life activity, this bacterium decomposes carbohydrates with the production of a gas which is mainly hydrogen, with 25 to 40 per cent of carbon dioxide. Bile has the property of inhibiting the growth of other gas-forming bacteria. The sample of water is therefore mixed with bile and lactose and kept at a temperature of 37J° C. for 48 hours. If gas is evolved during this incubation period, the water almost certainly contains B. coli. An idea of the number of these organisms present can be obtained by making three tests, 428 WATER ANALYSIS. using respectively .1 c.c, 1 c.c, and 10 c.c. of the water. The test is made in two test-tubes of such a size that the smaller one can be inverted within the larger one. 645. Bacterial Count. — In addition to the test for B. coli, the total number of bacteria in a water is determined. An excessive number of bacteria in a water indicates contamination with sew- age or other decomposing animal or plant tissue, and renders the water imfit for consumption. This test is made by adding a measured amount of the water to a plate containing a sterile nutrient medium which is composed of meat extract and pep- tone to nourish the growing bacteria, and gelatin or agar-agar to make a semi-solid media in which the bacteria will remain fixed. Each bacterium develops a colony around it, which ulti- mately becomes visible to the eye imder only very slight magni- fication. By counting these spots or colonies, the number of bacteria originally present in the water can be ascertained. This determination is only approximate, as all of the bacteria do not grow under the conditions of incubation sufficiently to be counted. All material used must be thoroughly cleaned and sterilized by heat, so as to kill all organisms present before the water to be tested is added. EXERCISE 72. Bacterial Count. 646. Sterilization of Apparatus. — Clean thoroughly 12 test-tubes and insert firmly a plug of cotton. Clean 6 Petri dishes and 6 one-c.c pipettes, also 2 small glass-stoppered bottles. Place the pipettes in a copper tube having a close-fitting cap, or in a large test-tube closed with a plug of cot- ton. Sterilize this apparatus by placing it in an air-oven heated to 150° for 1 hour. Prepare sterile water by nearly filling a 250-c.c. flask with distilled water and boiling vigorously for 15 to 20 minutes. Insert a plug of cotton and allow to cool. 647. Preparation of Nutrient Gelatin. — Place 500 grams of lean beef in a large beaker or flask, and add 1000 c.c. of distilled water. The meat must be as free as possible from fat and chopped fine or run through a sausage grinder. The beaker or flask is weighed and distilled water added to make up the loss by evaporation after standing 24 hours. Place a piece of new cotton flannel, with the wool side up, in a large funnel and cover with a layer of clean cotton wadding. Filter the meat infusion through the flannel, squeezing out the last portions. It is advisable to place a BACTERIOLOGICAL EXAMINATION OF WATER. 429 second smaller funnel containing the same filtering medium below the first funnel, so that the filtered solution from the first funnel passes through the second, thus giving a very clear solution. Allow the solution to flow into the inner vessel of au agate-ware double boiler of at least two quarts capacity. This vessel should be weighed empty as well as after receiving the meat infusion. Add an amount of Witte's peptone equal to 1 per cent of the weight of the infusion, and 10 per cent of the best quality of sheet gelatin (gold label). Dissolve the peptone and gelatin by stirring with a thermometer and heating the solution, not allowing the tempera- ture to rise above 60° C. For this purpose the outer vessel only of the double boiler is heated with the Bunsen burner, the inner vessel being surrounded by the hot water. After the gelatin is dissolved, the water in the outer vessel is brought to a boil and kept boiling for 30 minutes, the inner vessel being covered. 648. Adjusting the Acidity. — After the boiling has continued for 20 min- utes, withdraw 5 c.c. of the solution with a pipette and place in a porcelain dish or casserole. Add 45 c.c. of distilled water and boil for 1 minute over the Bunsen-burner flame. Add 1 c.c. of phenolphthalein solution and titrate while hot (preferably while boiling) with N/20 caustic soda. Add the soda solution until within a drop or two of the end-point. Cool by standing the dish in cold water, and if a distinct pink color does not develop, add the soda solution drop by drop until the ond-point is reached. So much of the acid in the solution must be now neutralized that the amount of acid remaining in 1000 grams will neutralize 10 c.c. of normal caustic-soda solution. The calculation of the amount of soda to be added is made as follows: If the 5-c.c. portion titrated required 4^ c.c. of the N/20 soda solution, and the weight of the solution is 950 grams, the entire solution will require 855 c.c. of the N/20 soda, or 42.7 c.c. of normal soda. As the acid equivalent to 10 c.c. of normal soda solution must be left free, 32.7 c.c. of normal soda solution is added. To coagulate finely suspended matter so that the solution may be filtered clear, the white of an egg is added at this point. The solution is cooled to 60°-70° by immersing the containing vessel in cold water, and the white of an egg added with stirring; it is then heated in the double boiler and finally boiled for 2 minutes over the free flame, with constant stir- ring. Weigh and add distilled water to make up for loss by evaporation. Take out 5 c.c. and titrate with N/20 sodium hydroxide and calculate the acidity as before. If the acidity per 1000 grams is less than S c.c. or more than 12 c.c. of normal acid, acid or alkali should be added to bring the acidity to the standard 1 per cent. Take out another 5-c.c. portion and titrate a third time to ascertain if the adjustment of the acidity has been correctly carried out. 649. Sterilization of the Culture Medium. — Filter the solution again through absorbent cotton and cotton flannel, passing the filtrate through the filter until clear. The nutrient gelatin must now be measured off in 10-c.c. 430 WATER ANALYSIS. portions into the sterilized test-tubes closed with cotton plugs. For this purpose a glass tube graduated every 10 c.c. is convenient. The cotton plug is removed from a sterilized test-tube and held between the first and second finger and again inserted into the test-tubes as soon as the 10-c.c. portion of gelatin has been added. The gelatin should not be allowed to touch the upper portion of the test-tube, which is set aside in an upright position to cool. When all of the gelatin has been measured out into test- tubes, it is sterilized by heating for 15 minutes in an autoclave at 120°, which is the temperature of steam at 1.5 pounds pressure. The tubes of gelatin must be kept in an ice-chest. If the gelatin has not been com- pletely sterilized, colonies will make their appearance in a few days. 650. Count of Bacteria. — To make the count of bacteria in water, the sample must be taken in a sterilized glass-stoppered bottle. It is advisable before sterilization to cover the stopper with tin-foil to prevent entrance of bacteria with the dust from the air; 1 c.c. is withdrawn with a sterilized pipette and transferred to a sterilized test-tube, and 9 c.c. of sterilized water added. This gives a dilution of 1 to 10. By diluting this solution in the same manner, a dilution of 1-100 is obtained. A clean pipette must be used for each dilution. To one of these dishes 1 c.c. of the sample of water is transferred; to the second dish 1 c.c. of the 1-10 dilu- tion, and to the third dish 1 c.c. of the 1-100 dilution. Tubes of the nutrient gelatin are melted by placing in warm water, the plug of cotton removed, the open end of the test-tube sterilized by passing it through the flame of a Bunsen burner, and the gelatin poured into three sterilized Petri dishes. The covers are raised only when pouring in water or gelatin, so as to prevent entrance of bacteria from atmospheric dust. The water and nutrient gelatin are mixed by slightly tilting the dish, so that the con- tents flow from one side to the other. The dish is then placed in a hori- zontal position in a thermostat or refrigerator kept at about 20° C. After 48 hours the number of colonies on the plate are counted. If the number is not over 200, the total number may be counted. If the number is large, some counting device must be employed. This generally consists of a plate marked off in sections or other divisions, so that the total number of divisions covered by the plate may be counted, and the average num- ber of colonies per division is obtanied by counting the number of colonies in several divisions, so as to obtain a fair average of the number per divi- sion. Small specks of dust must not be mistaken for colonies which are circular in shape. The count must generally be made with a small mag- nifying lens. When the colonies are small it is sometimes advisable to allow the sample to incubate for 72 hours. This fact should be stated, however, in reporting the analysis. The best results are generally obtained by counting the colonies on plates containing about 200. 651. Preparation of Nutrient Agar-agar. — When this medium is prepared the meat is soaked in one-half the quantity of water, that is, 500 c.c. Fifteen grams of thread agar are dissolved in 500 c.c. of water by boiling for one- INTERPRETATION OF RESULTS. 431 hau hour. The loss of water is then restored and the infusion allowed to cool to about G0° C. The remaining operations are identical with those used in preparing gelatin, except that 2 per cent of Witte's peptone is dissolved in the filtered meat infusion, after which the agar is added to the meat infusion, care being taken to keep the temperature below 60° C. The samples of water are plated as described for gelatin. It is neces- sary, however, to heat the agar to a higher temperature in order to melt it. It should be cooled to about 40° before pouring on the plate. The plates are incubated at body temperature (37J°-40° C.) for 48 hours. The number of colonies obtained is usually higher than with gelatin. It is advisable to use porous earthenware covers with agar to prevent the spreadmg of colonies by the condensation of water. EXERCISE 73. Test for B. Coli. Obtain a quantity of fresh undiluted ox-bile or a 10-per cent solution of dry fresh ox-gaU. Place in a flask and sterilize by heating in an auto- clave at 15 pounds pressure for 20 minutes, or heat in steam for one-half to three-quarters of an hour. Filter through cotton and add 1 per cent of lactose and 1 per of cent peptone. Add a sufiicient amount of the bile to a large test-tube containing an inverted smaller test-tube, so that the bile wUl fill the small tube and act as a seal at the bottom. Fill other test-tubes in the same maimer until all the bile has been utilized, and plug with cotton. Sterilize in an autoclave at 15 pounds pressure for 5 minutes. The air in the smaller tube will be expelled by the high tem- perature, and the steam produced, so that on cooling, the bile will entirely fill the smaller tube. Cool and store in a cool place. Take three of these tubes and add, with sterilized pipettes, 0.1 c.c, 1 c.c, and 10 c.c. of the water to be tested. Incubate at 37^° C. for 48 hours. The presence of gas in the small tube is e%'idence of the presence of B. coli in the water. INTERPRETATION OF THE RESULTS OF A SANITARY WATER ANALYSIS. No definite limits can be set for the amount of chlorides, nitro- gen, or other elements which may be present in water without rendering it unsafe for domestic use. None of these substances are themselves poisonous or injurious in the amounts found in water. The amount present can merely be used as evidence of more or less recent contamination of the water with sewage. A knowl- A32 WATER ANALYSIS. edge of the condition of the water before contamination is the best possible basis for a judgment. On comparing the analysis of a water which is suspected to have produced disease with the analysis of the same water before the disease appeared, if a marked increase in the amount of Chlorides or nitrogen is shown, contamination of the water may be suspected, and its source should be looked for by examining the watercourse from which the supply is drawn. In cities where the water is analyzed at regular intervals and where a careful record of the cases of typhoid and other diseases known to be due to impure water is kept the amount of nitrogen, chlorides, etc., in the water which indicates a contaminated water- supply is soon learned. 652. Normal Chlorine. — These difficulties in interpretation arise from the fact that the substances introduced into the water with sewage also find their way into the water from sources which cannot produce dangerous contamination. While, for instance, the chlorides present in water may be due to sewage contamina- tion, which always contains large amounts of chlorides, their pres- ence is by no means prima facie evidence of such contamination. As the earth contains many salt deposits with which the water may come in contact, and the salt from the ocean undoubtedly diffuses through the earth for considerable distances inland, natural waters may contain large amounts of chlorides derived from these sources. The amount of chlorides derived from these sources by the water of a given region is known as the normal chlorine for the region. Maps have been prepared for many states giving isochlors, or lines passing through regions of equal chlorine. When the amount of chlorine found is in excess of the normal chlorine, contamination of the water with sewage is indicated. 653. Nitrogen. — ^As has been already stated, the albuminoid AMMONIA may be derived from vegetable as well as from animal sources. A surface-water which has passed through a swampy region almost always contains a large amount of albuminoid ammonia. The presence of dissolved vegetable matter is almost always indicated by the high color of the water. On the other hand, the presence of much albuminoid ammonia in a light-colored water is almost always suspicious, ■ the absence of color making it probable that the organic matter is partly of animal origin. If INTERPRETATION OF RESULTS. 433 the free ammonia and nitrites are also high, the evidence of pol- lution is strong. 654, The Characteristics of Polluted Surface-waters are given by Geo. C. Whipple as follows: "They contain an excess of chlorine above the normal of the region. "They contain considerable organic matter, shown by the albuminoid ammonia and loss on ignition; but it should be remem- bered that the dissolved albuminoid ammonia may be explained by a high color. "They contain free ammonia and nitrites, if the pollution is recent and the organic matter is in a decomposing state. "They show coincident high nitrates and high hardness. " They are characterized usually by a mouldy or musty odor and, if much polluted, by turbidity and sediment." 655. Ground-waters are very different from surface-waters. As a rule they are clear, almost colorless, and without odor. Hav- ing been much in contact with the soil, they are rich in mineral matter. The fixed solids are higher, and usually the hasdness. The amount of chlorine and nitrogen may be high, indicating pollu- tion, but if the nitrogen is nearly all present as nitrate, the water has been purified by filtration through the soil and oxidation of the organic matter so that it may be safely used. The determination of the amount of total solids as well as HARDNESS is Carried out not so much to determine the sanitary character of the water as its suitability for laundry purposes, though the continued drinking of hard water is undoubtedly injurious to many constitutions. The presence of considerable amounts of iron renders water unfit for laimdry use, the presence of more than 0.5 part per million being objectionable. If the water contains much organic matter, more iron may be present without being objectionable. A hardness of 25 to 50 is most de- sirable for ordinary use. A water of hardness of 50 is noticeably hard, while if above 100 it is classed as very hard. The COLOR of water should not exceed 25. The amount of color which this figure implies may be judged from the fact that if the color of a water is 15 to 20 it will be noticed when the 434 WATER ANALYSIS. water is allowed to flow into a bath-tub, while if the color is 30 it may be noticed in a tumblerful. Waters having a turbidity of more than 3 or 4 are objection- able. ANALYSIS OF WATER FOR USE IN BOILERS. Aside from the sanitary analysis of water, the most commonly conducted analysis is that carried out for the purpose of deter- mining the adaptability of the water in question to use in boilers. The water may not be suitable for use in boilers for two reasons. It may, in the first place, contain substances which cause the CORROSION of the iron of which the boiler is made. Free acids, MAGNESIUM CHLORIDE, AMMONIUM SALTS, MUCH DISSOLVED OXY- GEN, HUMUS, AND FATTY SUBSTANCES are the commouly occurring substances which have been found to cause corrosion in boilers. A water may also be objectionable for boiler use from contain- ing SCALE-FORMING INGREDIENTS, among which the most impor- tant are calcium sulphate and carbonate and magnesium CARBONATE. Others occurring commonly, though in smaller amounts, are silica and the oxides of iron and aluminium. Although rarely forming parts of the boiler-scale, the alkalies should in a careful analysis be determined, as the distribution of the hydrochloric and sulphuric acids present may then be more correctly determined. 656. Hardness. — The fitness of a water for boiler use may be somewhat roughly determined by simply determining the tem- porary and permanent hardness, as given in Chapter XXI, page 269. All of the calcium and magnesium carbonates constituting temporar)'^ hardness are precipitated by boiling and become part of the boiler-scale. The calcium and magnesium existing as per- manent hardness may be present as chlorides or sulphates. The determination of permanent hardness does not give the amount of magnesium chloride present which causes corrosion, nor the amount of calcium sulphate present which is precipitated at the temperature existing in the boiler and constitutes part of the boiler-scale. 657. For a Complete Analysis an amount of water should be taken which will give from ^ to 1 gram of residue. It is evapo- BOILER-SCALE. 435 rated to dryness in a platinum dish and the total solids deter- mined as already directed. The ignition of the residue to volatilize organic matter and burn the carbon must be conducted at dull redness and the ignition continued only for a few minutes, or a loss of alkalies may occur. After weighing, the residue is digested with water and a few cubic centimeters of hydrochloric acid. The insoluble residue is filtered off, washed, ignited in the platinum dish and weighed. It is mainly silica, though a little iron, alxuninium, and calcium sulphate may be present. The silica may be volatilized by treatment with hydrofluoric acid and a drop or two of sulphuric acid, the residue weighed after ignition, and the weight deducted from that of the impure silica. The residue in the dish is dissolved by digestion with a little hydrochloric acid and water and is added to the main solu- tion. The IRON and alxjminium may be precipitated with ammonia and weighed as oxides in the usual manner. The CALCIUM is precipitated as oxalate from the filtrate and weighed as oxide. The MAGNESIUM and the alkalies in the filtrate may be weighed together as sulphates, the solution being evaporated to dryness in a platinum dish and the ammonium salts volatilized after the addition of a few drops of sulphuric acid. A small amount of silica and other insoluble matter generally remains with the sul- phates of magnesium and the alkalies. This is separated by treat- ing the residue in the dish with water and filtering. The filtrate is again evaporated to dryness in the dish, and after gentle igni- tion is weighed. The weighed sulphates are dissolved in water, the solution diluted to 100 c.c. and divided into two portions, in one of which the magnesium is determined in the usual manner as pyrophosphate, while in the other portion the potassium is precipitated as platinochloride, washed according to the Lindo- Gladding method and weighed. The amount of magnesium present as sulphate is calculated from the weight of the pyro- phosphate, and the weight of potassium as sulphate is calcu- lated from the weight of the potassium platinochloridfe. The difference between the sum of these two weights and the weight 435a WATER ANALYSIS. of the combined magnesium and alkali sulphates gives the weight of sodium sulphate present. The amount of chlorides present is determined volunletric- ally according to the method given under the sanitary analysis of water on p. 421. The amount of sulphates present is also determined in a separate portion. Unless a considerable amount is present, 500 c.c. are evaporated down to a bulk of about 100 c.c. and, after acidifying with a drop or two of hydrochloric acid, the sulphuric, acid is precipitated by the addition of barium chloride solution. After digestion on the water-bath the barium sul- phate is filtered off, washed with hot water, ignited, and weighed in the usual manner. If FREE MINERAL ACIDS are present, the amount is determined volumetrically by titration with a standard alkali. 658. Method of Combining Acids and Bases. — The elements present in the water having been determined, it remains to com- bine them in such a manner that the properties of the water when used in a boiler may be easily sho-nm, no attempt being made to combine them as they are believed to exist when in solution. The following method is generally followed : The chlorine is com- bined with the bases in the following order: sodium, potassium, magnesium, and calcium. The sulphuric acid is combined with the alkalies if the amount of chlorine was insufficient, then with the calcium and, if any remains, with the magnesium. The cal- cium and magnesium remaining are calculated as carbonates. That this method of combination does not exist in solution is evident from the consideration that the strong base calcium would undoubtedly be in combination with some of the hydro- chloric acid, while the alkalies would also be combined partly with the sulphuric and partly with the hydrochloric acid. When, however, the temperature was reached at which the calcium sulphate is almost absolutely insoluble in water this salt would be precipitated, even if the sulphuric acid were combined with the alkalies and the calcium with the hydrochloric acid. The ultimate result would therefore be that the alkalies would be found in combination with the chlorine, and the calcium with the sul- phuric acid. Some authorities recommend that as much magne- sium chloride as can be formed from the amounts of these elements 'CALCULATION OF RESULTS. 435& present be reported in order to bring out the fact that the presence of the two elements together leads to the formation of hydrochloric acid and magnesium oxide. This method of reporting the analy- sis probably overstates the danger, as some of the magnesium is undoubtedly precipitated as carbonate if that acid is present in the solution. 659. Calculation of Results. — The simplest way in which to state the results is in grams per liter or in milligrams per liter, which would be identical with parts per million. The results must frequently be given in grains per United States gallon (58,318 grains) or per imperial gallon (70,000 grains), to suit the conve- nience of engineers who still use the English instead of the metric system. The method of calculation may be more clearly understood from the following example. The analysis gave the following results: SiOa 0.0063 gram per liter SOs 0.0245 CI 0.0075 K2O 0.0021 Na^O 0.0065 Mgb 0.<)125 CaO 0.0223 Fe203+ AI2O3 0.0040 The sodium is first calculated to sodium chloride as follows: Na20:2NaCl:: 0.0065: a; 62. : 117.9 ::0.0065:a; a; = 0.0123 gram NaCl. The amount of chlorine in this weight of sodium chloride is then calculated: NaCl: CI :: 0.0123: a; 58.5:35.45:: 0.0123: a; x = 0.0075gramCl. As this is the amount of chlorine found by analysis, no other chlorides are present. 435c WATER ANALYSIS. The potassium is next calculated to sulphate : K^OiK^SO^:: 0.0021: a; 94.3:174.36: :0.0021:a; a; =0.0039 gram K2SO4 K20 = 0.0021 " 803 = 00018 " The amount of potassium oxide being subtracted, it is found that 0.0018 gram of sulphur trioxide has been taken to combine with the potassium, leaving 0.0227 gram (0.0245-0.0018) which is combined with calcium oxide to form sulphate. S03:CaS04:: 0.0227: a; 80:136.2 ::0. 0227:3; a; =0.0386 gram CaSO, SO3 = 0.0227 " CaO = 0.0159 " 0.0064 gram (0.0223-0.0159) of calcium oxide remains after all of the sulphur trioxide has been combined to form sulphate. This calcium oxide will be present as carbonate. CaO:CaCO3::0.0064:x 56 :100 ::0.0064:.x a; = 0.0114gramof CaCOa. The chlorine and sulphur trioxide being all combined, all of the magnesium will be present as carbonate. MgO:MgC03:: 0.0125: a; 40.3: 84.3: :0. 0125 :x X = 0.0261 gram MgCOj. The result of the analysis is therefore reported as follows: NaCl = 0.0123 gram per liter K2SO, = 0.0039 " CaSO, = 0.0386 " CaCO3 = 0.0114 " MgC03 = 0.0261 " FeA+AlA = 0.0040 " SiO,= 0.0063 " LITERATURE. 4S5d LITERATURE. Water Supply. Mas n (John Wiley & Sons) Am, Water, and Food. Richards & Woodman (John Wiley & Sons). Water Supply. Nichols (John Wiley & Sons). Sewage. Rideal (John Wiley & Sons) . Examination OF Water. Leffman (Blakiston). Water Supplies. T. C. Thresh (Rebman Pub. Co., London). Special Report on Examination of Water Si:pply, 1890, by the Mass. State Board of Health. Special Report on Purification of Sewage and Water, 1890, by the Mass. State Board of Health. Typhoid Fever, its Causation, Transmission, and Prevention. Geo. C. Whipple (John Wiley & Sons). Elements of Water Bacteriology, with Special Reference to Sani- tary Water Analysis. Prescott and Winslow (John Wiley Sr Sons). The Bacteriological Examination of Water-supplies. W. G. Savage (Blakiston & Sons, Philadelphia). CHAPTER XXX. ANALYSIS OF FATS AND OILS. 660. The Quantitative Analysis of a fat or oil is conducted for the purpose of identifying the constituent fats or oils as well as ascertaining the proportion in which they are present. The difficulties of the subject arise from the fact that not only are most pure fats and oils mixtures of a considerable number of chemi- cal substances, but a given fat or oil, being produced as a part of an animal or vegetable structure, is by no means uniform in com- position, but varies with the conditions under which it was pro- duced. Changes in climate, season, food, etc., produce consid- erable variations in the fat or oil produced. The number of oils and fats which are produced on a commercial scale is very large, thus adding very much to the complexity of the subject. The mineral oils, being obtained by distillation of crude products from various parts of the earth, differ still more widely in composition. The strictly chemical examination is confined almost exclusively to vegetable fats and oils. The methods employed are essen- tially inorganic, and therefore come within the scope of this work. The interpretation of the results should be undertaken only after the consultation of a more extensive work on the subject,* except in the case of pure oils and comparatively simple mixtures. 661. Chemical Composition. — ^All animal and vegetable oils and fats are organic salts or esters in which a common base, GLYCERINE, is always present. Besides glycerine, most fats and oils and especially the waxes contain a larger or smaller quantity of a base peculiar to themselves. Of these, Benedikt and Ulzer =^ A very excellent work for this purpose is that by Lewkowitsch on "Chemical Analysis of Oils, Fats, and Waxes." (Macmillan & Co.) A recent German publication is also excellent, "Analyse der Fette und Wachsarten." Benedikt und Ulzer. (Julius Springer, Berlin.) 436 FREE ACID. 437 mention twelve. Three of these belonging to the aromatic series furnish a means of distinguishing between fats and oils of vege- table or animal origin. Cholesterin and isocHOLESTBRi]>f occur exclusively in substances of animal origin, while vegetable products contain another base called phytosterin. These may be liber- ated from the fatty acids by saponification, and after extraction with ether, may be crystallized from alcohol, cholesterin and isocholesterin forming irregular plates or tables, phytosterin, groups of needles. The first melts at 147°, the second at 138°, and the third at 132°-134°. Failure to obtain these substances, however, is no indication that the oil or fat was not derived from its proper source, since they often exist in such small quantity as to be isolated with difficulty. The organic acids present in com- bination with the glycerine are very numerous and differ greatly in their properties. On these differences are founded most of the methods of identifying the oils and fats. 662. Acidity. — In many oils and fats, especially those of vege- table origin, part of the acid exists uncombined with the base. The amount of this free acid increases with the age of the oil, this being especially true of palm-oil. The amoimt of the free acid in most fresh animal oils is generally so small as to be neg- ligible. Old and rancid fats become strongly acid. The method of determining this value, as well as preparing the solutions required, is given in the following paragraphs. 663. Standard N/2 Hydrochloric Acid Solution. — Prepare 2 liters of a half-normal solution of hydrochloric acid by measuring out about 40 c.c. concentrated acid and diluting to a liter. Repeat the operation and pour the two solutions together and mix thor- oughly. Standardize the solution by one of the methods given in Chapter XXI, page 258, and dilute to exact strength. 664. Standard N/2 Caustic Potash Solution. — Prepare some pure alcohol by adding to about 2 liters of the ordinary alcohol a few pieces of caustic potash, shaking thoroughly, and after allowing the solution to stand for a few hours, distilling off the alcohol. To 1 liter of the alcohol add 30 grams of caustic potash. The alcohol should be poured into a liter bottle, the potash added, and the bottle closed with a rubber stopper. Shake until the caustic pot- ash is dissolved. The potassium carbonate will not dissolve, but 438 ANALYSIS OF FATS AND OILS. will adhere quite firmly to tlie sides and bottom of the bottle if the solution is allowed to stand undisturbed for a few hours, Decant or siphon off the clear liquid into another liter bottle, exposing the solution to the air as little as possible. The solution is standardized by titrating with the standard hydrochloric acid, using phenolphthalein as the indicator. It should be protected from the carbon dioxide of the air both when being withdrawn from the bottle and during a titration. A simple method of accomplishing this object is to arrange a siphon for withdrawing the solution from the bottle, as shown in Fig. 50, p. 290. The siphon should be passed through a rubber stopper with two holes, a straight tube containing soda-lime being inserted in the second hole of the stopper The end of the siphon exposed to the air should be kept closed with a piece of rubber tubing having one end closed with a glass plug. After being filled, the burette should' be immediately closed with a small rubber stopper through which passes a soda-lime tube. 665. Determination of Free Acid. — For the determination of the acid value of oils, calculate the volume of the N/l2 caustic potash solution to measure out and dilute with alcohol to 250 c.c. to make a fifth-normal solution. As the solution is not stable it must be used very soon after being made. All solid fats must be melted at as low a temperature as possible, and after standing until the impurities have settled, the clear fat is carefully decanted through a dry filter-paper. The vessel containing the oil or fat is weighed together with a small glass rod or spoon. From 1 to 2 grams are transferred to an Erlenmeyer flask, the exact amount being obtained by difference. 50 c.c. of the redistilled alcohol are now added and the flask shaken until the oil is dissolved. If the fat does not readily dissolve in alcohol, 50 c.c. of a mixture of equal parts of alcohol and ether may be used. One drop of phenolphthalein is now added. The solution is titrated with the fifth-normal caustic potash solution, which should be added drop by drop with constant and vigorous shaking. The first appearance of pink color is taken as the end-point. It may fade on standing a few minutes, but this is due to saponification of the neutral glycerides by the excess of alkali. When oils of high free acid value, such as palm-oil or old and rancid ^^ts are being tested, less than 1 gram may be taken for a KOTTSTORFER VALUE. 439 titration. When the acid value is small, as in most solid fats, as much as 5 or 10 grams may be taken to obtaia a reliable determina- tion. The number of milligrams of KOH per gram of fat or oil is calculated, which is the acid value. The acid value is also fre- quently expressed in terms of the amount of oleic acid corre- sponding to the amount of alkali used. The molecular weight of oleic acid being 282, 1 c.c. of N/5 alkaU is equal to 0.0564 gram of oleic acid. 666. Kottstorfer Value. — As the acids present in fats and oils differ in molecular weight and basicity, varying amounts of alkali will be required to neutralize the acid in a definite amount of the oU or fat. Caustic potash is the alkali invariably used, 1 gram of the fat or oil being taken for the determination. The so-called Kottstorfer value is the number of milligrams of KOH required to neutralize the acids present in 1 gram of the fat or oil. When an oil or fat is treated with an alcoholic solution of caustic potash the union between the glycerine and the fatty acids is broken, so that the acid is left in combination with the potassium. This process is called saponification. An excess of standard caustic potash dissolved in alcohol is added to a weighed amount of the fat or oil. When the saponification is complete the excess of alkali is titrated with standard acid. The number of milligrams of KOH neutralized by one gram of the fat or oil is then calcu- lated to find the Kottstorfer value. , 667. Determination of Kottstorfer Value. — ^To determine the Kottstorfer or saponification value, weigh out as directed for the determination of free acid from 1 to 2 grams of the fat or oil into a small Erlenmeyer flask fitted with a cork to a return condenser. Add 25 c.c. of the alcoholic caustic potash solution with a pipette or a burette, and heat on the water-bath until the oil or fat is dissolved, which will require from ten minutes to half an hour, depending on the fat or oil being investigated. Measure out 25 c.c. of the caustic potash solution into a similar flask connected with a return condenser, and digest on the water-bath for the same time. Cool for a few minutes and titrate the contents of each flask with the N/5 hydrochloric acid, using rather more phenolphthalein than usual as the indicator. The difference between the acid used in the two titrations gives the acid necessary to neutralize the alkali 440 ANALYSIS OF FATS AND OILS. used in the saponification. Calculate tlie number of milligrams of KOH necessary to saponify 1 gram of the fat or oil. 668. Ether Value. — ^The ether value is the difference between the saponification and the acid value. 669. "Reichert and Reichert-Meissl Value. — Some of the fatty acids of small molecular weight are quite volatile, and also soluble in water. It has been found that by liberating the acids con- tained in a given fat or oil, adding water, and distilling off this water again, a distillate is obtained which contains a considerahl<» percentage of the volatile acids. It is found in practice that the total amount of volatile acid present is not separated in this manner by one distillation. The amount which comes over depends upon the dilution of the solution, the amount of the distillate, and the rapidity of the distillation. The amount of acid in the distillate is determined by titration with decinormal caustic potash. The Reichert value indicates the number of cubic centimeters of N/10 potash required to neutralize the acid from 2..5 grams of fat contained in 100 c.c. of distillate, 110 c.c. of which has passed over in one-half hour. The Reichert-Meissl value is obtained in the same manner, 5 grams of fat having been used instead of 2.5 grams. This number is not necessarily twice the Reichert number. This determination has been used more especially to distinguish butter from the artificial substitutes which have been put on the market. 670. Determination of Reichiert-Meissl Value. — ^For the deter- mination of the Reichert-Meissl number a sample of butter-fat should be used which has been freed from casein, water, salt, etc., by melting and filtering as already directed. A quantity of this fat as nearly as possible 5 grams is weighed out into a 250-c.c. Erlenmeyer flask, 2 c.c. water, 8 c.c. alcohol, and a stick of caustic soda weighing about 2 grams are added. Attach the flask to an inverted condenser by means of a cork stopper and heat on the water-bath for one hour, shaking the contents of the fiask occa- sionally. Remove or incline the condenser and distil off the alco- hol by immersing the flask in the hot water and heating for about three-quarters of an hour. After the alcohol is removed add 100 c.c. recently boiled distilled water and dissolve the soap by warming on the water-bath and shaking. Cool the solution to POLENSKE VALUE. 44) 60° or 70° and add 50 c.c. of a solution of sulphuric acid containing 25 c.c. concentrated acid per liter. Again attach the flask to the inverted condenser and heat on the water-bath until the insoluble acids collect in a clear oily layer on top of the liquid. Allow the contents of the flask to cool to the room temperature. Introduce a few pieces of pumice-stone which have been heated to redness and dropped into distilled water. Attach the flask to an inclined condenser and distil off 110 c.c. into a graduated cylinder, regulating the heat so that thirty minutes as nearly as possible shall be required for the distillation. Mix the distillate thoroughly by pouring back and forth into a dry beaker. Filter through a dry paper into a 100-c.c. flask. Pour into a beaker, add a few drops of phenolphthalein, and titrate with N/10 alcoholic KOH made by diluting the half-normal solution. When the end-point is reached the 100-c.c. flask is rinsed and the titration completed. The num- ber of cubic centimeters of N/10 alkali used multiplied by 1.1 gives the Reichert-Meissl number. 671. The Polenske Value. — This method of examining butter- fat has come into use for the detection of cocoanut-oil, which has been largely used for the adulteration of butter. The Reichert- Meissl value has been generally relied on to detect the admixture of foreign fats in butter, on account of its high value in the case of butter-fat and the low value given by other fats, except cocoa- nut-oil, which gives a Reichert-Meissl value of 7 to 7.8, so that a considerable amount of this oil can be mixed with a butter having a high value, 30 for instance, producing a mixture which would have a Reichert-Meissl value above the legal limit of 20 to 24. Polenske found that the greater part of the volatile fatty acids of butter obtained in the Reichert-Meissl determination are solu- ble in water, while the reverse is true of cocoanut-oil. The amount of the soluble and insoluble acids is obtained, as in the Reichert-Meissl determination, by titration with decinormal caus- tic potash, and the values are given in cubic centimeters of the alkali. For butter-fat the insoluble volatile fatty acids required 1.5 to 3.0 c.c. of the decinormal alkalies, while 16.8 to 17.8 c.c. was required for the insoluble volatile fatty acids from cocoanut- oil. The Polenske value varies with the Reichert-Meissl value, as is shown in the table on page 459. 442 ANALYSIS OF FATS AND OILS. 672. The Determination of the Polenske Value is made in a manner very similar to the Reich ert-Meissl determination. It has been found that the details of the method must be adhered to very strictly in order to obtain concordant results. Five grams of the butter-fat are placed in a 300-c.c. flask and saponified with caustic soda and glycerine, for this purpose 2 c.c. of a caustic- soda solution, prepared from equal parts of caustic soda and water, and 20 grams of glycerine, are added, and the flask heated over the free flame until saponification is complete. The solution is allowed to cool below 100°, and 90 c.c. of water added. The mass is dissolved by warming on the water-bath to 50° C. The solu- tion must be clear and colorless. If it is brown it has been over- heated during the saponification and must be rejected and the saponification repeated. 50 c.c. of dilute sulphuric-acid solution containing 25 c.c. concentrated sulphuric acid per liter are now added, and the volatile acids distilled off. For this purpose i gram of pumice is added. This pumice must be broken and sifted so as to include only pieces of 0.5 to 1 mm. diameter. The flask is placed on copper or brass gauze of 0.5 mm. mesh. The flask is connected to an upright condenser by means of a bulb tube. The bulb is immediately above the stopper of the flask, and the tube passing to the condenser is inclined toward the bulb, which must be 70 mm. from the condenser. The heat is regulated so that 110 c.c. are distilled off in 19 to 20 minutes. The cooling water must flow rapidly enough through the condenser to cool the distillate to 20°-23° C. as it drops into a 110-c.c. flask, which serves as a receiver. When 110 c.c. have distilled over, the flask is replaced with a 20'C.c. measuring cylinder. The 110-c.c. flask containing the distillate is immersed, with- out shaking, in water at 15° C. After about 5 minutes the neck of the flask is tapped lightly, so as to cause the drops of oil floating on the surface of the liquid to adhere to the walls of the flask. After another 10 minutes the consistency of the oily drops is noted. If the butter is pure and has a high Reichert-Meissl value, the drops will be semi-sohd and milky. If the Reichert- Meissl value is low, or 10 per cent or more of cocoanut-oil is pres- ent, the insoluble fatty acids do not sohdify, but remain clear and oily. The contents of the flask are mixed by inverting it DETERMINATION OF THE POLENSKE VALUE. 443 several times without violent shaking, and 100 c.c. filtered through an 8-cni. filter-paper and titrated with tenth-normal caustic potash. The insoluble acids on the paper are washed free from soluble acids by washing with three 15-c.c. portions of water which have been passed through the condenser tube and the 20-c.c. cylinder and the 110-c.c. flask. The insoluble acids are then dissolved in neu- tral 90 per cent alcohol by passing three 15-c.c. portions through the condenser tube, the 20-c.c. measuring cylinder, the 110-c.c. flask, and the filter-paper. Each portion is allowed to drain completely before adding a fresh portion. The combined alco- holic solution is then titrated with tenth-normal alkali to obtain the Polenske value. 673. The Hehner Value is the converse of the Reichert value, the percentage of insoluble acids being determined by saponifying the fat, adding an excess of mineral acid so as to liberate the fatty acids, then filtering off, washing, and weighing the insoluble portion. 674. The Iodine Value of oils and fats is obtained by allowing a measured amount of an iodine solution to act on the oil or fat. It has been found that a considerable amount of the halogen is absorbed by the fatty acids present. The percentage 'of the halo- gen absorbed is taken as the iodine value. The chemical reactions which take place are not fully understood. In general it is thought that in the unsaturated acids the double bond is broken and the iodine enters at this point into the molecule. There seems to be no doubt, however, that substitution takes place to a limited extent, an atom of iodine taking the place of an atom of hydrogen, which unites with another atom of iodine, forming hydroiodic acid. Solutions of bromine have been used in the same manner, while recently solutions of iodine monochloride and also solutions of iodine monobromide have come into use. The results are almost universally computed as percentage of iodine absorbed. Slightly different values are obtained by the use of these different solu- tions. Strictly comparable results can be obtained only by working with the same solutions and under the same conditions, especially as to the time during which the halogen is allowed to act on the oil or fat. 675. Wijs' Iodine Monochloride Solirtion. — For the deter- mination of the iodine, the Wijs solution of iodine is the most 444 ANALYSIS OF FATS AND OILS. advantageous for use. This is a solution of iodine monochloride in glacial acetic acid. The acid must be the strongest and purest obtainable, giving no reduction with sulphuric acid and potas- sium dichromate. The percentage of acid must not be less than 99 and can be most readily determined by taking the melt- ing-point which varies with the amount of water present, as is evident from the following table: Per cent of H.C2H3O2. Melting-point. 100 16.75 99.5 15.65 99 14.80 97.09 11.95 95.24 9.40 The test is made by half filling an 8-inch test-tube with the dcetic acid. This test-tube is suspended within a larger test-tube by means of a cork, through which the smaller tube is passed. A delicate and accurate thermometer is placed in the acid, which is cooled by immersing the whole in ice-water As soon as the acetic acid begins to crystallize, the temperature rapidly rises to the true melting-point of the acid. The advantage of using this solution is that its titre remains unchanged for months, and the time necessary for a determination is reduced from four or six hours to a few minutes. Thirteen grams of pure iodine is dissolved in a liter of the acid. The exact strength is then determined by titration against a standard thiosulphate solution. Chlorine free from hydro- chloric acid is passed into the solution until the dark-brown color changes to a light-brown or yellow color. The strength should then be determined again by titration against the thiosulphate solution, and should be double the original strength. It is advis- able, however, to have a slight excess of iodine present rather than of chlorine. 676. Hanus' Monobromide Solution. — The Hanus solution serves the same purpose as the Wijs solution, and is more easily made. Instead of passing chlorine into the acetic-acid solution of iodine, liquid bromine is added until the titre is doubled. About 3.0 c.c. will be found sufficient. IODINE VALUE. 445 677. Sodium-thiosulphate Solution. — An N/10 sodium-thio- sulphate solution is made up and standardized by means of re- sublimed iodine. This solution is then titrated against the iodine solution. For this purpose 20 c.c. of the iodine solution are meas- ured out into a beaker, 100 c.c. distilled water added, and 1 gram potassium iodide dissolved in 10 c.c. water. The thiosulphate solution is then added until a faint yellow color is obtained, the starch solution is added, and the titration completed. 678. Determination of Iodine Value. — To determine the iodine value of an oil, from 0.15 to 0.2 gram of a drying-oil, 0.3 to 0.5 gram of a non-drying oil, or 0.7 to 1 gram of a solid fat is weighed out and transferred to a glass-stoppered bottle. The most con- venient method of weighing out oils is to pour a suitable amount into a small beaker, place in the beaker a short glass tube, and weigh carefully. By means of the glass tube, drop some of the oil into the bottle, replace the tube in the beaker, and weigh again. The oil may also be weighed in a tared homeopathic shell vial. The vial containing the oil is introduced into the glass-stoppered bottle. The oil or fat is dissolved in 10 c.c. of chloroform and 25 to 50 c.c. of the iodine solution are added. The iodine must be pres- ent in excess. If the iodine solution is very much decolorized by the oil, more must be added. The iodine should be allowed to act for fifteen minutes for non-drying oils, thirty minutes for semi- drying oils, and one hour for drying oils. A blank determination is carried on at the same time, 10 c.c. of the chloroform being treated with iodine solution for the same length of time. T^ter the iodine has acted on the oil, 10 c.c. of a 10% solution of potassium iodide free from iodate are added to the bottle, and this is followed, after thorough agitation, by ab9ut 100 c.c. of distilled water, care being taken that all of the iodine solution is washed down from the stopper, neck, and walls of the bottle. The excess of iodine is then titrated with thiosulphate. The num- ber of cubic centimeters of thiosulphate used, subtracted from the amount required for the blank, represents the amount of iodine absorbed by the oil. The equivalent weight of iodine thus found is divided by the weight of oil taken, and the result expressed in per cent as the iodine number. 446 ANALYSIS OF FATS AND OILS. 679. betermination of Rosin in Shellac by A. C. Langmuir's Method.* — The iodine value of unbleached shellac varies from 15 to 18 per cent, while that of rosin varies from 175 to about 260. It is possible, therefore, to ascertain approximately the amount of rosin present in a given sample of shellac by a determination of the iodine number. For calculating the result, 18 is used as the iodine number of the shellac, and 228 as that of the rosin. By using these figures the percentage of the adulterant which is found may be slightly lower than -that actually present. This is in accord with the accepted commercial principle that the inaccuracies in a method of analysis shall favor the seller rather than the buyer, by assigning the lowest possible value to the amount of impurities present. The calculation is made accord- ing to the formula , A-M where y = per cent of rosin ; M = iodine number of shellac; N= " " " rosin; A= " " " mixture. Substituting the fixed values of M and N, the formula becomes A-18 2/ =100 210 The iodine number is determined by means of the Wijs solu- tion. Exactly 0.2 gram of the ground sample is introduced into a dry 250-c.c. glass-stoppered bottle. 20 c.c. of glacial acetic acid is introduced and warmed until the shellac is dissolved, with the possible exception of a little wax. 10 c.c. of chloroform is then added and the solution cooled to 21°-24° C. This tem- perature must be maintained during the remainder of the test. 20 c.c. of the Wijs solution is introduced by means of a pipette. The bottle is allowed to stand in water having a temperature of 22° to 23° for exactly 1 hour. *Ber., 31, 750. J. Soc. Chem. Ind., 17, 609. Jour. Am. Chem. Soc, 29, 1221. ACETYL VALUE. 447 Pure shellac will scarcely alter the color of the Wijs solution. If rosin is present, a recklish-brown color will be produced, the depth of which depends on the amount of rosin present. After standing 1 hour, 10 c.c. of a 10 per cent potassium-iodide solution is added. The excess of iodine is inmiediately determined by means of a tenth-normal solution of sodium thiosulphate. When the iodine color is nearly gone, a little starch is added and the titration completed. Any color returning after a haK-minute is disregarded. On account of the large coefficient of expansion of the Wijs solution, a blank determination should be made with that of the shellac. The determhiation is carried out in exactly the same mamier, with the exception of the addition of the shellac. 68o. Acetyl Value. — ^This permits the identification of castor- oil, grapeseed-oil, and of oxidized fish oils in mixtures. Besides this, many fish oils show a comparatively high acetyl value. The test is founded on the action of acetic anhydride upon alco- holic hydroxyls as they exist in the oxyacids, as is shown in the following reaction: C„H3,(OH)COOH+(C2H30)20 = C,7H32(OCjH30)COOH+HC2H302. Upon saponification of the acetylated fatty acid thus produced, the hydroxyl is again replaced and the acetyl group liberated as an acetate. The amount of acetic acid liberated may be deter- mined, giving the acetyl value. Benedikt first proposed this method, operating on the fatty acids, but the process was modified by Lewkowitsch, who works on the oils or fats directly, giving more exactly the true content of hydroxy acids. 68 1. Determination of Acetyl Value. — ^The method adopted by the Association of Official Agricultural Chemists is carried out as follows: Boil a portion of the oil or fat with an equal volume of acetic anhydride for two hours and pour the mixture into a large beaker containing 500 c.c. of water, and boil for half an liour. To prevent bumping, a slow current of carbon dioxide is passed into the liquid through a finely drawn-out tube, reaching nearly to the bottom. Allow the mixture to separate into two layers, siphon off the water, and boil the oily layer with fresh water until it is no longer acid to litmus-paper. The acetylated fat is then sepa- rated from the water, filtered, and dried in a drying-oven. 448 ANALYSIS OF FATS AND OILS. Weigh from 2 to 4 grams of the acetylated fats into a flask ana saponify with alcoholic potash, as in the determination of the saponification number. If the distillation process is to be adopted it is not necessary to work with a standardized alcoholic-potash solution. In case the filtration method is used, which will be found much shorter, it is necessary that the alcoholic potash be measured exactly. In either case evaporate the alcohol after saponification and dissolve the soap in water. Now two pro- cedures are possible, either distillation or filtration. (a) Distillation Process. — Acidify with dilute sulphuric acid (1-10) and distil the liquid, as in the Reichert test. As several hundred cubic centimeters must be distilled, either a current of steam is run through or portions of water are added from time to time. From 500 to 700 c.c. of distillate wiU be found suffi- cient. Filter the distillates to remove any insoluble acids carried over by the steam, and titrate the filtrate with decinormal potas- sium hydroxide, using phenolphthalein as the indicator. Multiply the number of cubic centimeters of alkali employed by 5.61, and divide by the weight of substance taken. This gives the acetyl value, which is the number of milligrams of KOH required to neutralize the acetic acid liberated from one gram of the acetyl- ated fatty acid. (&) Filtration Process.— Add to the soap solution a quantity of the standard sulphuric acid exactly corresponding to the amount of alcoholic potash added, warm gently, and the free fatty acids will collect on top. Filter off the liberated fatty acids, wash with boiling water until the washings are no longer acid, and titrate the filtrate with decinormal potassium hydroxide, using phenolphthalein as the indicator. Calculate the acetyl value as before. PHYSICAL TESTS. Besides the constants obtained by strictly chemical quantita- tive means, there are a few physical tests that are extremely im- portant and widely used, which it seems advisable should be mentioned under the head of oil-analysis. 682. Specific Gravity. — This may be very accurately obtained by means of a Westphal balance as described on p. 275. PHYSICAL TESTS. 449 If the oil is viscous, however, a pycnometer must be used as described on p. 274. 683. Index of Refraction. — This is found by means of an instrument called a refractometer, of which there are several on the market. The Zeiss butyro-refractometer, one of the simplest of these, is also one of the most efficient. The oil or fat is intro- duced as a thin film between two halves of a Nicol's prism, which are enclosed in a hollow metal jacket through which water is allowed to flow, maintaining a constant temperature. The degree of diffraction is indicated on an arbitrary scale by a shadow which is observed through a telescope attached to the prism, and this reading, by comparison with a table, made especially for the instrument, is converted into the refractive index. The Abbe refractometer is better suited for general use, being constructed with prisms and water-jackets like the Zeiss butyro- refractometer, but instead of an arbitrary scale fixed in the tele- scope, the border of a shadow, visible through the latter, is brought to the intersection of cross-hairs by moving an arm at the side of the instrument, and the refractive index may then be read directly on a scale along which the movable arm slides. 684. Maumene Number and Specific Temperature Reaction. — It is found that by mixing various oils with concentrated sul- phuric acid a reaction ensues, causing a rise of temperature in the mixture, which for the same oil under the same conditions and with the same concentration of acid is a constant. In order to bring more uniformity into the results obtained by different operators working under different conditions and using different strengths of acid, the old Maumene test has been modified with considerable success. Maumene mixed 50 grams of oil with 10 c.c. of concentrated acid in a small beaker sunk in a larger vessel filled with cotton or some other insulating material. Before mix- ing, both oil and acid were at the same temperature, as close to 20° as possible. This temperature was noted, and, after com- bining, the highest point registered by the thermometer used for stirring the mass was observed, which, minus the initial tempera- ture, gave the Maumene number. The errors introduced by different degrees of insulation and strength of acid have been elimuiated by conducting another test exactly as the one just 450 ANALYSIS OF FATS AXD OILS. described, using the same apparatus and the same acid, but sub- stituting 50 c.c. of pure water for the oil. Then, by the following formula : .^ lOOA in which S = specific temperature number, A=Maumene number, 5 = rise of temperature obtained with pure water under the same conditions as those used in the Maumene test, results are obtained which compare admirably, provided the acid used is fairly concentrated — between 95 and 99 per cent.* 685. The determination is carried out as follows: j A beaker, 5 inches by 1-^ inches, is placed inside another, 6 inches by 3 inches, and a wet mixture of asbestos and plaster of Paris tightly packed around the inner one. This, when dried, makes a hard, solid packing which radiates heat very slowly. Remove the inner beaker, weigh into it 50 grams of fat or oil, and note the temperature carefully. Then from a pipette which will deliver it in approximately one minute add 10 c.c. of 96-98% sulphuric acid, which is at the same temperature as the oil. While the acid is being introduced, stir the oil and acid with an accurate thermometer. Then hold the thermometer-bulb carefully in the centre of the mixture, and when the mercury reaches the highest point note the reading. It is easy to determine this point, as the column of mercury remains stationary for some time. It is neces- sary to take care not to read the temperature too soon, as some oils take considerable time to reach their maximum point. The difference between the initial reading and the final reading, ex- pressed in degrees centigrade, gives the Maumf^nc- number. In order to get the specific temperature number the operation is repeated with the same apparatus and the same acid, using 50 c.c. of pure water in place of the oil. It is advisable in this case to cover the beaker with a piece of cardboard or heavy paper ♦References: Journal Soc. Chem. Ind., i;,91, 10, 233; Jour. Amer. Chem, Boc, 1902, 24, 266. ■f Allen, Com. Org. Anal., 3d Ed., Vol. II, Pt. 1, p. 76. PHYSICAL TESTS. 451 twice the diameter of the latter, the thermometer being thrust through a small hole in the centre. This interferes with the evaporation of the water and the consequent fall of the tempera- ture of the mixture, but if employed should be used with the oils also. 686. Melting-point of Fatty Acids. — -This test is much used in corroborating results obtained by other methods when the ques- tion of identity or purity of an oil or fat arises. The sample is saponified by boiling with an excess of alcoholic potash, and the fatty acids liberated, as in the Hehner method, with dilute sul- phuric acid and washed with warm water until the washings show a neutral reaction. The melted fatty acids are then drawn into a very thin-walled capillary tube 1 or 2 inches long according to the length of the bulb of the thermometer used. The capillary is sealed about one-fourth inch above the filled portion by holding Ln the outer mantle of a Bunsen flame and drawing out when the glass softens. The tube thus prepared, after cooling in a refrig- erator, is fastened to the bulb of the thermometer by an elastic band and immersed in a glycerine bath, which, to secure a slower and more even rise, may in turn be surrounded by a second bath of the same material which is directly heated by a Bunsen flame. By regulating the flame the rise' of temperature may be readily controlled. A slow rotary motion is maintained by the ther- mometer with its capillary until the fatty acids become trans- parent, when the temperature is noted. It is almost always necessary to conduct three tests; the first to get the approximate temperature, after which the heating may be done very slowly as the melting-point approaches, and the average of the two last determinations taken as the true point. 687. Detection of Phytosterol and Cholesterol. — Fats and oils of vegetable origin may be detected in those derived from animals by the presence of phytosterol and, vice versa, animal products in vegetable oils and fats by the presence of cholesterol. Boil 50 grams of the fat or oil in a flask having a reflux condenser, with 75 c.c. of 95% alcohol, for flve minutes, and separate the alcoholic solution. Repeat with another portion of alcohol and separate. Mix the alcoholic solution with 15 c.c. of 30% sodium hydroxide, 452 ANALYSIS OF FATS AND OILS. and boil in a flask with an air condenser until one-fourth of the alcohol is evaporated. Evaporate nearly to dryness in a porce- lain dish and extract the residue with successive portions of ether. Evaporate the ethereal solution to dryness, take up with a little ether, filter, again evaporate, dissolve in hot 95% alcohol, and allow to crystallize. Examine a few of the crystals under a micro- scope. Phytosterol thus prepared appears as tufts of needles, cholesterol as thin rhombic tables. The melting-point of phy- tosterol is 130°-137.5°, that of cholesterol 146°. The melting-point is better determined, however, on the acetates of these alcohols. The crystals obtained as described are heated in a dish, with 2-3 c.c. of acetic anhydride, over a small flame until it boils, the dish being covered with a watch-glass. The watch- glass is then removed and the excess of acetic anhydride evapo- rated off on the water-bath. The contents of the dish are next heated with the smallest quantity of absolute alcohol, and in order to prevent immediate solidification or crystallization, a few c.c. of alcohol are added, and the mass allowed to crystallize by spontaneous evaporation of the alcohol, of which one-half or a third is allowed to volatilize. The crystals are filtered off through a small filter and washed with a little 95% alcohol. They are then dissolved in 5-10 c.c. of hot absolute alcohol and again allowed to crystallize. These crystals are filtered off, pressed dry, and their melting-point determined. Cholesteryl acetate melts at 114.3° to 114.8° (corr.), and phj'tosteryl acetate at 125.6° to 137° (corr.). In doubtful cases the crystals should be redissolved in hot absolute alcohol and recrystallized at least three times before obtaining the melting-point. 688. Flash Point. — This is the temperature at which the vapors given off by the oil may be ignited, producing a fiash when brought in contact with a small flame. This test is made in order to deter- mine the temperature at which the use of the oil involves the danger of fires or explosions. The flash point varies with the form of apparatus and the method of making the test, so that the conditions must be rigidly fixed in order to obtain constant results. Various forms of apparatus have been devised, all of which embody a few essential features, such as a cup to hold the oil, some device for gradually raising the temperature of the oil, PHYSICAL TESTS. 453 a thermometer to indicate the temperature, and a small flame to be used for igniting the vapor. A cover containing two open- ings, one for the thermometer and another at which to apply the test-flame, is sometimes provided. Such an instrument is called a closed tester, while an instrument without a cover is called an open tester. The open testers give higher readings than the closed te'sters. More accurate determinations can be made with the closed testers. 689. Oil Testers. — A simple tester may be made from ordi- nary laboratory apparatus. A shallow porcelain crucible of about 100 c.c. capacity will serve as the cup to contain the oil. It may be heated on a sand-bath, or it may be placed in a circular opening in a piece of asbestos board and heated over a Bunsen-burner flame. A thermometer is suspended in the oil, which is then slowly heated while a small test-flame is passed over the surface of the oil. The test-flame may be produced by drawing out a piece of glass tubing to a fine tube and attaching it to the gas-supply by means of rubber tubing. The temperature is noted at the moment a flash is produced. After the oil has cooled some- what, the test is repeated, the oil being heated more slowly as the temperature approaches the point at which the oil flashes. By placing a metallic or glass cover containing two openings over the crucible, it can be used as a closed tester. Tagliabue's tester is constructed entirely of copper and con- tains two cups. The large one serves as a receptacle for a non- volatile oil in which the smaller cup containing the oil to be tested is immersed. The heat is applied to the larger cup. By this device the temperature of the oil in the inner cup is raised more uniformly. In a similar form of tester known as the "Say- bolt," the test-flame is replaced by a spark of electricity produced by an induction-coil. 690. Viscosity is that property of a liquid which determines the relative rate of flow imder fixed conditions. It is usually determined by noting the time required by a definite volume of oil at a definite temperature to flow through a small aperture. Water is usually taken as the standard. The simplest form of apparatus is that used in the Pennsylvania and other railroad laboratories. A 100-c.c. pipette, having a bulb 1.5 to 1.75 inches 454 ANALYSIS OF FATS AND OILS. in diameter and about 4J inches long, is graduated so that the 100-c.c. mark is just at the bottom of the bulb. The opening of the pipette is then made of such a size that 100 c.c. of water at 100° F. will run out of the pipette in 34 seconds. After dry- ing the pipette, the oil to be tested is warmed to 100° and 100 c.c. is drawn up into the pipette. The time required for the oil to flow out is noted. This gives the viscosity of the oil. It is also becoming customary to take the ratio between the number of seconds required by the water and the oil as the viscosity. A considerable number of instruments have been devised for this determination, Engler's, Taghabue's, Redwood's, and Gibbs' being standard instruments. All of these instruments contain a cup having a small opening at the bottom, through which the oil is allowed to flow. A jacket is generally provided which can be filled "\^'ith water or oil and heated, thus maintaining the oil at a constant temperature. The oil is delivered into a flask of known capacity. 691. Oil Analysis. — In the examination of an unknown oil or fat, the first test to be made is the determination as to whether the oil is of mineral or organic origin, or a mixture of these two classes of oils. For this purpose advantage is taken of the fact that organic fats and oils can be saponified, while mineral oils and waxes cannot be converted into a soap, but can be extracted with ether, while the soap produced from the organic fats and oils is insoluble in ether. On evaporating off the ether, the mineral oil can be recovered in pure condition. On acidifying the water solution of the soap, the fatty acids of the vegetable or animal oils separate out as an oily layer. As the fatty acids represent about 95% of most animal or vegetable oils, the weight of these acids can be used to calculate the amount of these constituents. A determination of the chemical and physical constants of the fatty acids can be made for the identification of the oils present. If mineral oils are absent, these determinations are made on the original oil. 692. The Saponification is carried out in the same manner as the determination of the Kottstorfer value (page 439). Stand- ard solutions of the caustic potash and acid are not necessary. PHYSICAL TESTS. 455 60 grams of caustic potash are dissolved in one liter of 95% alco- hol. 5 to 10 grams of the fat or oil to be tested are weighed out into an Erlenmeyer flask and 50 to 75 c.c. of the alcoholic caustic potash added. The flask is heated on the water-bath with a return condenser until the oil is saponified. 75 c.c. of water is then added and the heating continued without the condenser until the alcohol has been expelled. The soap solution is trans- ferred to a separatory funnel and 75 c.c. of ether added. The funnel is corked and the contents thoroughly shaken and then allowed to stand quietly until the upper ether solution of the unsaponified or mineral fat is completely separated from the lower water solution of the soap formed from the animal or vege- table fat. The lower layer is drawn off into a beaker, the stop- cock being closed when the last drop of the water solution has passed out. A little ether may be allowed to pass out with the water solution, but none of the latter must be allowed to remain with the ether. The ethereal solution is drawn off into a small weighed flask and the water solution returned to the funnel and the extraction with ether repeated. Three or four extractions with ether are generally sufficient to remove all the unsaponified fat or oil. The last portions of ether should leave no residue when allowed to evaporate on a watch-crystal. The flask contain- ing the ethereal solution is connected with an inclined condenser and the ether distilled off over a water-bath, or, still better, over an electric-light bulb. As the ether vapor is both inflammable and explosive when mixed with air, a flame should not be brought near vessels containing ether. The residue remaining after dis- tilling off the ether is dried on the water-bath, or at a lower tem- perature if volatile oils are present and weighed. This gives the weight of mineral oil present. 693. Determination of the Fatty Acids. — The aqueous solution of the soap is acidified with hydrochloric acid and heated on the water-bath until the fatty acids form a clear layer on the water solution. The oily layer is filtered off on a wet filter-paper which has been dried for one hour in the steam-oven together with a small beaker and weighed. The oil is washed with hot water without allowing the funnel to become more than half empty. When the chlorides have been completely washed out, the paper 456 ANALYSIS OF FATS AND OILS. containing the fatty acids is transferred to the beaker which was weighed with the paper, and the whole placed in the steam-oven to dry. After one hour it is weighed and again dried and weighed until the weight is constant. In order to determine what organic oils are present in the sample being analyzed, a determination of the various constants of the fatty acids is made. 694. Identification of Oils. — For the identification of an un- known organic oil, the most useful constant is the iodine value, on account of the great difference in this value for various oils. When two or more oils have the same iodine value as given in the table, page 457, other constants must be determined. An iodine value of 103, for example, might indicate cottonseed, mustard, rape, or sesame oil. A saponification value of 193 would show that the oil must be cottonseed or sesame oil, while an acetyl value of 16.6 would indicate that the oil must be cottonseed. This conclusion could be confirmed by qualitative color tests, of which some very excellent ones are known. CONSTANTS. 457 IODINE VALUE OF OILS. * By Habl's Solution. By Wij.s's Solution. ByHanus' Solution. Authority. Butter 35.3 34.8 30.7 36.2 35.9 35.3 35.4 30 6 Tolman and Munson a tl tl tl ii H3,nus ii 19.5 to 38.9 8.93 9.05 8.60 Cocoanut Tolman and Munson " 9.03 9.03 Hanus ti 8 to 9.5 Corn 119.0 119.0 123.3 124.8 123.2 122.2 129.2 128.4 120.2 119.6 126.0 Tolman and Munson n tt tl tl ({ tt tt tl il Wijs u 111.2 to 122.9 103.8 105.3 105.2 Cottonseed-oil. . . . Tolman and Munson It 106.2 107.3 107.8 tl tt tc it 104.8 100.9 to 116.9 106.2 106.7 tt tt te Konut 6.09 73.7 69.3 56.4 50.4 to 77.28 (Decreases to 22 6.43 74.5 70.5 with age) 6.40 73.9 69.8 56.9 tt te cs Lard-oil tt tc a '"^liich equals 96.06 __ _„ 246:532 «^ ^^-9^^- 748. Indirect Gravimetric Analyses are more difficult to cal- culate. General rules or factors cannot readily be given. The general method may be seen from the calculation of a determina- tion of potassium and sodium in a mixture of the chlorides of these metals. This determination is readily carried out by weigh- ing the sodium and potassium chlorides together and then deter- mining the 8.mount of chlorine present. As the percentage of chlorine in potassium chloride is 47.53, and in sodium chloride is 60.6, the percentage of chlorine in a mixture of these tv/o salts will be between these numbers. For example, a sample of the mixed chlorides which weighed 1.5802 grams gave .9023 gram of chlo- rine, which is equal to 57.1 per cent. If onlv potassium chloride 492 STOICHIOMETRY. had been present .7511 gram of chlorine would have been ob- tained. The difference between .9023 and .7511 is the amount of chlorine due to the presence of sodium chloride. This difference (.1512) divided by the difference in the percentage of chlorine in the two salts (.1307) gives the weight of sodium chloride present ( ^rT7yi=1.1568|. The weight of the mixed chlorides less the weight of sodium chloride gives the weight of potassium chloride present (1.5802- 1.1568 = .4234). The problem may be solved algebraically as follows: Let X =the weight of sodium chloride ; and y= " " " potassium chloride;, then x+ y = 1.5802 the weight of the mixed chlorides; and .606x +.47532/= .9023 the weight of chlorine. Solving these simultaneous equations we find that > .9023 -(1.5802X. 4753) x = - .606 -.4753 On carrying out the indicated operations we obtain a value for x, or the amount of sodium chloride present, which is identical with that already obtained. The disadvantage of using indirect methods of this kind is found in the fact that the difference upon which the weight of one of the constituents is calcvdated is very small. In the illustration given it must be multiplied by the factor 7.652 ( ^on? )- ^^^ experi- mental errors of the analysis are increased in this ratio in the per- centage found. The method is, therefore, not applicable to the determination of small amounts of one constituent in the presence of large amounts of the other. 749. Factor Weights. — ^When many determinations of a given constituent must be made it is found convenient to so choose the amount of substance to be weighed out that the weight of the precipitate obtained shall have a simple ratio to the percent- age to be calculated. For example, if 1.373 grams of a substance containing .375 per cent of sulphur is weighed out, the sulphur converted into sulphuric acid and weighed as barium sulphate, FACTOR WEIGHTS. 493 .0375 gram of the latter will be obtained. The number of milli- grams of barium sulphate obtained is evidently equal to the hun- dredths of per cent of sulphur present. The general method of finding the amount to be weighed out in a given case may be obtained from the following general case: Let w =the amoimt of substance weighed out; and V =the weight of precipitate obtained; and p =the per cent of the given constituent in the precipitate v. The percentage is obtained by the following indicated calcula- tions : pXv w If the value of w bears a simple ratio to p, the percentage will bear a simple ratio to v. In the illustration already given p, or the per- centage of sulphur in barium sulphate, is 13.73. w was made one- tenth of this number, or 1.373. The percentage is therefore ten times V or the weight of the bariimi sulphate obtained, as may be seen by substituting the values of v and w in the formula giving i;X 13.73 1.373 • If w is made 13.73 grams, the weight of v in grams will be equal to the percentage of sulphur in the substance analyzed. In general, then, the amoxmt of substance taken should bear a simple ratio to the percentage in the precipitate to be weighed of the constituent to be determined. This ratio will be inversely equal to the ratio of the percentage to the weight of the precipitate. For example, the percentage of chlorine in silver chloride being 24.72, the per- centage of chlorine in a given substance will be ten times the weight of silver chloride obtained if 2.472 grams of the substance to be analyzed are weighed out. CALCULATION OF THE FORMULAE OF SALTS AND ISOMORPHOUS MIXTURES. The results of many analyses may be expressed by a chemical formula. The composition of all pure chemical compounds as well as a large number of minerals may be expressed in this manner. All chemical formulas have been derived from quantitative anal- ^94 STOICHIOMETRY. yses and express primarily the proportions by weight in which the elements are present. 750. Salts. — Instead of using the percentage notation, the system of equivalent or atomic weights has been devised so that a single number or its simple multiple expresses the proportion in which a given element is present in its innumerable compounds. Sodium chloride, for example, contains 39.40 per cent of sodium, and 60.60 per cent of chlorine as foimd by analysis, but this fact is generally stated by giving the formula NaCl, which states that 23.05 parts of sodium are combined with 35.45 parts of chlorine. The ratio of the percentages 39.40 and 60.60 must be the same as the ratio of the numbers 23.05 and 35.45, and what is not quite so self-evident, the ratio of 23.05 to 39.40 must be the same as the ratio of 35.45 to 60.60. This ratio is 1.709; that is, 39.40 parts of sodium represent 1.709 atoms of this element and 60.60 parts of chlorine represent 1.709 atoms of chlorine. In this way it is found from the percentage composition and atomic weights that "n sodium chloride an equal number of sodium and chlorine atoms are present. If the percentages of copper and chlorine found in crystallized cupric chloride are divided by the atomic weights of these elements the numbers ( 37 30 41 58 ) and 1.1728 ]~^=.58Qi, 5.— ^ = 1.1728 [ .5864 c..^ ^.^.-^ 1 ggg .^^^., g.^g ..x.^^j are obtained. The number of clilorine atoms is evidently twice the number of .copper atoms. The formula then must be CuClj. The number of acid radicles, molecules of water, etc., may be found in the same manner, as may be seen from the calculation of the formula of potash alum from its analysis. Per Cent Found. Atomic Weights Atomic Ratios. K 8.17 39.14 ^^4^= -2087 Al 5.78 27.1 |^^= .2132 SO, 40.26 96.06 ^^7^7^=. 4190 40.26 96.06^ H,0 45.74 18.02 ||^=2.539 Total 99.95 CALCULATIUX OF THE FORMUL.H OF SALTS. 495 The atomic ratios of potassium and almniiiium being nearly equal, one atom of each of these elements will be present with two sul- phuric acid radicles and twelve molecules of water, since their atomic ratios are respectively twice and twelve times as great as that of potassium, giving the formula EA1(S04)2.12H20. 751. Many Minerals are found in the form of sharply defined homogeneous crystals, which contain varying amounts of several elements. These minerals contain several salts which crystallize in the same system, and are therefore called isomorphous. Dolo- mite is an example of such a mineral, in which varying amounts of magnesium and calcium carbonates are present. Small amounts of iron, aluminium, and silica are also present. The silicic acid is able to replace a portion of the carbonic acid forming calcium or magnesium silicates, while the iron and aluminium are able to replace a portion of the calcium and magnesium forming carbon- ates or silicates. A general formula, M"R03, will express the composition of the mineral. M" represents the bivalent bases, calcium, magnesium, ferrous iron, or even two valences of an aluminium atom. R represents the carbon or silicon. The method of obtaining the formula from an analysis is shown by the following analysis of dolomite : PercentofCO, 46.80 ||^ = 1.064 ^ 44. U fi2 " " "SiO, 62 -f-. = .010 ' 60.4 Total acid equivalents, R" 63 = 1.074 .21 '' " " FeO 21 ^Ym^ -^^ 32.35 *' " " CaO 32.35 -^g-j = .577 ** " " MgO 19.83 4^= -491 40 . 3d Total 99.81 Total basic equivalents, M"0 = 1 . 071 Ratio of acids to bases = .. ' „.. = 1 . 003 496 BTOU'HlOMETnY. The formula of the dolomite will therefore be MgCOgjCaCO,, small amounts of the base being replaced by FeO, and a little of the acid by SiO,. For each atom of bivalent metal there is present one acid radi- cle, the number 1.003 being so close to 1.000 that the difference may be ascribed to the unavoidable experimental errors of the analysis. Ratios as high as 1.02 or even 1.03 may be produced by errors which occur in fairly good analytical work, although the analytical worker should not be satisfied unless the ratio is usu- ally less than 1.01. BALANCING OF EQUATIONS. The calculation of nearly all volumetric determinations is based on the chemical equation which represents the reaction taking place. While it is true that the chemical equation is the expression of a quantitative analysis and should not be written unless the quantitative analysis has first been made, this view of the matter is taken only by the research chemist. In ordinary analyt- ical work the correctness of the equation is assumed, and the quantitative result is calculated from it. The equation must therefore be written before the calculation is made. 7S2. The Simple Synthetical Equation, such as 2H,+0,=2H,0 or CH4+203=C0,+2H,0, is required mainly for gas analysis. 753. The Analytical Equations, such as CuCp,=Cu+2C02 or 2Hp=2H,+ 02, are most commonly required to express electrical decompositions. When the elements or compounds produced by the reactiqn are known, the equations- are balanced by counting the number of atoms of each element on each side of the equation. For exam- ple, in the second equation given, the molecule of methane requires for its oxidation two atoms of oxygen for the carbon atom and two atoms for the four hydrogen atoms. The first member will BALA.\CLV<7 OP EQUATIONS. 497 therefore contain C ... 1, H ... 4, and ... 4, while the second member will contain C . . . 1, H . . . 4, and O . . . 4. These numbers being identical for each element, the equation is balanced. 754. Metathetical Equations are far more commonly met with, as they represent the two types of chemical reaction which are most commonly carried out; namely (a) those involving the pro- duction of a precipitate when two substances in solution are brought together, such as AgN03+ NaCl = AgCl + NaNO, or BaCl2+H,S04=BaSO,+2HCl; (6) those involving the production of a substance which is in the gaseous form under the conditions of the experiment, as CaCOj + 2HC1 = CaCl^ +00^+ H^O or 2NH,Cl+Ca(0H), =2'^Ii,+CB£\+2'RjJ. Equations such as these may be balanced by adding or sub- tracting molecules of one or the other substance until the number of atoms of each element is identical in each member of the equa- tion. It is far better, however, to take into consideration the fact that some of the elements are combined into groups or radicles which do not separate into their constituent atoms. NO3 is such a radicle, and in the first equation given the number of NO3 radicles on each side of the equation must be identical. SO^ acts as a radicle in the same manner. In the third equation the carbon remains combined with two atoms of oxygen as CO2, and two molecules of hydrochloric acid must be introduced, because one atom of calcium requires two atoms of chlorine to form calcium chloride, while the two atoms of hydrogen unite with the remain- ing atoms of oxygen in the calcium carbonate. In a similar man- ner the ammonia radicle NH3 is considered as a unit in the fourth equation. 755. Oxidation and Reduction Equations are more difficult to balance since two distinct reactions generally take place. The 498 STOICHIOMETRY. valence of two substances changes, involving the transference of oxygen or its equivalent, and in the second place an acid or an alkali reacts with some of the elements present to form salts. In balancing these equations the oxidizing and reducing substances must first be introduced in equivalent amounts. Then the amount of acid or base necessary to combine with the bases or acid liber- ated by the first reaction must be introduced. In order to balance the oxidizing and reducing equations the amount of oxygen absorbed or liberated must be known for each substance present. For instance, a molecule of potassium dichro- mate breaks up as follows: KjCr.O, = Kjd + Cvfi, + 30, liberating three atoms of oxygen. Potassium permanganate breaks up so as to give five atoms of oxygen for every two molecules of the salt, as follows : 2KMn04 =^,0 + 2MnO + 50. Amongst reducing substances two atoms of ferrous iron take up one atom of oxygen to form ferric iron while one atom of stan- nous tin takes up one atom' of oxygen or its equivalent. Two molecules of potassium permanganate giving five atoms of oxygen are therefore able to oxidize ten atoms of ferrous iron. If the iron is present as ferrous sulphate one molecule of sulphuric acid will be required to convert two atoms of ferrous iron into ferric iron as follows: 2FeS04+H2SO,+0=Fe2(SOJ3+H20. . . . (1) As five atoms of oxygen are given by the two molecules of potas- sium permanganate we must multiply each member of the equa- tion by five, giving: 10FeSO,+5H,SO,+ 5O=5Fe2(SO,)3+5H2O. . . . (2) The potassium and manganese of the potassium permanganate are also converted into sulphates, two molecules of potassium per- manganate requiring three molecules of sulphuric acid as follows: 2KMn04+3H,S04=K2SO,+ 2MnSO,+3H20 + 50. . . (3) CALCULATIOX OF VOLUMETRIC DETERMINATIONS. 499 Combining equation (2) and (3), we obtain the complete equation 10FeSO,+ 2KMnO,+ 8H2SO4 =5Fe,(SOJ3+ K,SO,+ 2MnS0, + 8Hp. The free oxygen disappears from both equations since it is present in the first member in the potassium permanganate and in the second as water. The oxidation of stannous chloride with potassium dichrornute can also be written in two equations which may then be combiner 1 into one. The potassium dichromate liberates oxygen in tlic presence of , , reducing agent according to the following equation : K,Cr207+8HCl=2KCl+2CrCl3+30+4H,0. . . . (1) The stannous chloride in the presence of acid and an oxidizing agent reacts as follows: SnCl2 + 2HCl+0=SnCl^+H20 (2) As one molecule of potassium dichromate liberates three atoms oi oxygen equation (2) must be multiplied by three, giving: 3SnCl,+ 6HCl + 30=3SnCl,+3Hp (3) By combining equations (1) and (3) we obtain the complete equation as follows : K,Cr,07+ 3SnCl,+ 14HC1 =2KC1+ 2CrCl3+ 3SnCU+ 7H,0. CALCULATION OF VOLUMETRIC DETERMINATIONS. As volumetric solutions are generally made on the normal basis of strength the calculation of the results is quite simple. Roundabout and very cumbersome methods of calculation are nevertheless often used. 756. Standard Solutions. — When solutions such as hydrochloric or sulphuric acid have been standardized by weighing precipi- tates of silver chloride or barium sulphate the weight of the acid radicle present in the precipitate may be computed. From this weight the corresponding weight of acid may be com- puted, and finally the ratio of this* weight to the theoretical wci;rht for a normal solution may be calculated. The same inverse ratio is taken I 7-^^Xo- I, the number .85712 is obtained 50O STOIC/IIOMETRY. result may, however, be obtained by a single calculation. As a liter of normal hydrochloric acid contains the gram molecular weight of the acid it must give a gram molecular weight of sil- ver chloride according to the equation, HC1+ AgNOg = AgCl+ HNO3. If 10 c.c. of the acid has been taken for the determination the silver chloride precipitate should weigh 1.4338 or 1/lOOth of the gram molecular weight. The ratio of this number to the weight found will give the strength of the acid in terms of normal. For example, if 1.6728 grams of AgCl were obtained, the acid is (1 6728\ YJoooj- To dilute the acid to exact strength there must be added 166.7 c.c. of water to one liter of the acid. If the a.4338 \ and the dilution must be made by diluting 857.12 c.c. of the strong acid to one liter. 757. Acidimetry and Alkalimetry. — Having obtained a normal acid solution the calculation of the amount of any substance titrated is equally simple. For all univalent bases titrated one hter of the acid is equal to the gram molecular weight of the base. For example, one liter of a normal acid, whether it is sulphuric or hydrochloric, is equal to 17.034 grams of ammonia, while one c.c. is equal to one-thousandth of this amount or .017034 gram. If 10 c.c. of acid has been used to titrate a given amount of ammo- nia, there is present .17034 gram of NH3. If this anmaonia has been formed from the nitrogen contained in a substance analyzed, the amount of nitrogen may be computed directly, since one liter of a normal acid will be equivalent to 14.01 grams of nitrogen in the form of ammonia. If, for example, the ammonia has been obtained from an ammonium salt, such as ammonium sulphate, the value of the normal acid in terms of this salt is obtained from its molecular weight, half of which is taken since each molecule contains two molecules of ammonia. The value of a standard acid in terms of a given salt will some- times vary with the indicator used, as already explained in Chap- ter XX, page 250. If disodic phosphate is titrated with a strong CALCULATION OF VOJMMETRIC DETERMINATIONS. 501 acid, using methyl orange as the indicator, the following reaction will take place: Na^HPO.+HCl =NaCl+NaH2P0,. A liter of normal acid will therefore be equal to a gram molecular weight of disodic phosphate, which will be 358.4 grams of the crys- tallized salt. While the disodic phosphate is alkaline to methyl orange it is neutral to phenolphthalein, to which the trisodium salt is alkaline. If the latter salt is titrated with acid, using phenol- phthalein as the indicator, the following reaction takes place : Na3P0,+ HCl =Na2HP0,+ NaCl. One liter of a normal acid is therefore equal to the gram molecular weight of this salt or 380.44 grams of the crystallized salt. A mixture of trisodic and disodic phosphates may therefore be ana- lyzed by titration with a standard acid, using phenolphthalein and methyl orange as the indicators. The method of calculation ia seen from the following illustration: One gram of a sample of disodic phosphate was titrated with fifth-normal acid. With phenolphthalein 1.50 c.c. of acid were used, and when methyl orange had been added 13.5 c.c. acid were required to again give the acid reaction. The acid used with the phenolphthalein rep- resents the amount of trisodic phosphate present, 1 c.c. of fifth- normal acid being equal to .07608 gram of this salt. Percentage of Na3PO,.12H,0 - '^^qqq^^ =11.41. The 13.50 c.c. of acid used with the methyl orange neutralized the disodic phosphate present in the original material and that pro- duced by the titration with phenolphthalein. The acid used for the latter purpose will be exactly equal to that used with the phenolphthalein, in this case 1.50 c.c. The acid necessary to neutralize the disodic phosphate originally present will be 12.00 c.c. (13.50-1.50). The percentage of Na2HP04.12H20 will be 12.00X.0717 LOOO ^^^■^' 502 STOICHIOMETRY. . 758. Percentage Given by Number of Cubic Centimeters of Acid Used. — ^Frequently it is convenient to use a solution of such a strength that when a given weight of substance is taken the number of cubic centimeters used in the titration shall be equal to the percentage of a given constituent. For this purpose the solution must be made of such a strength that 100 c.c. are equal to the amount of substance to be weighed out, provided the substance is absolutely pure. If sodium carbonate is the substance to be titrated and one gram is the amount to be 'weighed out, 100 c.c. of the acid must be equal to 1 gram of pure sodium carbonate. As 100 c.c. of a normal acid are equal to 5.3 grams of sodium carbonate, the acid required must be -53rd normal. If an acid double this strength is used the num- ber of cubic centimeters used must be multiplied by 2 to give the percentage. The amount of substance to be weighed out may also be so chosen that the percentage shall be equal to the number of cubic centimeters used in a titration. The acid may then be of any desired strength. The amount of substance weighed out must in every case be equal to the weight of the pure substance, which will just neutralize 100 c.c. of the acid. In the case of a fifth-nor- mal acid 100 c.c. are equal to 1.06 grams of pure sodium carbonate. If this amount of the impure article be weighed out and titrated with the acid the number of cubic centimeters used will be equal to the percentage of NajCOa present. This is evident from the consideration that while 1.06 grams of the pure compound will require 100 c.c, any decrease in the purity of the substance will produce a proportionate decrease in the amount of acid used in its titration. 759. Oxidation and Reduction Titrations. — In calculating the resillts obtained by means of oxidizing and reducing solutions the amount of oxygen absorbed or liberated by the substances acted upon must always be kept in mind. Equivalent weights of reducing or oxidizing substances absorb or give up the same amount of oxygen. It has already been shown that 31.63 grams of KMn04 contain the same amount (8 grams) of available oxygen as 49.05 grams of KjCrjO,. A Uter of solution containing either of these amounts of oxidizing substances is a normal solution. 126.97 grams of iodine liberate the same amount of oxygen to CALCULATION OF VOLUMETRIC DETERMINATIONS. 503 reducing substances, and dissolved in a liter give a normal oxidiz- ing solution. These different amounts of oxidizing substances are therefore equal to each other, because the same amount of free oxygen is furnished by each. The method of shortening calculations by applying this princi- ple may be seen from the following examples: A potassium per- manganate solution which had been standardized by means of pure iron wire was used to standardize a sodium thiosulphate solution. The standardization of the potassium permanganate solution with iron takes place according to the following equation: 2KMnO,+ 10FeSO4+ 8H3SO4 =K,SO,+ 5Fe2(SO,)3+ 2MnS04 + 8H,0. Two molecules of potassium permanganate are therefore equal to ten atoms of iron. A measured volume of the permanganate solution was acidified with hydrochloric acid after the addition of potassium iodide. Iodine was liberated according to the following equation: KMnO^ + 8HC1 + SKI = 51 + 6KC1 + MnCl^ + 4H,0. The iodine liberated is then titrated with sodium thiostdphate solution. By this equation two molecules of potassium perman- ganate are equal to ten atoms of iodine. As by the first equation the same amount of potassium permanganate is equal to ten atoms of iron, one atom of iron is equal to one atom of iodine ; that is, 55.9 parts of iron are equal to 126.97 parts of iodine. As 1 c.c. of the potassium permanganate solution was found equal to .01115 gram of iron the value of 1 c.c. in iodine is obtained from the following proportion: 55.9 : 126.85 :: .01115 : x x = .0253 gram. As 25 c.c. of the potassium permanganate solution required 23 c.c. of the thiosulphate solution to titrate the iodine liberated, the 25 X .0253 iodine value of 1 c.c. of the latter solution is equal to ^5 > °^ ,0275 gram. 504 STOICHIOMETRY 760. lodometric Titrations. — In a similar manner the value of the thiosulphate solution in terms of any substance it may be used to titrate may be calculated. The intermediate reactions are used simply to find the number of iodine atoms equivalent to one mole- cule of the substance titrated. For example, manganese dioxide may be boiled with hydrochloric acid, the chlorine evolved being absorbed in potassium iodide, and the iodine liberated titrated with the thiosulphate solution. The following reactions take place: Mn02+ 4HC1 =MnCl2+ CI2+ 2H2O Cl,+2KI=2KCl+l2 I^+NaAOs =2NaI+Na2S40e. One molecule of manganese dioxide is therefore equal to two atoms of iodine. The value of 1 c.c. of the sodium thiosulphate solution in terms of manganese dioxide is obtained from the following pro- portion: I2 : MnO, : : .0275 : x 253.7: 87 : : .0275 : a; a; =.00943 APPENDIX. REAGENTS. It is important that the reagents used should be of known defi- nite strength and that the strength should be indicated on the bottles. So far as possible also the reagents should be of uniform strength. This is not always possible on account of the solubilities of the salts employed. The strength is most easily indicated by the normal system of nomenclature. Chemically pure material should always be used. The impurities which may be present and the tests to be applied are given in the notes on pp. 32-35. Many salts when in solution, especially those that react alkaline, act on the glass becoming contaminated with silica and other constituents of the glass. Only recently prepared solutions of such reagents should be used. It is therefore advisable to keep the dry salts at hand and make up the solutions only as required for immediate use. Reagents of Five Times Normal Strength. ' Label, 5 N. Sulphuric Acid, HjSO^ Equivalent =49 Ordinary concentrated acid, sp. gr. 1.84, has a strength thirty- six times normal, and may be designated 36 N. One volume of strong acid diluted with 6 volumes of water =5 N. Nitric Acid, HNO3 Equivalent =63 Concentrated acid, sp. gr. 1.40 = 15 N. One volume of strong acid diluted with 2 volumes of water =5 N. Hydrochloric Acid, HCl Equivalent =36.5 Concentrated acid, sp. gr. 1.20 = 13 N. Five volumes of strong acid diluted with 8 volumes of water =5 N. 505 506 APPENDIX. Acetic Acid, H.C2H3O2 Equivalent =60 Glacial acid, M.P. 10° C, =17 N. One volume of glacial acid diluted with 2^ volumes of water =5 N. Potassium Hyd--3zide, KOH Equivalent =56 280 grams dissolved in water and diluted to one liter =5 N. Sodium Hydroxide, NaOH Equivalent =40 200 grams dissolved in water and diluted to one liter =5 N. Ammonia (solution in water, NH4OH) Equivalent =35 The strong solution, sp. gr. 0.90 = 15 N. One volume of the strong solution diluted with 2 volumes of water =5 N. Ammonium Sulphide, (NH4)2S Equivalent =34 600 c.c. of 5N ammonia are saturated with sulphuretted hydrogen; this gives hydrogen ammonium sulphide, NH^HS. This is made up to 1 liter by adding 5N ammonia. (This reagent is slowly decomposed by atmospheric oxygen, ammo- aia is evolved, and yellow ammonium sulphide, (NH4)2S2, is formed.) Sodium Sulphide, NazS Equivalent =39 200 grams of sodium hydroxide are dissolved in 800 c.c. of water. 400 c.c. of this solution are saturated with sulphuretted hydrogen and the remaining half added, together with water sufficient to make the volume up to 1 liter. When hydrogen sodium sulphide, NaHS, is required, the sodium hydroxide is simply saturated with sulphuretted hydrogen, without the further addition of sodium hydroxide. Ammonium Chloride, NH4CI Equivalent =53.5 267.5 grams of the salt dissolved in water and diluted to one liter = 5 N. Ammonium Carbonate, (NH4)2C03 Equivalent =48 200 grams of ammonium sesquicarbonate (commercial carbon- ate) dissolved in 350 c.c. of 5 N ammonia, and diluted with water to one liter =5 N. Ammonium Acetate, NH4C2H3O2 Equivalent =77 To 300 c.c. of glacial acetic acid (17 N) an equal volume of water is added, the acid neutralized with strong ammonia and diluted to one liter =5 N. REAGENTS. 507 Reagents of Normal Strength. Label, N. The following normal reagents are prepared by dissolving the equivalent weight in grams of the various salts in water, and diluting to one liter. In all cases the nearest whole number to the equivalent weight may be taken as sufhciently exact. Barium chloride, BaCl2.2H20 Equivalent weight 122.0 Disodium phosphate, HNa2P04.12H20. " " 119.3 Lead acetate, Pb(C2H302),.3H20 " " 189.5 " Magnesia mixture " (MgCl2. (NH,C1)„ and NH.OH). 68 grams of MgCl2.6H20, together with 165 grams of ammo- nium chloride, are dissolved in 300 c.c. of water; 300 c.c. of 5 N ammonia are added, and the solution diluted with water to one liter. Stannous chloride, SnCl2.2H20 . .. . Equivalent weight 112.5 112 grams of the salt are dissolved in 200 c.c. of 5 N hydro- chloric acid and the solution diluted with water to one liter. Fragments of granulated tin should be placed in the solution. Reagents op Various Strengths. Ammonium oxalate, (NH4)2C204.2H20. Equivalent weight 80.0 N 40 grams dissolved in one liter = y solution. Molybdate Solution. — ^The molybdate solution is made as fol- lows: 75 grams of ammonium molybdate are dissolved in 500 c.c. of water with the addition of a little ammonia if necessary. If still turbid, the solution is filtered and then poured with constant stirring into 500 c.c. of a mixture of 250 c.c. of concentrated nitric acid (sp. gr. 1.40) and 250 c.c. of water. The solution may also be made from molybdic oxide by dissolving 60 grams in 440 c.c. of water and 60 c.c. strong ammonia (sp. gr. 0.90), and pouring the solution into 500 c.c. of nitric acid diluted as already directed. The freshly made molybdate solution must be allowed to stand in a warm place for several days. The clear solution is decanted or filtered off for use. Large amounts of the solution should not be 503 APPENDIX. made up at one time, since it will not keep more than several months, as the molybdic acid is slowly precipitated. Mercuric chloride, HgClz Equivalent weight 135.5 N 27 grams dissolved in one liter =- solution. 5 Silver nitrate, AgNOs Equivalent weight 170.0 N 3.4 grams dissolved in 100 c.c. =^ solution. Potassium dichromate, K2Cr207 Equivalent weight 294.5 100 grams are dissolved in one liter of water. INTERNATIONAL ATOMIC WEIGHTS FOR 1916 0=16 Aluminum Al 27 . 1 Antimony Sb 120.2 Argon... A 39.88 Arsenic As 74 . 96 Barium Ba 137 . 37 Bismuth Bi 208.0 Boron B 11.0 Bromine Br 79.92 Cadmium Cd 112.40 Caesium Cs 132.81 Calcium Ca 40.07 Carbon C 12.005 Cerium Ce 140.25 Chlorine CI 35.46 Chromium Cr 52 . Cobalt Co 58.97 Columbium Cb 93 . 5 Copper Cu 63.57 Dysprosium Dy 162 . 5 Erbium Er 167.7 Europium Eu 152.0 Fluorine F 19.0 Gadolinium Gd 157 . 3 Gallium Ga 69.9 Germanium Ge 72 . 5 Glucinum Gl 9.1 Gold Au 197.2 Helium He 4.00 Holmium Ho l63 . 5 Hydrogen H 1.008 Indium In 114.8 Iodine I 126.92 Iridium Ir 193.1 Iron Fe 55.84 Krypton Kr 82.92 Lanthanum La 139.0 Lead Pb 207.20 Lithium Li 6.94 Lutecium Lu 175 . Magnesium Mg 24.32 Manganese Mn 54.93 Mercury Hg 200.6 = .16. Molybdenum Mo 96 . Neodymium Nd 144.3 Neon Ne 20.2 Nickel Ni 58.68 Niton (radium emanation) Nt 222.4 Nitrogen N 14.01 Osmium Os 190.9 Oxygen O 16.00 Palladium Pd 106. 7 Phosphorus. P 31 . 04 Platinum Pt 195.2 Potassium K 39.10 Praseodymium Pr 140 . 9 Radium Ra 226.0 Rhodium Rh 102.9 Rubidium Rb 85.45 Ruthenium Ru 101 . 7 Samarium Sa 150 . 4 Scandium Sc 44 . 1 Selenium Se 79.2 Silicon Si 28.3 Silver Ag 107.88 Sodium Na 23.00 Strontium Sr 87 . 63 Sulfur S 32 09 Tantalum Ta 181.5 Tellurium Te 127.5 Terbium Tb 159.2 Thallium Tl 204.0 Thorium Th 232.4 Thulium Tm 168.5 Tin Sn 118 7 Titanium Ti 48 1 Tungsten W 184.0 Uranium U 238.2 Vanadium V 610 Xenon Xe 130.2 Ytterbium (Neoytter- bium) Yb 173 5 Yttrium Yt 88.7 Zinc Zn "65.37 Zirconium Zr 90 . 6 Compiled by the International Committee on Atomic Weights, con- sisting of F. W. Clarke, W. Ostwald, T. E. Thorpe, and G. Urbain. Jour. Am. Chem. Soc, 1915, 37, 2451. .509j 610 TABLES. TABLE II.— CHEMICAL FACTORS AND THEIR LOaARITHMS. Factor. • Atomic Weight. Weighed. Bequired Per Cent. Loga- rithms.* Aluminium 27.1 Al,Os Al 53.03 9.72455 AlPO. Al 22.19 9.34621 il AUOs 41.84 9.62166 Antimony 120.2 SbaOi Sb 78.97 9.89749 Arsenic 74.96 As=S3 As 60.92 9.78475 " ASjOa 80.41 9.90531 MgaASaOT As 48.27 9.68371 (C AsjOs 63.73 9.80435 It As,0, 74.03 9.86941 Barium ........ 137.37 BaS04 Ba 58.85 9.76973 (( BaCl, 89.23 9.95051 ** BaCl3.2H20 104.66 0.01980 ii BaO 65.70 9.81758 BaCOa Ba 69.60 9.84259 li BaO 77.70 9.89044 " C 06.08 8.78390 n CO2 22.29 9.34820 BaCrO. Ba 54.20 9:73396 it BaO 06.51 9.78181 Bismuth 208.0 BUO, Bi 89.65 9.95257 BiOCl Bi 80.17 9.90399 U BijOs 89.42 9.95142 Bromine 79.92 AgBr Br 42.56 9.62896 ii HBr 43.11 9.63461 Cadmium 112.40 CdO Cd 87.54 9.94220 CdS Cd 77.83 9.89112 i( CdO 88.90 9.94892 Calcium 40.07 CaO Ca 71.47 9.85409 " CaCOa 178.50 0.25159 CaCOa Ca 40.04 9.60250 il CaO 56.03 9.74841 CaSOi Ca 29.43 9.46886 " CaO 41.19 9.61477 Carbon 12.005 CO, C 27.27 9.43570 Chilorlne 35.46 AgCl CI 24.74 9.39337 " HCl 25.44 9.40543 Ag CI 32.87 9.51680 HCl 33.80 9.52886 Chromium 52.0 Cr.Os Cr 68.42 9.83519 li CrOa 131.57 0.11916 " Cr04 152.63 0.18364 PbCrOi Cr 16.09 9.20653 li CrOa 30.94 9.49053 " CrO. 35.89 9.55499 li CrjOa 23.51 9.37134 BaCrOi Cr 20.53 9.31236 il CrOa 39.47 9.59626 " CrOi 45.78 9.66072 (( Cr.Oa 29.99 9.47707 Cobalt 58.97 Co CoO 127.14 0.10426 CoSO, Co 38.04 9.68022 ii CoO 48.36 9.6.S448 2CoS0..3K,SO, Co 14.16 9.15110 ti CoO 18.00 9.25636 Copper 63.57 Cu CuO 125.17 0.09750 CuSO^.SHjO 392.82 0.59419 * The-10 after each of the logarithms has been omitted. TAjBLES. 511 CHEMICAL FACTORS AND THEIR LOGARITHMS- (Oora6'l8 22 3424 3444 3464 3483^3502 3522 3541 3560 3579 3598 2 4 6 8 10 12 U x5 17 23 3617 3636 3655 3674;3692 3711 3729 3747 3766 3784 2 4 6 9 11 13 15 17 , 24 3802 3820 383 J 385613^74 3892 3909 3927 3945 3962 2 4 6 7 9 11 12 14 16 > 25 3979 3997 4014 4031 4048 4065 4082 4099 4116 4133 2 3 S 7 9 10 12 14 15 26 4150 4166 4183 4200 4216 4232 4249 4265 42 1 4298 2 3 5 7 8 10 11 13,15 27 4314'4330 4346 4362 4378 4393 4409 4425 4440 4456 2 3 5 6 8 9 11 13^14 23 4472|44S7 4502 4518|4533 4548 4564 4579 4594 4609 2 3 S e? 8 9 11 12 14 29 4624 4639 4654 4669 4683 4698 4713 4728 4742 4757 1 3 4 6 7 9 10 12 13 30 4771 4786 4800 48144829 4843 4857 4871 4886 4900 3 4 6 7 9 10 1113 31 4914 4928 4942 4955 4969 49«3 4997 5011 5024 5038 3 4 6 7 8 10 11:12 32 5051 5065 5079 50925105 5119 5132 5145 5159 5172 3 4 5 7 8 9 11 12 33 5185 5198 5211 522415237 5250 5263 5276 5289 5302 3 4 5 6 8 9 1012 34 5315 5328 5340 5353 5366 5378 5391 5403 5416 5428 3 4 5 6 8 9 lo'ii 1 ) 35 5441 5453 5465 5478 5490 5502 5514 5527 5539 5561 2 4 5 6 7 9 1 10 11 36 5563 5575 5587 5599 5611 5623 5635 5647 5658 5670 2 4 6 6 7 8 10 11 37 56 2 5694 5705 5717 5729 5740 5752 5763 5775 5786 2 3 S 6 7 8 9 10 38 5798 5809 5821 5832 5843 5855 5866 6877 58SR 5899 2 3 S 6 7 8 9 10 39 5911 5922 5933 5944 5955 5966 5977 6988 6999 6010 2 3 4 5 7 8 9 10 40 6021 6031 6042 6053 6064 6075 6085 6096 6107 6117 3 4 5 6 8 9 10 41 6128 6138 6149 6160 6170 6180 6191 6201 6212 6222 2 3 4 5 6 7 8 9 42 6232 6243:6253 6261 6274 6284 6294 6304 6314 6325 2 3 4 5 6 7 8 9 43 6335 6345l6355'6365 6375 6385 6395 6405 6415 642S 2 3 4 5 6 7 8 9 44 6435 6444 6454 6464 6474 6484 6493 6503 6613 6522 2 3 4 5 6 7 8 9 45 6532 6542 6551 6561 6571 6580 6590 6599 6609 6618 2 3 . 5 6 7^8 9 46 6628 6637 6646 6656 6665 6675 6684 6693 6702 6712 2 3 4 5 6 7 7 8 47 6721 6730 6739 6749 6758 6767 6776 6785 6794 6803 9 3 4 5 5 7 8 48 6812 6821 6830 6839 6848 6867 6866 6875 6884 6893 2 3 4 4 5 u' 7: 8 49 6902 6911 6920 6928 6937 6946 6955 6964 6972 6981 2 3 4 4 5 6 7' 8 1 50 6990 6998 7007 7016 7024 7033 7042 7050 7059 7067 2 3 3 4 5 6 8 51 7076 7034 7093 7101 7110 711S 7126 7135 7143 7162 2 3 3 4 5 6 7 8 52 7160 7168 7177 7185 7193 7202 7210 7218 7226 7236' 1 2 2 3 4 5 6 7 7 53 7243 7251 7259 7267 7275 7284 7292 7300 7308 7316] 1 2 2 3 4 5 616:7 54 7324 7332 7340 734? 7356 7364 7372 7380 7388 7396 1 o 2 3 4 5 C 6 1 ' TABLES. 515 LOGARITHMS. 2 Phoportional Pabts. ll 1 2 3 4 5 6 7 s 9 si 1 3 2 3 2 i 3 5 4 6 5 7 5 8 7 9 55 7404 7412 7419 7427 7435 7443 7461 7459 7466 7474 7 56 74S2 7490 7497 7505 7513 7520 7528 7536 7543 7551 2 2 3 4 S 5 6 7 57 7559 7566 7574 7682 7689 7697 7604 7612 7619;7627 2 2 3 4 5 5 6 7 58 7634 7642 7649 7667 7664 7672 7679 7686 769417701 2 3 4 4 5 6 7 59 7709 77167723 7731 7738 7745 7752 7760 7767 7774 2 3 4 4 6 6 7 60 7782 77897796 7803 7810 7818 7825 7832 7839 7846 2 3 4 4 5 6 6 61 7853 7860 7868 7875 7882 7889 789 7903 7910 7917 2 3 4 4 5 fi 6 62 7924 793117938 7946 7952 7959 7966 7973 7980 7987 2 3 3 4 5 6 6 63 7993 8000 8007,8014 8021 8028 8035 8041 8048 8055 2 3 3 4 S 5 6 .64 8062 8069 80758082 8089 8096'8102 8109 81168122 2 3 3 4 6 5 6 65 8129 8136 8142 8149 8166 8162 8169 8176 81828189 2 3 3 4 5 6 6 66 . 8196 8202 820918215 8222 8228 8235 8241 8248 8254 2 3 3 4 6 5 6 67 8261 8267 8274 8280 8287 8293 8299 8306 8312 8319 2 3 3 4 5 5 6 68 S325S33l'8338:8344 8351 8367 8363 8370 8376 8382 2 3 3 4 4 5 6 69 8388 8395 8401 '8407 8414 8420 8426 8432,8439,8445 1 2 2 3 4 4 5 6 70 8461 8457 8463 8470 8476 8482 8488 8494 8600 '8506 2 2 3 4 4 5 6 71 8513 8519'8525 8531 8537 8543 8549 8556 8561 8567 2 2 3 4 4 5 5 72 857318579 8635 8591,8597 8603 8609 8615 86218627 2 2 3 4 4 5 5 73 8633 8639 8645 86518657 8663 8669 8675 86818686 2 2 3 4 4 5 5 74 8692 86988704 8710 8716 8722 8727 8733 8739 8745 2 2 3 4 4 5 5 75 8751 8756 8762 8768 8774 8779 8785 8791 8797,8802 2 2 3 3 4 5 5 76 8808 8814,8820 88258831 8837 8842 8848|8854 8859 2 2 3 3 4 5 5 77 8865 887r8876;8882 8887 8893 8899 8904 8910 8915 2 2 3 3 4 4 5 73 8921 8927 8932 893818943 8949 8954 8960 8965'8971 2 2 3 3 4 4 5 79 8976 8982 8937 8993 8993 9004 9009 9015 9020:9025 1 2 2 3 3 4 4 5 SO 9031 9036 9042 9047 9053 9058 9063 9069 9074 9079 2 2 3 3 4 4 5 81 9086 9090l9096'910] ;9106l9112 9117 9122 9128 9133 2 2 3 3 4 4 5 82 9138 9143 9149 9154 9159 9165 9170 9175 9180 9186 2 2 3 3 4 4 5 83 9191 9196 9201 9206 9212 9217 9222 9227 9232 9238 2 2 3 3 4 4 5 84 9243 9248 9253 9258 9263 9269 9274 9279 9284 9289 2 2 3 3 4 4 5 86 9294 9299 93049309 9315 9320 9326 9330 9336 9340 2 2 3 3 4 4 5 £6 9345 9350' 9355 9360 9366 9370 9375 9380 9385 9390 2 2 3 3 4 4 5 87 9395 9400 9405 941019415 9420 9426 9430 9436 9440 2 2 3 3 4 4 88 9445 9450 9456 9460 9465 9469 9474 9479 9484 9489 2 2 3 o 4 4 89 9494 9499 9504|9509 1 9613 951896239528 9533,9538 2 2 3 3 4 4 90 9542 9547 9552,9567 9662 9566 9671 9576 95819586 2 2 3 3 4 4 91 9590 9695 9600 96059609 9614 9619 9624 962819633 2 2 3 3 4 4 92 963S 9643 9647 9652'9657 96619666 9671 9675 9680 2 2 3 3 4 4 93 9685 9689 9694 96999703 9703 9713 9717 9722 9727 2 2 3 3 4 4 94 9731 9736 9741 97459750 9754 9769 9763 9768 9773 2 2 3 3 4 4 95 9777 9782 9786 97919795 9800 9805 9809 9814 9818 2 2 3 3 4 4 96 9823 9827 9832 9836:9841 9845 9850 9854,9869 9863 2 2 3 3 4 4 97 9868 9872 9877:9881 '9886 9890 9894 9899,9903 9908 2 2 3 3 4 4 98 9912 9917 9921 9926 9930i9934 9939 994319948 9952 2 2 3 3 4 4 99 9956 9961 9966 9969,9974 1 r I 9978 9983 9987 9991 9996 i: 2 2 3 3 3 4 516 TABLES. TABLE IV.— ANTILOGARITHMS. .00 .01 .02 .03 .04 .05 .06 .07 .08 .09 .10 .11 :i2 .13 .14 .15 .16 .17 .18 .19 .20 .21 .22 .23 .24 .25 .26 .27 .28 .29 .30 .31 .32 .33 .34 .35 .36 .37 .38 .39 .40 .41 .42 .43 .44 .45 .46 .47 .48 .49 1000 1023 1047 1072 1096 1122 1148 1175 1202 1230 1259 1288 1318 1349 1380 1413 1445 1479 1514 1549 1585 1622 1660 1698 1738 1778 1820 1862 1905 1950 1995 2042 2089 2138 2188 2239 2291 2344 2399 2455 2512 2570 2630 2692 2754 1002 1026 1050 1074 1099 1125 1151 1178 1205 1233 1262 1291 1321 1352 1384 1416 1449 1483 1517 1552 1589 1626 1663 1702 1005 1028 1052 1076 1102 1127 1153 1180 1208 1236 1265 1294 1324 1355 1387 1419 1452 1486 1521 1556 1592 1629 1667 1706 1007 1030 1054 1009 1033 1057 1079 1081 1104 1107 1130 1156 1183 1211 1239 1268 1297 1327 135° 1390 1132 1159 1186 1213 1242 1271 1300 1330 1361 1393 1012 1035 1059 1084 1109 1135 1161 1189 1216 1422 1426 1455 1459 148911493 A 1 1560 1596 1633 1671 1710 1750 17421746 1782|l786 1791 1824 1866 1910 1954 2000 2046 2094 2143 2193 1828 1871 1914 1959 2004 2051 2099 2148 2198 1832 1875 1919 1963 2818 28^5 2244 2296 2350 2404 2460 2518 2576 2636 2698 2761 1563 1600 1637 1675 1714 1754 1795 1837 1879 1923 1968 1014 1038 1062 1086 1112 1138 1164 1191 1219 1016 1040 1064 1089 1114 1140 1167 1194 1222 1250 1246,1247 127411276 1279 1282 1285 1019 1042 1067 1091 1117 1143 1-169 1197 1225 1253 1021 1045 1069 1094 1119 1146 1172 1199 1227 1266 1303 1334 1365 1396 1429 1462 1496 1531 1567 1603 1641 1679 1718 1758 1799 1841 1884 13061309 13371340 1368!l371 1400 1403 1312 1343 1374 1928 1932 2009 2014 2056 2061 2104 2153 2203 2249 22.54 2301 2307 2355 2410 2466 2523 2582 2642 2704 2767 2773 2884 2951 3020 3090 2891 2958 3027 3097 1432 1466 1500 1535 1570 1607 1644 1683 1722 1762 1803 1845 1888 1436 1469 1503 1538 1674 1611 1648 1687 1726 1766 1807 1849 1892 1936 1315 1346 1377 1406,1409 1439 1442 1472 1476 15071510 1542 1645 1678 1681 1614 1618 1652 1666 16901694 1730il734 1770 1774 1811 1854 1897 1941 1972 1977|l982 1986 2360 2415 2472 2529 2588 2649 2710 2831 2897 2965 3034 3105 2109 2158 2208 2259 2312 2366 2421 2477 2636 2594 2656 2716 2780 2018 2023 2028 2065 2070 2075 211312118 21632168 2213,2218 2123 2173 2223 2266 2317 2371 2427 2483 2270 2323 2377 2432 2489 2838 2904 2972 3041 3112 2541 2600 2661 2723 2786 2851 2917 2985 3055 2032 2037 2080 2084 2128^2133 2178' 2183 2228 2234 2275 2280 2328 2333 2844 2911 2979 3048 31193126 2647 2606 2667 2729 2793 2858 2924 2992 3062 3133 23s 2 2438 2495 12653 2612 2673 2736 2799 2388 2443 2500 2659 2618 2679 2742 2805 1 2286 1 2339 1 2393 1 2449 2506 Propobtiokai. Pakts. 13 3 4 1816 1858 1901 1945 1991 2''64 2871 2931 2999 3069 3141 2938 3006 3076 3148 2564 2624 2635 2748 2812 2877 2944 3013 3083 3156 5 6 7 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 3. 3 4 3 4 3 4 3 4 3 4 3 4 5!5 5I5 ■si sle 6 5 5 5 6 TABLES. 517 ANTILOGARITHMS. 1 .50 .51 .52 .53 .54 .55 .56 .57 .53 .59 .60 .61 .62 .63 .64 .65 .66 .67 .68 .69 •70 .71 .72 .73 .74 .75 .76 .77 .78 .79 .80 .81 .82 .83 .84 .85 ,86 .87 .90 .91 .92 .93 .94 .95 .96 .97 .98 .99 3162 3170 3236 3311 3388 3467 3548 3631 3715 3802 3890 3981 4074 4169 4266 4365 4375 3243 3319 3396 3476 3177 3251 3327 3404 3483 3556 3565 3639 364 3724 3811 3899 3990 4083 4178 4276 4467 4571 4677 4477 4581 4688 47864797 4898 4909 5012 5129 5248 5370 5495 5623 5754 5888 6026 6166 6310 6457 6607 6761 6918 7079 7244 7413 7586 7762 7943 8128 8318 8511 8710 8913 9120 9333 9550 9772 5023 5140 5260 5383 5508 3733 3819 3908 3999 4093 4188 4285 4385 4487 4592 4803 4920 3184 3258 3334 3192 3266 3342 3412 3420 3491 3499 3573 3656 3741 3828 3917 4009 4102 4198 4295 4395 449C 4603 4710 4819 4932 5636 5768 5902 6039 6180 6324 6471 6622 6776 6934 7096 7261 7430 7603 7780 7962 8147 8337 8531 8730 8933 9141 9354 9572 9795 5035 5047 5152'5164 6272] 5284 5395,5408 55215534 5649 5662 57815794 59165929 6053 6067 6194 6339 6209 6353 64866501 66376653 67926808 69506966 7112 7278 7447 7621 7798 7980 8166 8356 8551 8750 8954 9162 9376 9594 9817 7129 7295 7464 7638 7816 7998 8185 8375 8670 8770 8974 9183 9397 9616 9840 3681 3664 3750 3837 3926 4018 4111 4207 4305 4406 450: 4613 4721 4831 4943 5058 5176 5297 5420 5646 6675 5808 5943 6081 6223 6368 6516 6668 6823 6982 7145 7311 7482 7656 7834 3199 3273 3350 3428 3508 3589 3673 3758 3846 3936 4027 4121 4217 4315 4416 4519 4624 4732 4842 4955 5070 6188 5309 6433 5559 5689 5821 6957 6096 6237 6383 6531 6683 6839 8017 8204 8396 8590 8790 8995 9204 9419 9638 9863 7161 7328 7499 7674 7852 8035 8222 8414 8610 8810 9016 9226 9441 9661 9886 3206 3281 3357 3436 3516 3597 3681 3767 3855 3945 4036 4130 4227 4325 4426 4529 4634 4742 4863 4966 50S2 5200 5321 5445 5572 5702 5834 5970 6109 6252 6397 6546 6699 6855 7015 7178 7345 7516 7691 7870 8054 8241 8433 8630 8831 9036 9247 9462 9683 9908 3214 3239 3365 3443 3524 3606 3690 3776 3864 3954 4046 4140 4236 4335 4436 4539 4645 4753 4864 4977 5093 6212 6333 6453 6585 5716 5848 5984 6124 6266 6412 6561 6714 6871 7031 7194 7362 7534 7709 7889 8072 8260 8463 8660 8851 3221 3228 3296 3304 3373 3381 3451 3532 3614 3698 3784 3873 3963 4055 4150 4246 4345 4446 4550 4666 4764 4876 4989 5105 6224 5346 6470 5598 5728 6861 5998 6138 6281 6427 6577 6730 6887 7047 7211 7379 7551 7727 7907 9067 9268 9484 9705 9931 3459 3540 3622 3707 3793 3882 3972 4064 4169 4256 4355 4457 4560 4667 4775 4887 5000 6117 6236 635S 5483 6610 5741 5875 6012 6152 6295 6442 6592 6745 6902 7063 7228 7396 7668 7745 7926 8091 8279 8472 8670 8872 9078 9290 9506 9727 9954 Pkopobtional Parts. 133456789 8110 8299 8492 8690 8892 9099 9311 9528 9750 9977 7 6i7 6 7 6 7 6 7 9 11 loln lOlll lO'll 10 12 5 7 9 9 10 10 10 10 11 11 11 11 13 12 12 12 13 13 13 14 14 14 15 15 15 16 16 16 17 17 17 18 IS 19 19 20 20 20 518 TABLES. TABLE v.— SPECIFIC GRAVITY OF AMMONIA SOLUTIONS AT 15° C. Lunge and Wibenik.* Specific Gravity in Vacuo. Per Cent NH3, Grams. 1 Litef Contains NHa, Grams. Correction of Specific Gravity for±l«. Specific Gravity i 15° at 4^ in Vacuo. Per Cent NH3, Grams. 1 Liter Contains NH3, Grams. Correction of Specific Gravity for ± 1°. 1.000 0.00 0.0 0,00018 0.940 15.63 146.9 0.00039 0.998 0.45 4.5 0.00018 0.938 16,22 152,1 0.00040 0.996 0.91 9.1 0.00019 0.936 16,82 157,4 0.00041 0.994 1.37 13.6 0.00019 0.934 17,42 162.7 0.00041 0.992 1.84 IS. 2 0.00020 0.932 18, "3 163,1 0.00042 0.990 2.31 22.9 0.00020 0.930 18,64 173,4 0.00042 0.988 2.£0 27.7 0.00021 0.928 19,25 173,6 0.00043 0.986 3.30 32.5 0,00021 0.926 19.87 184,2 0.00044 0.984 3.80 37.4 0.00022 0.924 20.49 1^9,3 0.00045 0.932 4.30 42.2 0.00022 0.922 21.12 194,7 0.00046 0.9S0 4. SO 47.0 0.00023 0.920 21,75 200,1 0.00047 0.97^ 5.30 51.8 0.00023 0.918 22,39 205.6 0.00048 0.976 5.S0 56.6 0.00024 0.916 23.03 210.9 0.00049 0.974 6.30 61.4 0.00024 0.914 23.63 216.3 0.00050 0.972 6.80 68.1 0.00025 0.912 24.33 221.9 0.00051 0.970 7.31 70.9 0.00025 0.910 24.99 227.4 0.00052 0.968 7.82 75.7 0.00026 0.903 25.65 232.9 0.00053 0.966 8.33 £0.5 0.00026 0.906 26.31 233.3 0.00054 964 8.84 £5.2 0.00027 0.904 26.93 243.9 0.00055 0.962 9.35 £9.9 0.00028 0.902 27.65 249.4 0.00056 0.960 9.91 95.1 0.00029 0,900 23.33 255.0 0.00057 0.958 10.47 100.3 0.00030 0,898 29.01 260.5 0.00058 0.956 11.03 105.4 0.00031 0,896 29.69 266.0 0.00059 0.954 11.60 110.7 0.00032 0.894 30.37 271.5 0.00060 0.952 12.17 115.9 0.00033 0.892 31.05 277.0 0.00060 0.950 12.74 121.0 0.00034 0.890 31.75 232.6 0.00061 0.948 13.31 126.2 0.00035 0,888 32.50 238,6 0.00062 0.946 13.88 131.3 0.00036 0,886 33.25 294,6 0.00063 0.944 14.46 136.5' 0.00037 0,884 34,10 301,4 0.00064 0.942 15.04 141.7 0.00038 0.882 34,95 303.3 0.00065 * Zeit. £. angew. Chem., 1889, 181. TABLES. 519 TABLE VI.— SPECIFIC GRAVITY OF HYDROCHLORIC ACID AT 15° C, LUNQB AND MaRCHLEWSKI.* Specific Grav- 100 Parts by 1 Liter Contains ityatM! Degrees Baum^.t Degrees Twaddell. Weight Contain Grams, in Kilograms, HCl. in Vacuo. HCl. 1.000 0.0 0.16 0.0016 1.005 0.7 1 1.15 0.012 1.010 1.4 2 2.14 0.022 1.015 2.1 3 3.12 0.032 1.020 2.7 4 4.13 0.042 1.025 3.4 5 5.15 0.053 1.030 4.1 6 6.15 0.064 1.035 4.7 7 7.15 0.074 1.040 5.4 8 8.16 0.085 1.045 6.0 9 9.16 0.096 1.050 6.7 10 10.17 0.107 1.055 7.4 11 11.18 0.118 1.060 8.0 12 12.19 0.129 1.065 8.7 13 13.19 0.141 1.070* 9.4 14 14.17 0.152 1.075 10.0 15 15.16 0.163 1.020 10.6 16 16.15 0.174 1.085 11.2 17 17.13 0.186 1.090 11.9 18 18.11 0.197 1.095 12.4 19 19.06 0.209 1.100 13.0 20 20.01 0.220 1.105 13.6 21 20.97 0.232 1.110 14.2 22 21.92 0.243 1.115 14.9 23 22.86 ■ 0.255 1.120 15.4 24 23.82 0.267 1.125 16.0 25 24.78 0.278 1.130 16.5 26 25.75 0.291 1.135 17.1 27 26.70 0.303 1.140 17.7 28 27.66 0.315 1 . 1425 18.0 28.14 0.322 1.145 18.3 29 28.61 0.32S 1.150 18.8 30 29.57 0.340 1.152 19.0 29,95 0.345 1.155 19.3 si 30.55 0.353 1.160 19.8 32 31.52 0.366 1.163 20.0 32.10 0.373 1.165 20.3 33 32.49 0.379 1.170 20.9 34 33.46 0.392 1.171 21.0 33.65 0.394 1.175 21.4 35 34.42 0.404 1 180 22.0 36 35.39 0.418 1.185 22.5 37 36.31 0.430 1.190 23.0 38 37.23 0.443 1.195 23.5 39 38.16 0.456 1.200 24.0 40 39.11 0.469 * Zeit. angew. Chem., 1891, 135. i' Fut the American Standard Baura^ scale and the specific-gravity table adopted by the Manufacturing Chemists' Association of the Uijited States, see Van Nostrand s Chemical Annual. 520 TABLES. TABLE VII.— SPECIFIC GRAVITY OF NITRIC ACID AT 15 "C. Lunge akd Rey.* Specific 100 Parts by Weight 1 Liter Contains in Gravity Contain Grams, Kilograms, 15° Degrees Baumd.t Degrees TwaddeU. ^^4^ in Vacuo. N2O5. HNO3. N^Os. HNO3. 1.000 o.os 0.10 0.001 0.001 1.005 0.7 1 0.85 1.00 0.003 0.010 1.010 1.4 2 1.62 1.90 0.016 0.019 1.015 2.1 3 2.39 2.80 0.024 0.028 1.020 2.7 4 3.17 3.70 0.033 0.038 1.025 3.4 5 3.94 4.60 0.040 0.047 1.030 4.1 6 4.71 6.50 0.049 0.057 1.035 4.7 7 6.47 6.38 0.057 0.066 1.040 5.4 8 6.22 7.26 0.064 0.075 1.045 6.0 9 6.97 8.13 0.073 0.085 1.050 6.7 10 7.71 8.99 0?1 0.094 1.055 7.4 11 8.43 9.84 0.0~9 0.104 1.060 8.0 12 9.15 10.63 0.097 0.113 1.065 8.7 13 9.87 11.51 0.105 0.123 1.070 9.4 14 10.57 12.33 0.113 0.132 1.075 10.0 15 11.27 13.15 0.121 . 0.141 1.0?0 10.6 16 11.96 13.95 0.129 0.151 1.0; 5 11.2 17 12.64 14.74 0.137 0.160 1.090 11.9 18 13.31 15.63 0.145 0.169 1.095 12.4 19 13.99 16.32 0.153 0.179 1.100 13.0 20 14.67 17.11 0.161 0.188 1.105 13.6 21 15.34 17.89 0.170 0.198 1.110 14.2 22 16.00 18.67 0.177 0.207 1.115 14.9 23 16.67 19.45 0.186 0.217 1.120 ■ 15.4 24 17.34 20.23 0.195 0.227 1.125 16.0 25 18.00 21.00 0.202 0.236 1.130 16.5 26 18.66 21.77 0.211 0.246 1.135 17.1 27 19.32 22.54 0.219 0.256 1.140 17.7 28 19.93 23.31 0.228 0.266 1.145 18.3 29 20.64 24.08 0.237 0.276 1.150 18.8 30 21.29 24.84 0.245 0.286 1.155 19.3 31 21.94 25.60 0.254 0.296 1.160 19.8 32 22.60 26.36 0.262 0.306 1.165 20.3 33 23.25 27.12 0.271 0.316 1.170 20.9 34 23.90 27.88 0.279 0.326 1.175 21.4 35 24.54 28.63 0.238 0.336 I.ISO 22.0 36 25.18 29.38 0.297 0.317 1.185 22.5 37 25.83 30.13 0.306 0.357 1.190 23.0 38 26.47 30.88 0.315 0.367 1.195 23.5 39 27.10 31.62 0.324 0.378 1.200 24.0 40 27.74 32.36 0.333 0.388 1.205 24.5 41 23.36 33.09 0.342 0.399 1.210 25.0 42 23.99 33.82 0.351 0.409 1.215 25.5 43 29.61 34.55 0.360 0.420 1.220 26.0 44 30.24 35.28 0.369 0.430 1.225 26.4 45 30.88 36.03 0.378 0.441 1.230 26.9 46 31.53 36.78 0.387 0.452 1.235 27.4 47 32.17 37.53 0.397 0.463 1.240 27.9 48 32.82 38.29 0.407 0.475 1.245 2S.4 49 33.47 39.05 0.417 0.486 1.250 28.8 50 34.13 39.82 0.427 0.498 *Zeit. angew. Chem.. 1891, 165. + For the American Standard Baum^ scale and the specific-gravity table adopted by the Manufacturing Chemists' Association of the United States, see Van Nostrand's Chemical Annual. TABLES. 521 SPECIFIC GRAVITY OF NITRIC ACID AT 15° C— (Continued). Specific 100 Parts by Weight 1 Liter Contains in Gravitv Contain, Grams, Kilograms, . 15° Degrees Degrees _ at^ Eaumd. Twaddell. in Vacuo. N2O5. HNO3. N2O5. HNO3. 1.255 29.3 51 34.78 40.53 0.437 0..509 1.260 29.7 52 35.44 41.34 0.447 0.52] 1.265 30.2 53 36.09 42.10 0.457 0.533 1.270 30.6 54 36.75 42.87 0.467 0.544 1.275 31.1 55 37.41 43.64 0.477 0.556 1.280 31,5 56 38.07 44.41 0.4S7 0.568 1.285 32.0 57 38.73 45.18 0.498 0.581 1.290 32.4 5S 39.39 45.95 0.50S 0.593 1.295 32.8 59 40.05 46.72 0.519 0.605 1.300 33.3 60 40.71 47.49 0.529 0.617 1.305 33.7 61 41.37 48.26 0.540 0.630 1.310 34.2 62 42.06 49.07 0.551 0.643 1.315 34.6 63 42.76 49.89 0.562 0.656 1.320 35.0 64 43.47 50.71 0.573 0.669 1.325 35.4 65 44.17 51.53 0.5 5 0.6^3 1.330 35.8 66 44. &9 52.37 0.597 0.697 1.3325 36.0 665 45.26 52.80 0.603 0.704 1.335 36.2 67 45.62 53.22 0.609 0.710 1.340 36.6 63 46.35 54.07 0.621 0.725 1.345 37.0 69 47.03 54.93 0.633 0.739 1.350 37.4 70 47.82 55.79 0.645 0.753 1.355 37.8 71 48.57 56.66 0.65 i 0.768 1.360 38.2 72 49.35 57.57 0.671 0.783 1.365 38.6 73 50.13 53.48 0.634 0.798 1.370 39.0 74 50.91 59.39 0.69? 0.814 1.375 39.4 75 51.69 60.30 0.711 0.829 1.380 39.8 76 52.52 61.27 0.725 0.846 1.3833 40.0 53.03 61.92 0.735 0.857 1.3S5 40.1 77 33.35 62.24 0.739 0.862 1.390 40.5 78 54.20 63.23 0.753 0.879 1.395 40.8 79 55.07 64.25 0.763 0.896 1.400 41.2 80 55.97 65.30 0.783 0.914 1.405 41.6 81 56.92 66.40 0.800 0.933 1.410 42.0 82 57.86 67.50 0.816 0.952 1.415 42.3 83 53.83 63.63 0.832 0.971 1.420 42.7 84 59.83 ' 69.80 0.849 0.991 1.425 43.1 85 60.84 70.98 0.867 1.011 1.430 43.4 86 61.86 72.17 0.885 1.032 1.435 43.8 87 62.91 73.39 0.903 1.053 1.440 44.1 88 64.01 74.63 0.921 1.075 1.445 44.4 89 65.13 75.98 0.941 1.098 1.450 44.8 90 66.24 77.23 0.961 1 . 12! 1.455 45.1 91 67.38 78.60 0.931 1.144 1.460 45.4 92 63.56 79.98 1.001 1.168 1.465 45.8 93 69.79 81.42 1.023 1.193 1.470 46.1 94 71.06 82.90 1.045 1.219 1.475 46.4 95 72.39 84.45 1.063 1.246 1.480 46.8 96 73.76 86.05 1.092 1.274 1.485 47.1 97 75.18 87.70 1.116 1.302 1.490 47.4 98 76.80 89.60 1.144 1.335 1.495 47.8 99 78.52 91.60 1.174 1.369 522 TABLES. SPECIFIC GRAVITY OF NITRIC ACID AT 15° C— {Continued). Lunge and Rey. Specific 100 Parts by Weight 1 Liter Contains in Gravity Contain Grams, Kilograms, 15° Degrees Degrees '^^-^ Baum^. Twaddell. in Vacuo N2O5. HNO3. N2O5. HNO3. 1.500 48.1 100 80.65 94.09 1.210 1.411 1.501 81.09 94.60 1.217 1.420 1.502 81.50 95.03 1.224 1.428 1.503 81.91 95.55 1.231 1.436 1.504 82.29 96.00 1.233 1.444 1.505 48 i ioi 82.63 96.39 1.244 1.451 1-.506 82.94 96.76 1.249 1.457 1.507 83.26 97.13 1.255 1.464 1.50S 48 5 • • • 83.58 97.50 1.260 1.470 1.509 83.87 97.84 1.265 1.476 1.510 48 7 i62 84.09 93.10 1.270 1.481 1.511 84.28 93.32 1.274 1.486 1.512 84.46 98.53 1.277 1.490 1.513 84.63 98.73 1.250 1.494 1.514 84.78 98.90 1.283 1.497 1.515 49 6 ios 84.92 99.07 1,237 1.501 1.516 86.04 99.21 1.2?9 1.504 1.517 85.15 99.34 1,292 1.507 1.51S 85.26 99.46 1.294 1.510 1.519 85.35 99.57 1.296 1.512 1.520 49 4 i64 85.44 99.67 1.299 1.515 TABLES. 523 TABLE VIII.- -SPECIFIC GRAVITY OF SULPHURIC ACID. Lunge, Islbr, and Nasf.* :> Specific Gravity 15° 100 Parts by Weight 1 Liter Contains in Degrees Degrees Contain , Grams, Kilograms, at — r- Baum^.f Twaddell. 4° in Vacuo. SO3. H2SO4. SO3. nso,. 1.000 0.07 0.09 0.001 0.001 1.005 0.7 1 0.68 0.83 0.007 0.008 1.010 1.4 2 1.28 1.57 0.013 0.016 1.015 2.1 3 1.88 2.30 0.019 0.023 1.020 2.7 4 2.47 3.03 0.025 0.031 1.025 3.4 5 3.07 3.76 0.032 0.039 1.030 4.1 6 3.67 4.49 0.038 0.046 1.035 4.7 7 4.27 5.23 0.044 0,054 1.040 5.4 8 4.87 5.96 0.051 0.062 1.045 6.0 9 5.45 6.67 0.057 0.071 1.050 6.7 10 6.02 7.37 0.063 0.077 1.055 7.4 11 6.59 8.07 0.070 0.085 1.060 8.0 12 7.16 8.77 0.076 0.093 1.065 8.7 13 7.73 9.47 0.0 2 0.102 1.070 9.4 14 8.32 10.19 0.0-9 0.109 1.075 10.0 15 8.90 10.90 0.096 0.117 1.080 10.6 16 9.47 11.60 0.103 0.125 1.085 11.2 17 10.04 12.30 0.109 0.133 1.090 11.9 18 10.60 12.99 0.116 0.142 1.095 12.4 19 11.16 13.67 0.122 0.150 1.100 13.0 20 11.71 14.35 0.129 0.158 1.105 13.6 21 12.27 15.03 0.136 0.166 1.110 14.2 22 12.82 15.71 0.143 0.176 1.115 14.9 23 13.36 16.36 0.149 0.183 1.120 15.4 24 13.89 17.01 0.156 0.191 1.125 16.0 25 14.42 17.66 0.162 0.199 1.130 16.5 26 14.95 18.31 0.169 0.207 1.135 17.1 27 15.48' 18.96 0.176 0.215 1.140 17.7 28 16.01 19.61 0.183 0.223 1.145 18.3 29 16.54 20.26 0.189 0.231 1.150 18.8 30 17.07 20.91 0.196 0.239 1.155 19.3 31 17.59 21.55 0.203 0.248 1.160 19.8 32 18.11 22.19 0.210 0.257 1.165 20.3 33 18.64 22.83 0.217 0.266 1.170 20.9 34 19.16 23.47 0.224 0.276 1.175 21.4 35 19.69 24.12 0.231 0.2S3 1.180 22.0 36 20.21 24.76 0.238 0.292 1.185 22.5 37 20.73 25.40 0.246 0.301 1.190 23.0 38 21.26 26.04 0.253 0.310 1.195 23.5 39 21.78 26. 6S 0.260 0.319 1.200 24.0 40 22.30 27.32 . 2118 0.328 1.205 24.5 41 22.82 27.95 0.275 0.337 1.210 25.0 42 23.33 28.58 0.282 0,346 1.215 25.5 43 23.84 29.21 0.290 0,366 1.220 26.0 44 24.36 29.84 0.297 0.364 1.225 26.4 45 24.88 30.48 0.305 0.373 1.230 26.9 46 26.39 31.11 0.312 0.382 1.235 27.4 47 25.88 31.70 0.320 0.391 1.240 27.9 48 26.35 32.28 0.327 0.400 1,245 2S.4 49 26.83 32.86 0.334 0.409 1.250 28.8 50 27.29 33.43 0.341 0.418 1.255 29.3 51 27.76 34.00 0.348 0.426 1.260 29.7 52 28.22 34.57 0.366 0.43S *Zeit. angew. Chem., 1890, 131; Chem. Ind., 1883, 39. , , j , ^ v t For the American Standard Baum^ scale and the specific-gravity table adopted by the Manufacturing Chemists' Association of the United States, see Van Nostrand s Chemical Annual- 524 TABLES. SPECIFIC GRAVITY OF SULPHURIC A.CID— (Continued). Specific 100 Parts by Weight 1 Liter Contains in Gravity Degrees Baum^. Degrees Twaddell. Contain Grams, Kilog rams, at^ in Vacuo. SO3. H2SO4. SO3. HzSOi. 1.265 30.2 53 28.69 35.14 0.363 0.444 1.270 30.6 54 29.15 35.71 0.370 0.454 1.275 31.1 55 29.62 36.29 0.377 0.462 1.280 31.5 56 30.10 36.87 0.385 0.472 1.285 32.0 57 30.57 37.45 0.393 0.481 1.290 32.4 58 31.04 38.03 0.400 0.490 1.295 32.8 59 31.52 38.61 0.403 0.500 1.300 33.3 60 31.99 39.19 0.416 0.510 1.305 33.7 61 32.46 39.77 0.424 0.519 1.310 34.2 62 32.94 40.35 0.432 0,529 1.315 34.6 63 33.41 40.93 0.439 0,538 1.320 35.0 64 33.88 41.50 0.447 0,548 1.325 35.4 65 34.35 42.08 0.455 0.557 1.330 35.8 66 34. SO 42.66 0.462 0.567 1.335 36.2 67 35.27 43.20 0.471 0.577 1.340 36.6 68 35.71 43,74 0.479 0.586 1.345 37.0 69 36.14 44.28 0.486 0.596 1.350 37.4 70 36.58 44.82 0.494 0,605 1.355 37.8 71 37.02 45.35 0.502 0,614 1.360 38.2 72 37.45 45.88 0.509 0.624 1.365 38.6 73 87.89 46.41 0.517 0.633 1.370 39.0 74 38.32 46.94 0.525 0.643 1.375 39.4 75 38.75 47.47 0.533 0.653 1.380 39.8 76 39.18 48.00 0.541 0.662 1.385 40.1 77 39.62 48.53 0.549 0.672 1.390 40.5 78 40.05 49.06 0.557 0.682 1.395 40.8 79 40.48 49.59 0.564 0.692 1.400 41.2 80 40.91 50.11 0.573 0.702 1.405 41.6 81 41.33 50.63 0.581 0.711 1.410 42.0 82 41.76 51.15 0.589 0.721 1.415 42.3 83 42.17 51.66 0.697 0.730 1.420 42.7 84 42.57 52.15 0.604 0.740 1.425 43.1 85 42.96 52.63 0.612 0.750 1.430 43.4 86 43.36 53.11 0.620 0.759 1.435 43.8 87 43.75 53.59 0,628 0,769 1.440 44.1 88 44.14 54.07 0.636 0.779 1.445 44.4 89 44.53 54.55 0.643 0.789 1.450 44.8 90 44.92 55.03 0.651 0.798 1.455 45.1 91 45.31 55.50 0.659 0.808 1.460 45.4 92 45.69 55.97 0.667 0.817 1.465 45.8 93 46.07 56.43 0.675 0.827 1.470 46.1 94 46.45 56.90 0.6S3 0.837 1.475 46.4 95 46.83 57.37 0.691 0.846 1.480 46.8 96 47.21 57.83 0.699 0.866 1.485 47.1 97 47.57 58.28 0.707 0.866 1.490 47.4 98 47.95 58.74 0.715 0.876 1.495 47.8 99 48.34 59.22 0,723 0.885 1.500 48.1 100 48.73 59.70 0.731 0.896 1.505 48.4 101 49.12 60.18 0.739 0.906 1.510 48.7 102 49.51 60.65 0.748 0.916 1.515 49.0 103 49.89 61.12 0.756 0.926 1.520 49.4 104 50.28 61.59 0.764 0.936 1.525 49.7 105 50.66 62.06 0.773 0,946 1 530 50.0 106 51.04 62.53 0.781 0.957 TABLES. 525 SPECIFIC GRAVITY OF SULPHURIC ACID— (Continued). Specific Gravity . 15° Degrees Baumd. Degrees Twaddell. 100 Parts by Weiglit Contain, Grams, 1 Liter Contains in Kilograms, at^^ in Vacuo. SO3. H2SO4. SO3. H2SO,. 1.535 50.3 107 51.43 63,00 0,789 0,967 1.540 50.6 108 51.78 63.43 0,797 0,977 1.545 50.9 109 52.12 63.!~5 0,£05 0,987 1.550 51.2 110 52.46 64.26 0,813 0,996 1.555 51.5 111 52.79 64.67 0.821 1.0U6 1 . 560 51.8 112 53.12 65.03 0.829 1,015 1 . 565 52.1 113 53.46 65,49 0.837 1,025 1.570 52.4 114 53.80 65,90 0,845 1,035 1.575 52.7 115 54.13 66,30 0,853 1,044 1.580 53.0 116 54.46 66.71 0,861 1,054 1.5S5 53.3 117 54. JO 67.13 0,869 1.064 1.590 53.6 118 55.18 67,59 0,877 1.075 1.595 53.9 119 55.55 63,05 0,886 1.085 1.600 54.1 120 55.93 63,51 0,895 1,096 1.605 54.4 121 56.30 63.97 0,904 1,107 1.610 54.7 122 56.6! 69.43 0,913 1,118 1.615 55.0 123 57.05 69.89 0,921 1,128 1.620 55.2 124 57.40 70.32 0,930 1,139 1.625 55.5 125 57.75 70.74 0,938 1,150 1.630 55.8 126 53.09 71.16 0.947 1,160 1.635 56.0 127 53.43 71.57 0,955 1,170 1,640 56.3 128 58.77 71.99 0,964 1,181 1.645 56.6 129 59.10 72.40 0,972 1,192 1.650 56.9 130 59.45 72.82 0,981 1,202 1,655 67.1 131 59.75 73.23 0.989 1,212 1.660 57.4 132 60.11 73.64 0.998 1,222 1.665 57.7 133 60.46 74.07 1,007 1,233 1.670 57.9 134 60.82 74.51 1,016 1,244 1.675 58.2 135 61.20 74.97 a, 025 1,256 1.6S0 58.4 136 61.57 75,42 1,034 1,267 1.6?5 58.7 137 61.93 75.86 1,043 1,278 1.690 58.9 138 62.29 76,30 1,053 1,289 1.695 59.2 139 62.64 76,73 1,062 1,301 1.700 59.5 140 63.00 77,17 1,071 1,312 1.705 59.7 141 63.35 77,60 1,0.0 1,323 1.710 60.0 142 63.70 78,04 1,0 9 1,334 1.715 60.2 143 64.07 78,48 1,099 1,346 1.720 60.4 144 64.43 78,92 1,103 1,357 1.725 60.6 145 64.78 79,36 1,118 1,369 1.730 60.9 146 65.14 79,80 1,127 1,381 1.735 61.1 147 65.50 SO, 24 1,136 1,392 1.740 61.4 14S 65.86 80,68 1,146 1,404 1.745 61.6 149 66.22 81.12 1,156 1,416 1.750 61.8 150 66.58 81,56 1,165 1,427 1.755 62.1 151 66.94 82,00 1,175 1,439 1.760 62.3 152 67.30 82,44 1,1E5 1.451 1.765 62,5 153 67.65 82.88 1,194 1,463 1.770 62.8 154 68.02 83.32 1,204 1,475 1.775 63.0 155 68.49 83.90 1,216 1,489 1.780 63.2 156 68.98 84.50 1 , 228 1.504 1 . 7.^5 63.5 157 69.47 85,10 1,240 1,519 1.790 63.7 15-! 69.96 85,70 1,252 1.534 1.795 64.0 159 70.45 86,30 1,265 1,549 1.800 64.2 160 70,94 86,90 1,277 1.564 526 TABLES. SPECIFIC GRAVITY OF SULPHURIC ACID— {Continued) . Specific 100 Parts by Weight 1 Liter Contains in Gravity Degrees Degrees Contain Grams, Kilograms, at — - TlmTm ti TwaddelL at 40 in Vacuo. J->aiUllHi. SO3. H2SO4. SO3. H2SO4. 1.805 64.4 161 71.50 87.60 1.291 1.581 1.810 64.6 162 72.08 88.30 1.305 1.598 1.815 64.8 163 72.69 £9.05 1.319 1.621 1.820 65.0 164 73.51 90.05 1.338 1.639 1.821 73.63 90.20 1.341 1.643 1.822 65! i 73.80 90.40 1.345 1.647 1.823 73.96 90.60 1.348 1.651 1.824 65! 2 74.12 90.80 1.352 1.656 1.825 ies 74.29 91.00 1.356 1.661 1.826 65^3 74.49 91.25 1.360 1.666 1.827 74.69 91.50 1.364 1.671 1.828 65]4 ... 74.86 91.70 1.368 1.676 1.829 75.03 91.90 1.372 1.681 1.830 iee 75.19 92.10 1.376 1.685 1.831 65.5 75.35 92.30 1.380 1.690 1.832 75.53 92.52 1.384 1.695 1.833 65!6 75.72 92.75 1.388 1.700 1.834 75.96 93.05 1.393 1.706 1.835 65!7 i67 76.27 93.43 1.400 1.713 1.836 76.57 •93.80 1.406 1.722 1.837 76.90 94.20 1.412 1.730 1.838 65^8 77.23 94.60 1.419 1.739 1.839 77.55 95.00 1.426 1.748 1.840 65^9 ies 78.04 95.60 1.436 1.759 1.8405 78.33 95.95 1.451 1.765 1.8410 79.19 97,00 1.45S 1.786 1.8415 79.76 97.70 1.469 1.799 1.8410 .... 80.16 98.20 1.476 1.808 1.8405 80.57 98.70 1.453 1.816 1.8400 80. 9S 99.20 1.490 1.825 1.8395 81.18 99.45 1.494 1.830 1.8390 81.39 99.70 1.497 1.834 1.8385 81.59 99.95 ■ 1.500 1.838 TABLE IX.— VAPOR TENSION OF WATER IN MILLIMETERS OF MERCURY. Magnus. Temp. Millimeters. Temp. Millimeters. Temp. Millimeters. Temp. Millimeters. 4.525 8 7.964 15 12.677 23 20.909 + 1 4.867 9 8.525 16 13.519 24 22.211 2 5.231 10 9.126 17 14.409 25 23.5«2 3 5.619 11 9.756 18 15.351 26 25.026 4 6.032 12 10.421 19 16.345 27 26.547 5 6.471 13 11.130 20 17.396 28 2S.148 6 6.939 14 11.882 21 18.505 29 29.532 7 7.436 22 19 . 675 30 31 . 602 TABLES. 527 TABLE X.— DENSITY OF WATER. ROSSETTI. Tempera- Density (Density Volume (Volume Density (Density Volu.*ne (Volume ture. atO° = l). at 0° = 1). at 4° = 1). at-l» = l). -10 0.998274 1.001729 0.998145 1.001858 - 9 0.998556 1.001449 0.998427 1.001575 - 8 0.998814 1.001191 0.998685 1.001317 - 7 0.999040 1.000963 0.998911 1.001089 - 6 0.999247 1.000756 0.999118 1.000883 - 5 0.999428 1.000573 0.999298 1.000702 - 4 0.999584 1.000416 0.999455 1.000545 - 3 0.999719 1.000281 0.999590 1.000410 - 2 0.999832 1.000168 0.999703 1.000297 - 1 0.999926 1.000074 0.999797 1.000203 1.000000 1.000000 0.999871 1.000129 + 1 1.000057 0.999943 0.999928 1.000072 2 1.000093 0.999902 0.999969 1.000031 3 1.000120 0.999880 0.999991 1.000009 4 1.000129 0.999871 1.000000 1.000000 5 1.000119 0.999881 0.999990 1.000010 6 1.000099 0.999901 0.999970 1.000030 7 1.000062 0.999938 0.999933 1.000067 8 1.000015 0.999985 0.999886 1.000114 9 0.999953 1.000047 0.999824 1.000176 10 0.999876 1.000124 0.999747 1.000253 11 0.999784 1.000216 0.999655 1.000345 12 0.999678 1.000322 0.999549 1.000451 13 0.999559 1.000441 0.999430 1.000570 14 0.999429 1.000572 0.999299 1.000701 15 0.999289 1.000712 0.999160 1.000841 16 0.999131 1.000870 0.999002 1.000999 17 0.998970 1.001031 0.998841 1.001160 18 0.998782 1.001219 0.998654 1.001348 19 0.998588 1.001413 0.998460 1.001542 20 0.998388 1.001615 0.998259 1.001744 21 0.998176 1.001828 0.998047 1.001957 22 0.997956 1.002048. 0.997828 1.002177 23 0.997730 1.002276 0.997601 1.002405 24 0.997495 1.002511 0.997367 1.002641 25 0.997249 1.002759 0.997120 1.002888 26 0.996994 1.003014 0.996866 1.003144 27 0.996732 1.003278 0.996603 1.003408 28 0.996460 1.003553 0.996331 1.003682 29 0.996179 1.003835 0.996051 1.003965 30 0.99589 1.00412 0.99577 1.00425 31 0.99560 1.00442 0.99547 1.00455 32 0.99530 1.00473 0.99517 1.00486 33 0.99498 1.00505 0.99485 1.00518 34 0,99465 1.00538 0.99452 1.00551 35 . 0.99431 1,00572 0.99418 1.00586 36 0.99396 1.00603 0.99383 1.00621 37 0.99360 1.00645 0.99347 1.00657 38 0.99323 1.00682 0.99310 1.00694 39 0.99286 1.00719 0.99273 1.00732 40 0.99248 1.00757 0.99235 1.00770 1.00809 41 0.99210 1.00796 0.99197 42 0.99171 1.00,36 0.99158 1.00849 528 TABLES. DENSITY OF WATER— (Continued). Tempera- Density (Density Volume (Volume Density (Density Volume (Volume ture. atO° = l). atO° = l). at 4"' = 1). at 4° = !). 43 0,99131 1.00876 0.99118 1.00889 44 0.99091 1.00917 0.99078 1.00929 45 0.99050 1.00953 0.99037 1.00971 46 0.99009 1.01001 0.9S996 1.01014 47 0.93967 1.01044 0.9S954 1.01057 48 0.98923 1.01088 0.9S910 1.01101 49 0.98878 1.01134 0.98865 1.01148 50 0.98832 1.01182 0.98819 1.01195 51 0.98785 1.01230 0.98772 1 . 01243 52 0.98737 1.01279 0.98725 1.01292 53 0.98689 1.01328 98677 1.01341 54 0.98642 1.01377 98629 1.01390 55 0.98594 1.01426 98531 1.01439 56 0.98547 1.01475 0.9^534 1.01488 57 0.98499 1.01524 0.93486 1.01537 53 0.98450 1.01574 0.',)8437 1.01587 59 0.98401 1.01625 0.93388 1.01638 60 0.98350 1.01678 0.98338 1.01691 61 0.98299 1.01731 0.93236 1.01744 62 0.98247 1.01785 0.98234 1.01798 63 0.98194 1.01839 0.98182 1.01852 64 0.98140 1.01895 0.93123 1.01908 65 0.9S0S6 1.01951 0.98074 1.01964 66 0.9S032 1.02008 0.93019 1.02021 67 0.97977 1.02065 0.979>')4 1.02078 68 0.97921 1.02124 0.97903 1.02137 69 0.97864 1.02183 0.97851 1.02196 70 0.97807 1.02243 0.97794 1.02256 71 0.97749 1.02303 0.97736 1.02316 72 0.97690 1.02365 0.97677 1.02378 73 0.97631 1.02427 0.97618 1.02440 74 0.97571 1.02490 0.97553 1.02503 75 0.97511 1.02553 0.97493 1.02566 76 0.97450 1.02617 0.97438 1.02630 77 0.973S9 1.02681 0.97377 1.02694 78 0.9732S 1.02745 0.97316 1.02758 79 0.97267 1.02809 0.97255 1.02822 80 0.97206 1.02874 0.97194 1.02887 81 0.97145 1.02939 0.97132 1.02952 82 0.97033 1.03005 0.97070 1.03018 83 0.97020 1.03072 0.97007 1.03085 84 0.96956 1.03139 0.96943 1.03153 85 0.96S92 1.03207 0.96379 1.03221 86 0.96828 1.03276 0.96315 1.03289 87 0.96764 1.03345 0.96751 1.03358 88 0.96699 1.03414 0.96637 1.03427 89 0.96634 1.03484 0.96622 1.03497 90 0.96563 1.03554 0.96556 1.03567 91 0.96502 1.03625 0.96490 1.03638 92 0.96435 1.03697 0.96423 1.03710 93 . 0.96368 1.03770 0.96356 1 03782 TABLES. DENSITY OP WATER— (Coniinwed). 529 Tempera- Density (Density Volume (Volume Density (Density Volume (Volume ture. aiO^-l). at 0° = 1). at4» = l). at 4'' = 1), 94 0.96300 1.03844 0.9628& 1.03S56 95 0.96231 1.03918 0.96219 1.03931 96 0.96161 1.03993 0.96149 1.04006 97 0.96091 1.04069 0.96079 1.04082 9S 0.96020 1.04145 0.96008 1.04158 99 0.95949 1.04222 0.95937 1.04235 100 0.95579 1.04299 0.95866 1.04312 TABLE XL— SOLUBILITY OF OXYGEN IN FRESH WATER AND IN SEA WATER OF STATED DEGREES OF SALINITY AT VARIOUS TEMPERATURES WHEN EXPOSED TO AN ATMOSPHERE CONTAINING 20.9 PER CENT OF OXYGEN AND UNDER A PRESSURE OF 760 MM.* (Calculated by Geo. C. Whipple and M. C. Whipple from Measurements of C. J. Fox) Chlorine in Sea Water (Parts per Million) . Difference per 100 Parts 5000 1000 15000 20000 of Chlorine per Million. 14.62 14.23 13.84 13.48 13.13 12.80 12.48 12.17 11.87 11.69 11.33 11.08 10.83 10.60 10.37 10.15 9.95 9.74 9.54 9.35 9.17 8.99 8.83 8.68 8.53 8.38 8.22 8.07 7.92 7.77 7.63 13.79 13.41 13.05 12.72 12.41 12.09 11.79 11.51 11.24 10.97 10.73 10,49 10.28 10.05 9.85 9.65 9.46 9.26 9.07 8.89 8.73 8.57 8.42 8.27 8.12 7.96 7.81 7.67 7.53 7.39 7.25 Milligram 12.97 12.61 12.28 11.98 11.69 11,39 11.12 10.85 10.61 10.36 10.13 9.92 9.72 9.52 9.32 9.14 8.96 8.78 8.62 8.45 8.30 8.14 7.99 7.85 7.71 7.56 7.42 7.28 7.14 7.00 6.86 i per Liter 12.14 11.82 11.52 11.24 10.97 10.70 10.45 10.21 9.98 9.76 9.55 9.35 9.17 8.98 8.80 8.63 8.47 8.30 8.15 8.00 7.86 7.71 7.57 7.43 7.30 7.15 7.02 6.88 6.75 6.62 6.49 11.32 11.03 10.76 10.50 10.25 10.01 9.78 9.57 9.36 9.17 8.98 8.80 8.62 8.46 8.30 8.14 7.99 7.84 7.70 7.56 7.42 7.28 7.14 7.00 6.87 6.74 6.61 6.49 6.37 6.25 6.13 .0165 1 .0160 2 .0154 3 .0149 4 .0144 5 .0140 6 .0135 7 .0130 8 .0125 9 .0121 10 .0118 11 .0114 12 .0110 13 .0107 14 .0104 15 .0100 16 .0098 17 .0095 18 .0092 19 .0089 20 21 .0088 .0086 22 .0085 23 .0083 24 .0083 25 .0082 26 .0080 27 .0079 28 .0078 29 .0076 30 ,0075 * Under any other barometric pressure, B, the solubility can be obtained from the corre- sponding value in the table by the formula: B _„ B' S' = solubility at 5mm. 5' inches. ^"760" 29.92 S =Bolubility at 760 mm. 29.92 inchea. INDEX. Accelerators, use of, in Parr calorim- eter, 401. Accuracy, limit of, 4, 122. Acetic acid, reagent, preparation of, 506. Acetyl value of fats and oils, 447. determination of, 447. distillation process for, 448. filtration process for, 448. table of values of, 461. Acid, best for general use, 255. boric, titration of, 293. carbonic. See carbon dioxide, chromic, titration of, 334. free, in fats and oils, determina- tion of, 438. hydrochloric. See hydrochloric acid, oxalic. See oxalic acid, phosphoric. See phosphoric acid, standard, most desirable strength for general use, 256. sulphuric. See sulphuric acid, sulphurous, oxidation of, 306. Acid radicles, secondary, electrolytic reactions of, 200. Acids, determination of, 94. normal, 247. standard, 258. dilution of, to exact strength, 262, 265. standardization of, by anhydrous oxalic acid, 260. crystallized oxalic acid, 2596. evaporation with ammonia, 261, 263. gravimetric methods, 261 , 265, 272. potassium dichromate, 260. " tetroxalate,260. sodium carbonate, 258, 264. specific gravity, 261, 271. titration of, by iodometric meth- ods, 343. Acids, weak, titration of, 255. titration of base in salts of, 253. Acidimetry, 247. calculation of titrations in, 500. Acidity of fats and oils, 437. determination of, 438. of nutrient gelatin, adjusting of, Ex. 72, 429. Air, collection of samples of, 302, 305. determination of carbon dioxide in, 301, Ex. 55, 304. calculation of results of, 305. displacement in weighing, correction for, 11, 234. effect of, on weighing, 11. drying of, by calcium chiloride, 45. exclusion of, from ferrous solu- tions, 312. leaks, testing for, 184. Albuminoid ammonia, determination of, 411, 416. Alcohol, absolute, absorption of ben- zine vapor by, 467. methyl, distillation of boric acid with, 116. Alkali, caustic, action on glass by, 53, 60. precipitation of copper, man- ganese, nickel, and cobalt by, 60. cyanides, analysis of, by iodomet- ric methods, 342. Alkalies, action on glass by, 60. contamination of precipitates with, 60. determination of, in silicates, 190, 195. in feldspar, 195. in water, 435. separation of, irom each other, 187. from magnesium, 186. from zinc, 145. standard, 279. titration of, 265. 531 532 INDEX. Alkalimetric method of titrating phos- phoric acid, 360. Alkalimetry, calculation of titrations in, 500. Alkaline earth, carbonates, purification of, 30, 33. metals, separation from zinc, 145. Alkaline solutions, method of keeping, 289. standard, not permanent, 279. Alkalinity of water, 268. determination of, 269. Alloys, analysis of, 121. brass or bronze, 146. Britannia metal, 142. determination of silver in, 355. German silver, 148. manganes e-phosphorus-bronze, 149. Rose's metal, 129. silver coin, 123. soft solder, 128. taking samples of, for analysis, 153. type-metal, 140. Aluminium, determination of, in chromite, 164. in cinnabar, 175o. in dolomite, 181. in feldspar, 194. in potash alum, Ex. 10, 54. in salts of volatile and organic acids, 71. in silicates, 189. in smaltite, 170. in water, 435. electrolytic separation from lead, 222. hydroxide, precipitation of, 51. effect of organic matter on, 54. washing and ignition of, 52, 54. oxide, determination of silica in, 53. pure compounds of, 31 . separation, as basic acetate from manganese, cobalt, nickel, and zinc, 157. as basic carbonate from man- ■ ganese, cobalt, nickel, and zinc, 159. as hydroxide, from cobalt, nicke), and zinc, 156. from chromium, 159, 164. from iron, 160, 162. from zinc, 145. vohimetric from iron, 162. weighing of, as phosphate, 85. Ammeter, 205. Ammonia, albuminoid, determination of, in water, 411, 416. Ammonia, determination of, as am- monium chloride, 105. as platino-ctJoride, 106. in ammonium cUoride, Ex. 52, 282. free, determination of, in water, 411, 413. free water, preparation of, 412. reagent, preparation of, 506. solution, standard, 412. specific gravity of, table of, 518. standard solutions of, 282. not permanent, 279. titration of, 287. Ammonium, acetate, reagent, 506. carbonate, precipitation of metals by, 64. reagent, 506. solution of arsenic sulphide in, 134. chloride, determination of ammo- nia as, 105. pure, preparation of, 32. reagent, 506. hydroxide, precipitation of metals by, 51. reagent, 506. oxalate, precipitation of calcium by, 66, 67. reagent, preparation of, 507. testing for purity, 32. platino-chloride, determination of ammonia as, 106. salts, removal of, 86. sulphate, pure, preparation of, 33. sulphocyanate, standard solution of, 355. Analysis, limit of accuracy in, 122. taking of sample for, 152. technical, 364. Analytical equations, balancing of, 496. Anions, definition of, 197. Anode or cathode, rotation of, 209, 217, 221, 225. Anodes, platinum, 205. Anthracite coal, definition of, 391, 395. Antilogarithms, table of, 516. Antimony, determination of, as sul- phide, 78, 80. by lodometric metliods, 343. in bronze or brass, 147. in pyrites, 174. in smaltite, 169. in type-metal, 141. in Wood's metal, 131. electrolytic separation, from ar- senic and tin, 137. from lead, 22^. oxide, analysis of, by iodometrio methods, 342. pure compounds of, 31. INDEX. 533 Antimony, separation, from arsenic and tin, 134. by H. Rose'smethiod, 136. from copper, 214. from lead, 127. from tin by F. W. Clarke's method, 134, 144. by reduction witii iron, 136, 141. from zinc, 145. Apparatus, calibration of, 226. volumetric, 227. Arms of balance, determination of rela- tive length of, Ex. 2, 15. Arsenic, compounds, solution of, 139. determination, as magnesium py- roarsenate, 89. as sulphide, 78, 81. in arsenious oxide, Ex. 21, 81. in brass or bronze, 147. in pyrites, 174. in smaltite, 169. in soft solder, 128. in type-metal, 142. in Wood's metal, 131. electrolytic separation from anti- mony, 137. pure compounds of, 31. separation, as arsenious chloride, 138. from antimony by H. Rose's method, 136. from antimony and tin, 134. from copper, 214. from tin by F. W. Clarke's method, 134, 144. Arsenic acid, separation of, from phos- phoric acid, 118. Arsenious chloride, distillation of, 139. oxide, analysis of, by iodometric methods, 342. determination of arsenic in, Ex. 21, 81. oxidation of, 306. percentage of arsenic in, 81 . pure, testing of, 33. standardization of iodine solutions with, 339, 341. Arsenites, oxidation of, 306. Arsenopyrite, analysis of, Ex. 36, 172. Ash, determination of, in coal, 391, 393. Atomic weights, limit of accuracy in, 486. international table of, 509. limit of accuracy in, 486. Available oxygen, 307. determination, in manganese ores, 322, Ex, 68, 325. in potassium dichromate, 307. Available oxygen, determination, in potassium permanganate, 307. in pyrolusite, Ex. 58, 325. B. coli, in water, test for, 427. present in intestines, 407. test for, in water, Ex. 73, 431. Bacteria present in water, 427. Bacterial count, in water, 428, Ex. 72, 430. preparation of, nutrient agar- agar for, Ex. 72, 430. nutrient gelatin for, Ex. 72, 428. Bacteriological examination of water, 427. Balance, centre of gravity of, 9. construction of, 7. determinati.m of relative length cf arms, Ex. .2, 15. maximum load of, 15. parts of, 7. rules for using, 20. sensibility of, 9. determination of, Ex. 1, 15. testing for equality of arms, 13. Balancing of equations, 496. Barium, carbonate, ignition of, 66, 71. pure, preparation of, 33. weighing of, 66, 71. chloride, determination of barium in, Ex. 18, 76. of water in, Ex. 8, 44. percentage of bariam in, 76. pure, testing of, 33. reagent, preparation of, 507. determination, cs chrom te, 90. in barium chlor'de, Ex. 18, 76. in salts of organic acids, 71 . in salts of volatile acids, 76. hydroxide, standard solutions of, 281. titration of, with oxalic acid, 303, 305. use of, in determining carbon dioxide in gases, 301, 303, 304. precipitation, as carbonate, 64. as sulphate, 73, 75. pure compounds of, 31. separation as chromate from cal- cium and strontium, 178. as sulphate from magnesium and the alkalies, 177. from calcium as nitrate, 180. from calcium and strontium, 176. from lead, 127. from the alkali metals, 176. sulphate, determination of sulphur in, 103. dissociation of, 21. 534 INDEX. Barium, sulphate, ignition of, 74. properties of, 73. thiosulphate, standardization of iodine solutions with, 338, 340. weighing as chromate, 179. Bases present in fats and oils, 436. Basicity of acids with indicators, 251 . Baum^ scale, 276. .American Standard, 277. Beam of balance, 7. Benzene vapor, absorption by absolute alcohol, 467. Berthelot, bomb calorimeter of, 396. Bishop and Allen, method of precipi- tating barium sulphate, 74, 173. Bismutb, determination, as oxide, 130; as oxychloride,93, 127; as sulphide, 78,80; in Britannia metal, 143; in Rose's metal, 130; in salts of volatile and organic acids, 71; in Wood's metal, 131. oxide, pure, preparation of, 33. oxychloride, properties of, 93. pure compounds of, 31. separation, from copper, 214. from lead as chloride, 126. from zinc, 145. Bituminous coal, definition of, 391 , 395 . Bleaching powder, active substance in, 346. determination of available chlo- rine in, Ex. 67, 347. iodometric titration of, 347. sampling of, 347. titration with sodium arsenite, 347. Bolting-cloth, use of, in sifting, 155. Bomb calorimeter of Berthelot, 396. Borates, solution of, 116. Borax, anhydrous, fusion of carbonates with. 111. Boric acid, apparatus for determina- tion of, 115. decomposition of organic matter containing, 295. determination of, by the method of Gooch, 115. distillation of, with methyl alcohol, 116. titration of, 293. by Jones' mannitol meth- od, 294. by Thompson's glycerine method, 293. weighing of, 116. oxide, decomposition of silicates by, 190. purification of, 191. volatilization of, 191. Brass, analysis of, Ex. 31, 146. Britannia metal, analysis of, Ex. 30, 142. Britannia metal, decomposition of, by means of chlorine, 142. British Thermal Unit, definition of, 396. Bromates, analysis of, by iodometric methods, 343. Bromine, absorption of hydrogen sulphide by, 369. of "illuminants" by, 466. determination, as silver bromide, 95. in metallic salts, 95. distillation of, 99. free, determination of, by iodo- metric methods, 343. separation from chlorine and io- dine, 97. from hydrocyanic acid, 100. from iodine as palladous io- dide, 96. Bronze, analysis of, Ex. 31, 146. manganese-phosphorus, analysis of, Ex. 33, 149. B. typhus in water, 427. Bunsen, conditions for explosion of hydro- gen, 468. distilling apparatus of, 344. filter-pump, 24. method of producing pure oxygen, 468. valve, 311. Burettes, calibration of, Ex. 47, 244. by Morse-Blalock bulbs, 238. curve of corrections for. Fig. 41, 243. errors in reading, 229. Hempel gas, 472. irregularities of bore, 241 reading of, 228. total capacity of, 240. use of floats with, 229. Winkler modified gas, 472. Butter fat, table of Polenske values of, 459. Cadmium, absorption of hydrogen sulphide by alkahne solution of, 369, 372. carbonate, properties of, 68. ignition of, 69. determination, as sulphide, 79. in Wood's metal, 181. electrolytic separation, from cop- per, 214. from lead, 222. iodide, pure, testing of, 33. oxide, ignition of, 69. pure compounds of, 31 . separation, from copper, 130, 131. from zinc, 145. INDEX. 535 Cadmium, precipitation by sodium carbonate, 68. Calcium, carbonate, 64. determination of calcium in, Ex. 15, 67. ignition of, 65. percentage of calcium oxide in, 68. precipitation of carbon diox- ide as, 299. pure, preparation of, 33. weighing of, 65. chloride, fused, drying properties of, 44, 45. effect of temperature on dry- ing properties of, 45. determination, in calcium carbon- ate. Ex, 15, 67. in cinnabar, 175a. in dolomite, 181. in salts of organic acids, 71. in salts of volatile acids, 76. in water, 435. oxalate, conversion into carbonate and oxide, 67, 68. precipitation and digestion of, 66, 67. titration of, 328. oxide, weighing of boric acid with, 117. phosphate, standardization with, 359.. precipitation, as carbonate, 64, 195. by ammonium oxalate, 66 , 67 . pure compounds of, 31. separation, as nitrate from barium, 180. as nitrate from strontium, 179. as oxalate from the alkalies, magnesium, barium, and strontium, 176, 181. as sulphate from the alkalies and magnesium, 177. from barium as chromate, 178. from bariimi and strontium, 176. from the alkali metals, 176. volumetric determination of, 328. weighing, as carbonate, 65, 68. as oxide, 65, 68, 71. Calculation, by logarithms, 487. of acidimetry and alkalimetry, 500. of calibration of weights, 17. of determinations of iron as ferric oxide, 59, 60. of iodometric titrations, 504. of oxidation and reduction titra- tions, 502. of results of analysis of illuminat- ing-gas, 481. Calculation, of titrations with acids, 248. of volumetric determinations, 499. of water analysis, 4356. Calibration, 233. of burettes, Ex. 47, 244. by means of Morse-Blalock bulbs, 235. by weighing water, 234. of flasks, Ex. 46, 244. by the Morse-Blalock bulbs, 238. by weighing water, 233. temperature of water used in, 237. of Morse-Blalock bulbs, 236. of volumetric apparatus, 226. of weights, Ex. 3, 16. calculation of results of, 17. Calorie, definition of, 395. Calorific value, of coal, determination of, Ex.71, 403. of fuels, determination of, 395. effect of composition on, 395. Calorimeter, Parr, 397. Calorimeters, description of, 396. Cann el-coal, definition of, 391, 395. Capsule, Hofmeister, use of, 290. Carbon, combined, determination of, 383. combustion of, in a stream of oxy- gen, 377. with chromic acid, 380. condition in which present in iron and steel, 364. determination by the colorimetric method of Eggertz, 383. in chrome steel or iron liigh in silica, 381. in iron and steel, 374. direct weighing of, 377. filtration of, on a platinum boat, 377. on a carbon filter, 378. fixed, determination of, in coal, 393. graphitic, determination of, 382. oxidation of, by chromic acid, 374. apparatus for, 374. separation from iron by copper solution, 376. total, determination of, 374. Carbon dioxide, absorption, by caustic potash solu- tion, 464. by soda-lime, 114. in caustic potaSh, 114, 183. determination of, 109. by fusion of carbonates with borax, 111, 182. by fusion of carbonates with microcosmic salt. 111. 536 INDEX. Carbon dioxide, determination of, by loss, by laboratory apparatus, 110. by the Schrotter appa- ratus, 109. by the apparatus of Fre- senius, 112, 183. in air, 301, Ex. 55, 304. calculation of results of, 304. in bottled waters, 300. in dolomite, 182. in gases, 300. by Pettenkofer's method, 301. in illuminating-gas, 478. in water, 296. by Pettenkofer's method, 298. free, 296. semi-combined, 297. total, 298. direct weighing of, 112, 183. in dolomite, 183. drying and absorption of, 114. effect on indicators, 250. precipitation of, 266, 268. as calcium carbonate, 299. purification and absorption of , 378. titration of, 265, 267, 293. Carbon di-sulphide, extraction of sul- phur by, 79. Carbon monoxide, absorbents for, 464. Carbonates, determination of carbon dioxide in, by fusion with borax or microcosmic salt. 111 . Cathions, definition of, 197. Cathode, rotation of. 209, 217, 221, 225. Cathodes, platinum dishes as, 205. roughening of, 205. Caustic potash, bulbs, Geissler, use of, 114, 183. standard alcoholic solution of, 281, 437. See potassium hydroxide. Caustic soda. See sodium hydroxide. Cells, primary and storage, 203. Centre of gravity of balance, 9. Centrifugal machine, separation of lead peroxide by, 389. Chalcopyrite, analysis of, Ex. 26, 172. Characteristic of logarithms, method of finding, 363. Chemical examination of water, 410. Chlorates, analysis of, by iodometric methods, 343. Chloric acid, dejierminationof, 100, 343. Chlorides, determination of, in water, 424. titration of, with silver nitrate, 350, 351. Chlorine, absorption of, by silver-foil, 379. Chlorine, available, determination of, in bleaching powder, 346, Ex. 67, 347. decomposition of Britannia metal by, 142. determination, as silver chloride, 94, 265. in metallic salts, 95. free, determination of, by iodo- metric methods, 343. normal in water, 432. oxidation of sulphur compounds with, 104. separation from bromine and io- dine, 97. from hydrocyanic acid, 100. from iodine as palladous io- dide, 96. from iodine as thallous iodide, 97. Cholesterol and phytosterol, detection of in tats and oils, 151. melting-points of, 452. Chromate of silver as indicator, 350. Chromates, analysis of, by iodometric methods, 343. Chrome-iron ore, decomposition of, with sodium peroxide, 333. Chromic acid, oxidation of carbon by, 374, 380. titration of, 334. Chromite, analysis of, Ex. 34, 163. Chromiimi, chromic, precipitation of, as hydroxide, 51. determination, as barium or lead chromate, 90. in iron ores, 333, Ex. 63, 333. in chromite, 163. in salts of volatile and organic acids, 71 in smaltite, 170. hydroxide, effect of organic matter on precipitation of, 54. washing and ignition of, 52. oxidation with chlorine or bro- mine, 160. with potassium nitrate, 159. with sodium peroxide, 160, 163. oxide, determination of silica in, 53. pure compounds of, 31. separation, as hydroxide from cobalt, nickel, and zinc, 156. from aluminium, 160, 164. from iron, 160, 163. from iron, aluminium, man- ganese, cobalt, nickel, and zinc, 159. from zinc, 145. Chromous chloride, absorption of oxy- gen by, 464. INDEX. 537 Cinnabar, analysis of, Ex. 37, 175. Clarke's method of separation of ar- senic and antimony from tin, 134. Classen's method of determining, zinc, 224. copper, 215. Cleaning mixture for volumetric appa- ratus; 240. Coal, anthracite, definition of, 391 . bituminous, definition of, 391. cannel-, definition of, 391. determination of, coke, 391, 393. ash, 391, 393. calorific value of, Ex. 71, 403. fixed carbon, 393. moisture, 390, 392. sulphur, 392, 393. by means of the Parr calorimeter, 394. volatile combustible matter, 391, 393. forms in which sulphur is present in, 392. proximate analysis of, 390, Ex. 70, 392. Cobalt, determination of, in salts of volatile acids, 76. electrolytic, determination of , 219. by ammonia or Fresenius and Bergmann's meth- od, 220. by ammonium oxalate or Classen's method, 219. separation, from copper, 214. from lead, 222. from manganese, 167. oxide, ignited, composition of, 61. precipitation with caustic alkali, 60. pure compounds of, 31. separation, as sulphide from man- ganese, 166. from iron, ]60. from iron and aluminium, 157. from iron, aluminium, and chromiimi, 156, 157. from manganese and nickel, 166. from nickel, as tri-potassium cobaltic nitrite, 167, 171. by means of nitroso-/?- naphthol, 168. from zinc, 145. Cobaltous hydroxide, properties of, 61. oxide, properties of, 61 . Cochineal, use as indicator, 251, 253. Coke, definition of, 391. determination of, 391, 393. Color, comparison of, in carbon determi- nations, 384. Color, determination of, in water, 409. Combustible matter, volatile, determi- nation of, in coal, 391, 393. Composition of fats and oils, 436. Conditions in which water is held, 39. Congo red, as indicator, 251. Constant weight, 41. Constitution, water of, determination of, 40. Copper, compounds of, conversion into sulphides, 78. determination, as sulphide, 78-80. by iodometric methods, 343. in brass or bronze, 147. in Britannia metal, 143. in cinnabar, 175a. in copper pyrites, 175. in copper sulphate, Ex. 12, 62, Ex. 20, 80. in Rose's metal, 130. in salts of volatile and organic acids, 71. in a silver coin, 124, 219. in type-metal, 141. in Wood's metal, 132. electrolytic, deposition, from a nitric or sulphuric acid solution, 213, Ex. 41, 217. from an ammonium ox- alate solution accord- ing to Classen, 215. determination of, 213. in copper sulphate, Ex. 40, 215. in a nickel coin, Ex. 44, 220. in a silver coin, 219. in ores, 214. separation, from lead, 222. from other metals, 214. from silver, 217. hydroxide, properties of, 61. oxide, ignited, composition of, 61. use of, in Kjeldahl digestion, 283. precipitation, as cuprous sulpho- cyanate, 79. with caustic alkali, 60, 62. pure compounds of, 31, 33. separation, from iron, 160. from zinc, 145. sulphate, determination, of copper in, 40, Ex. 12, 62, Ex. 20, 80, 215. of water in, Ex. 7, 41. percentage of copper in, 62, 81. water in, 42, 50. pure, preparation of, 33, Ex. 4,36. 538 INDEX. Correction factors for the Parr calorim- eter, 402. Counterpoise, use of, in weighing, 14. Crucible, Gooch, filtration with, 25. Rose, use of, 81. Crystallization, determination of water of, 40. Cubic centimeter, definition of, 232. Culture medium, for bacterial count, preparation of, Ex. 72, 428. sterOization of. Ex, 72, 429. Cupric hydroxide, properties of, 61. Cuprous chloride, absorption of carbon monoxide by, 464. ammoniacal solution of, 465. hydrochloric acid solutionof , 465. preparation of, 465. Current, electrical, method of obtaining, 202, 203. reduction, of voltage of, 202. of strength by incandescent lamps, 202. significance of density of, in elec- trolytic methods, 200. Cyanides, titration of, with silver ni- trate, 351, 354. Cyanogen, determination of, in potas- sium cyanide, Ex. 68, 354. Cylinders, platinum, as electrodes, 206. Daniell gravity cell, 203. Decantation, washing by, 26. Density, current, significance of, in electrolytic methods, 200. of water, table of, 527. Deshay's method of determining man- ganese, 388. Desiccator, use of, 19. Digestion-flaslis, Kjeldahl, 284. Digestion of precipitates, 23. Dihydric potassium phosphate, stand- ardization by, 358. Dimethyl amidobenzine, use of, as indi- cator, 251. Dishes, platinum, as cathodes, 205. roughening of, 205. Disodium phosphate, standardization by, 358; reagent, preparation of, 507. Dissolved oxygen, in water, method of determination, 421. Distillation, apparatus, of Bunsen, 344; of Mohr, 345; of Kjeldahl, 285. of arsenious chloride, 139. Dolomite, analysis of, Ex. 38, 180. Ealoulation of formula of, 495. Double salts, preparation of, 29. Drown's method of determining silicon in iron, 366. Drying-agents for gases, 44. Dudley, C. B., on coal analysis, 390. Duplicates, agreement of, 5. Economizing time, 124. Edison, Lalande cell, 204. primary cell, 204. Eggertz, colorimetric method for com- bined carbon, 383. Electric motors, use of, for rotation of electrodes, 212. Electrical charge on ions, amount of, 198. potential of, 199. Electrification, error in weighing by, 14. Electrodes, for rotation, 209. platinum cylinders as, 206. platinum gauze cylinders as, 207. rotation of, 207. Electrolysis, ionic theory of, 197. of warm solutions, 208. Electrolytic, deterrnination of metals, 213. methods, 197. significance of current density in, 200. secondary reactions, of acid radi- cles, 200. of metals, 201. separation of metals, by current density, 200. by potential, 199, 205. stands, 205, 207. Elements, difficulty of complete sepa- ration of, 121. Equations, analytical, 496. balancing of, 496. metathetical, 497. oxidation and reduction, 497. Errors, determination of amount of, 323. in reading, burettes, 229. pipettes, 231. in weighing, 14. of graduation, 232, 233. sources of, in water determina- tions, 40. unavoidable, 5. Frythrosene, use of, as indicator, 270. Eschka's method of determining sul- phur in coal, 392, 393. Ether value of fats and oils, 440. Factors, from a single precipitate, 490. method of calculating, 487. table of, 510. Factor weights, method of obtaining, 492. Faraday's law, 1 98. Fats and oils, acetyl value of, 447, 461. acidity of, 437. analysis of, 436. chemical composition of, 436. INDEX. 539 Fats and oils, detection of phytosterol and cholesterol in, 451 . detennination of fatty acid^ of, 455. ether value of, 440. free acid in, determination of, 438. Hehner value of, 443, 459. identification of, 456. iodine value of, 443, 457. Kottstorfer value of, 439, 459. Maumen6 number of, 449, 460. melting-point of fatty acids of, 451 . physical tests on, 448. Polenske value of, 441. determination of, 442. refraction, index of, 449. Reichert and Reichert-Meissl value, 440, 459. specific gravity of, 448. specific temperature reaction of, 449, 461. Fatty acids, determination of, 455. melting-point of, 451 . Feldspar, analysis of, Ex. 39, 192. Ferric, chloride, analysis of, by iodo- metric methods, 343. sulphocyanate, as indicator, 350. Ferrous, ammonium sulphate, prepa- ration of, 29. pure, testing of, 33. standardization of dichro- mate solutions by, 331, 332. of permanganate solu- tions by, 313, 316. iron, determination of, in ores, Ex. 57, 320. salts, determination of available oxygen in pyrolusite by, 322, 326. silicate, decomposition of, 319. sulphate, purification of, 30. titanate, decomposition of, 319. Filter-paper, ashless, 54. ignition of, 52, 56. on platinum wire, 58. Filter-pump, Bunsen, use gf, 24. Filtration, 24. by suction, 24. with Gooch crucible, 25. Flash point of oils, determination of, 452. Flasks, standard, 227. calibration of, Ex. 46, 244. by Morse-Blalock bulbs, 238. by weighing water, 233. for delivering capacity, 234. Floats for burettes, 229. Flue gases, interpretation of results of analy- sis of, 484. taking sample of, 484. Ford's method for the determination of manganese, 385. Formulae of salts and isomorphous mix- tures, calculation of, 493. Forster's modification of the Kjeldahl method, 288. Fractional weights, 8. Fresenius, apparatus for determining carbon dioxide, according to, 112. Fuels, determination of calorific value of, 395. Fuming nitric acid, oxidation of sul- phur compounds with, 103. Funnel, hot-water, 28. Gas, analysis, apparatus for, 471. double absorption pipette for, 475. method of filling, 475. explosion pipette for, 476. general methods of, 462. Hempel's apparatus for, 472. Orsat apparatus for, 481. manipulation of, 483. pipette for explosion with hot platinum wire, 477. simple absorption pipette for, 474. burette, simple, 472. AVinkler's modified, 472. explosive, definition of, 467. illimiinating-, analysis of, Ex. 72, 478. determination of, carbon di- oxide in, 478. carbon monoxide in, 479. hydrogen in, 480. "illuminants" in, 479. methane in, 480. oxygen in, 479. non-explosive, definition of, 467. pressure, regulation of, 471 . temperature, regulation of, 471. Gases, confining liquids for, 471 . determination of carbon dioxide in, 300. drying-agents for, 44. Geissler caustic-potash bulbs, use of, 114, 183. Gelatinous precipitates, filtration of, by means of paper pulp, 27, 54. General operations, 21. German silver, analysis of, Ex. 32, 148. Glass, expansion of, 232. solution of, in ammonia, 53. Glycerine, use of, in titrating boric acid, 293. Gold, determination of, in silver coin, 123. electrolytic separation of, from lead, 222. 540 INDEX. Gooch, crucible, use in filtering by suc- tion, 25. determination of boric acid by method of, 115. Graduation, errors of, 232. Graphitic carbon, determination of, 382. Gravimetric methods, advantage of, 2. definition of, 1, 226. standardization of acids by, 261, 263, 265, 272. Greiss, method of, for determining nitrites in water, 416. Gunning's modification of the Kjeldahl method, 287. Gunning-Jodlbauer modification of the Kjeldahl method, 288. Halogens, determination, 94. as silver salts, 94. in metallic salts, 95. separation of, 96. from hydrocyanic acid, 100. Hanus' monobromide solution, 444. Hardness, permanent, determination of, in water, 269, Ex. 50, 269. temporary, determination of, in water, 268, Ex. 50, 269. Heat of combustion, definition of, 396. Hehner value of fats and oils, 443. table of, 459. Hempel, double absorption pipette for gases, 475. method of filling, 475. explosion pipette, 476. gas, apparatus, 472. burettes, 472. simple absorption pipette for gases, 474. Hildebrand,W. F., on coal analysis, 390. Hofmeister capsule, use of, 290. Hot-water funnel, 28. Hydriodic acid, pure compounds of, 32. Hydrobromic acid, pure compounds of, 32. Hydrocarbons, formation of, in solu- tion of iron and steel, 364. heavy, absorbents for, 466. Hydrochloric acid, absorption of, by copper sulphate, 379. fifth-normal, preparation of, Ex. 48, 263. gas, generation of, 37, 140. normal, 247. pure compounds of, 32. reagent, preparation of, 505. specific gravity of, table of, 519. standard, advantages of, 255. preparation of, 263, 437. Hydrocyanic acid, determination of, 100. Hydrocyanic acid, separation of, from the halogens, 100. Hydrogen, basis for computing atomic weights, 486. gaseous, combustion by palladium sponge, 469. determination of, 467. in illuminating-gas, 480. in presence of methane, 470, 480. explosion of, with oxygen, 467, 480. explosion pipette for, 476. ignition of sulphides in, 81. slow combustion of, 468, 480. theory of combustion with palladium sponge, 469. peroxide, absorption of hydrogen sulphide by, 369. determination of, by iodo- ' metric methods, 343. pipette for combustion with hot platinum wire, 477. sulphide, formation of, in solution of iron and steel, 364. reduction of iron by, 320. solutions for absorbing, 369. Hydrometers, use of, 275. Hygroscopic, substances, weighing of, 19. water, determination of, 40. Hypochlorites, analysis of, by iodo- metric methods, 343. " lUuminants," absorbents for, 466. determination of, in gas, 479. Illummating-g:^s, analysis of, Ex. 72. 478. Incrustants, 269. determination of, 270. Index of refraction, determination of, 449. Indicator, iodine as, 308. potassium, ferrocyanide as, 308. penjianganate as, 308. Indicators, 250. basicity of acids with, 251. effect of carbon dioxide on, 250. for oxidation reactions, 307. for precipitation methods, theory of, 349. properties of various, 251, 253. Indirect determinations, by means of silver-jiitrate solution, 353. calculation of, 491 . Industrial processes, controlled by quantitative analysis, 6. International atomic weights, table of, 509. lodate, potassium acid, purity of, 35. INDEX. 541 Iodine, determination, as silver iodide, 95, 99. in metallic salts, 95. distillation of, 99. monobromide, Hanus' solution of, 444. monochloride, Wijs' solution of, 443. oxidizing power of, 306, 307. pure, preparation of, 34. reducing solutions used with, 336. resublimed, standardization of thiosulphate solution with, 337, 341. separation, from chlorine as thal- lous iodide, 97. from chlon.ie and bromine as palladous iodide, 96. from hydrocyanic acid, 100. solution, preparation and stand- ardization of, Ex. 64, 339. titration of, with sodium thio- sulphate solution, 340. solutions, conditions of use, 335. standard, preparation of, 336. standardization of, 336. by arsenious oxide, 339, 341. by barium thiosulphate, 338, 340. by potassium dichromate, 337. by potassium perman- ganate, 337. by resublimed iodine, 337, 341. by the iodates of potas- sium and sodium, 338, 341. solvents of, 335. value of fats and oils, 443. determination of, 445. table of, 457. lodometric, determination of, antimony, 343. copper, 343. sulphur, 369. methods. Chap. XXV, 335. analysis of, oxidizing sub- stances by, 342. reducing agents by, 342. sulphides by, 342. titration, of acids by, 294, 343. of bleaching powder by, 336. titrations, calculation of, 504. Ionization, constant, 21. law of, 21. of barium sulphate, 21. of metallic salts, 197. Ions, amountof electrical chargeon, 198. definition of , 21, 197. potential of electrical charge on, 199. Iron, analysis of, 364. determination, as sulphide, 83. in brass or bronze, 148. in Britannia metal, 143. in ehromite, 163. in cinnabar, 175a. in dolomite, 181. in. feldspar, 194. in German silver, 148. in ores, 318, Ex. 57, 320, 331, Ex. 62, 333. in pyrites, 175. in salts of volatile acids, 71. in silicates, 189. in smaltite, 169. in soft-iron wire, Ex. 11, 57. in soft solder, 129. in type-metal, 141. in water, 435. in Wood's metal, 132. of manganese in, by Deshay's method, 388. by Ford's method, 385. by Volhard's method, 387. of phosphorus in, 372. of silicon in, 365. after solution in hydro- chloric acid, 366. by Drown's method, 366. of sulphur in, 367. by solution in nitric acid, 367. of total carbon in, 374. electrolytic separation of, from copper, 214. from lead, 222. ferric, precipitation as hydroxide, 51. effect of organic matter on, 54. washing and ignition of, 52, 58. in zinc, titration of, 312. method of obtaining sample of, for analysis, 365. oxide, determination of silica in, 53. pig, method for determining sul- phur in, 368. pure compounds of, 31. reduction of, with hydrogen sul- phide, 194, 320. with stannous chloride, 319. with sulphurous acid, 320. with zinc, 311, 316, 319. 542 INDEX. Iron, separation, as basic acetate from manganese, cobalt, nickel, and zinc, 157. as basic carbonate from man- ganese, cobalt, nickel, and zinc, 159. as hydroxide from cobalt, nickel, and zinc, 156. by extraction with ether from chromium, aluminium, manganese, cobalt, nickel, and copper, 160, 169. by J. W. Roth's method, 160, 169. from aluminium as aluminate, 162. from manganese, 385. from zinc, 132, 145. volumetric, from aluminium, 162. solution in potassium cuprie chlo- ride solution, 376. standardization of permanganate solution by standard solution of, 310. volatilization of chloride, 381 . volumetric determination of, 194, 318; in manganese-phosphorus- bronze, 149. weighing of, as phosphate, 85. wire, determination of iron in, Ex. 11, 57. standardizations of dichro- mate solutions by, 330, 332. of permanganate solu- tions by, 309, 315. Jannasch and Aschoff, separation of halogens according to, 97. Jannasch and Heidenreich, decomposi- tion of silicates according to, 191. Jones' mannitol method of titrating boric acid, 294. Kjeldahl, digestion, 284, 290. digestion-flasks, 284. distillation, 285, 291. apparatus for, 286. method of determining nitrogen, 283. Forster's modification of, 288. determination of nitro- gen in potassium ni- trate by, Ex. 54, 291. Gunning's modification of, 287. determination of nitro- gen in milk by, Ex. 53, 289. Gunning-Jodlbauer modifica- tion of, 288. Kjeldahl, method of, substances decom- posed with difficulty by, 289. Wilfarth's modification of, 287. Knife-edge, of balance, 7. Kottstorfer value of fats and oils, 439. determination of, 439. table of, 459. Lacmoid, use as indicator, 251. Lamps, incandescent, reduction of cur- rent strength by, 202. Langmuir's method of determining rosin in shellac, 446. Lead, absorption of hydrogen sulphide by alkaline solution of, 369, 370, 371. acetate, reagent, preparation of, 507. carbonate, 64. ignition of, 64. determination of, as chromate, 90. as sulphate, 74. , as sulphide, 78. in brass or bronze, 147. in Britannia metal, 143. in lead nitrate, Ex. 17, 72 in ores, 223. in Rose's metal, 129. in salts of volatile and organic acids, 71, 77. in soft solder, 128. in type-metal as chloride, 140. in Wood's metal, 131. electrolytic deposition of, 201, 213, 222. conditions of, 222. separation and determination of, 213. separation of, from silver, 217. peroxide, dissolving off platinum, 223. oxidation of manganese bv, 388. separation of, by centrifugal machine, 389. precipitation by ammonium car- bonate, 64. pure compounds of, 31. separation, as chloride from bis- muth, 126. as sulphate, 126, 129. from antimony and barium, 127. of silver from, 123. sulphate, determination of sulphur in, 103. Ignition of, 75. properties of, '75. INDEX. 543 Limit of accuracy, in analysis, 122. in atomic weights, 486. liindo-Gladding method of detennining potassium, 91. Linen square, use of, in filtration, 24. use of, in sifting, 155. Liquids, taking samples of, 152. Liter, definition of, 232. Mohr, 232. various standards, in use, 232. Litmus, use of, as indicator, 251. Logarithms, division by, 488. method of using, 487. multiplication by, 488. table of, 514. Lunge, and Marchlewski, table of spe- cifiu. Tavityof hydrochloric acid, 519. and Rey, table of specific gravity of nitric acid, 520. and Weirnik, table of specific grav- ity of ammonia solutions, 518. Isler and Naef, table of specific gravity of sulphuric acid, 523. Lunge's nitrometer, use of, in standard- izing permanganate solutions, 314, 316. Magnesia mixture, 117. preparation of, 507. esium, ammonium arsenate, determination of phosphoric acid as, 117. ignition of, 87, 88. precipitation and ignition of, 88. ammonium phosphate, precipita- tion of, 85, 86. carbonate, determination of mag- nesium in, Ex. 19, 78. determination, as phosphate, 86, 87. in chromite, 165. in dolomite, 182. in magnesium carbonate, Ex. 19, 78. in magnesium sulphate, Ex. 23, 88. in salts of organic acids, 71. in salts of volatile acids, 71, 76. in water, 435. interference of, in determination of carbon dioxide in water, 298. pure compounds of, 31. pyroarsenate, determination of aresnic as, 89, 135. separation from the alkalies, 186. by means of barimn hydrox- ide, 187. by means of mercuric oxide, 186. Magnesium, sulphate, determination of magnesium in, Ex. 23, 88. of water in, Ex. 9, 47. percentage of magnesium in, 89. pure, preparation of, 35. Magnus, table of vapor tension of water, 526. Manganese, ammonium phosphate, pre- cipitation of, 85. ignition of, 87. carbonate, properties of, 69. ingition of, 69. determination, as phosphate, '85. as sulphide, 82. in chromite, 164. in cinnabar, 175 1. in dolomite, 181. in iron and steel, 385. Deshay's method for, 388. Ford's method for, 385. Volhard's method for, 387. in potassium permanganate, Ex. 22, 83. in pyrolusite, Ex. 59, 326. in salts of organic acids, 71. in salts of volatile acids, 71, 76. in smaltite, 171. volumetrically, 387, 388. dioxide, analysis of, by iodometric methods, 343. decomposition of potassium permanganate by, 308. determination of available oxj^gen in, Ex. 65, 322, 344, Ex. .58, 325. precipitation of, 385. electrolytic, deposition of, 201 . separation, from cobalt and nickel, 167, 171. from copper, 214. from lead, 222. hydroxide, properties of, 61. ores, analysis of, 322. oxidation of, to permanganic acid, 388. oxide, ignited, composition of, 61 . phosphorus-bronze, analysis of, Ex. 33, 149. precipitation, as dioxide, 385. with caustic alkali, 60. with sodium carbonate, 68. protosesquioxide, composition of, 69. pure compounds of, 31. separation, from chromium, 325. from copper, 214. from iron, 160, 3S5. from iron and aluminium, 159. 544 INDEX. Manganese, separation, from iron, aluminium, and chromium, 157. from lead, 222. from nickel and cobalt, 166, 167, 171. from zinc, 145. sulphide, green, precipitation of, 82. ignition of, 84. the two sulphides of, 82. volumetric determination in man- ganese-phosphorus-bronze, 149. weighing as oxide, 69. Mannitol, use of, in titrating boric acid, 294. Mantissa, definition of, 488. method of finding, 488. Marshand XJ-tube, 46. Mass, measurement of, 9. Maumen€ number, determination of, Maur enS numbers, table of, 460. MaziiJi-tiir load of balance, 15. MedicuSj solution of lead ores, accord- in;; '.o, 1^23. Melting-point of fatty acids, determi- nation of, 451. Mercuric chloride, pure, testing of, 34. reagent, 508. Mercurous chloride, properties of, 93. Mercury, as confining liquid for gases, 471. determination, as mercurous chloride, 93. as sulphide, 79, 175. by distillation, 1753. in cinnabar, 175, 175o. electrolytic separation of, from lead, 222. pure compounds of, 31. pure, testing of, 34. separation from copper, 214. from silver, 123. from zinc, 145. use of, in Kjeldahl digestion, 283. Metallic oxides, properties of, 50. Metals, determination, as oxide, 50. as phosphate, chromate, and chloride, 85. as sulphate and as sulphide, 73. as sulphide, 78. by evaporation with sulphuric acid, 76. by ignition of salts of volatUe acids, 70. electrolytic determination of, 213. secondary electrolytic reactions of, 201. specific heat of, 401 . taking samples of, for analysis, 153. Metals, washing and drying of electro- lytically deposited, 208. Metathetical equations, balancing of, 497. Methane, combustion of, 470. determination, of hydrogen in presence of, 470, 480. of, in Uluminating-gas, 480. Methods of weighing water, 39. Methyl alcohol, distillation of boric acid with, 116. Methyl chloride, volatilization of boric acid by, 191. Methyl orange, as indicator, use of, 253. effect of carbon dioxide on, 250. theory of action as indicator, 349. Microcosmic salt, determination of car- bon dioxide in carbonates by means of. 111. standardization by, 358. Milk, determination of nitrogen in, by the Kjeldahl-Gunning method, Ex. 53, 289. weighing sample of, 290. Mineral oils, detection of, 454. Minerals, analysis of, 152. arsenopyrite, 172. chalcopyrite, 172. chromite, 163. cinnabar, 175. dolomite, 180. feldspar, 192. pyrite, 172. smaltite, 169. calculation of formulas of, 493. taking samples of, for analysis, 153. Mohr, distillation apparatus of, 345. Mohr's salt, standardization of perman- , ganate solutions by, 316. Moisture, determination of , in coal, 390, 392. Molybdate solution, preparation of, 507. Morse-Blalock bulbs, calibration by means of, 235. calibration of, 236, Ex. 45, 242. burettes by, 238. flasks by, 238. description of, 235. temperature of water used with, 237. Mortars, agate, 155. steel, 154. Motors for rotation of electrodes, 210. MuUers, steel, 153. Naphthylamine acetate, solution of, for determining nitrites, 417. Neatness, necessity of, 3. Nessler, solution, preparation of, 411, tubes, 41 2. INDEX. 545 Wesslerizing, 414. Nickel, ammonium sulphate, determina- tion of nickel in, 62. coin, electrolytic analysis of, Ex. 44, 220. determination of, in German sil- ver, 148. in nickel ammonium sulphate, Ex. 13, 62. • in salts of organic acids, 71. in salts of volatile acids, 71, 76. in smaltite, 171 electrolytic, determination of, 219. by ammonia or Fresenius and Bergmann's meth- od, 220. by ammonium oxalate or Classen's method, 219. in a nickel coin, 221. separation from copper, 214. from lead, 222. from manganese, 167. oxide, ignited, composition of, 61. precipitation of, with caustic al- kali, 60, 62. pure compounds of, 31. separation, as sulphide from man- ganese, 166. from cobalt, as tri-potassium cobaltic nitrite, 167, 171. by means of nitroso-/?- naphthol, 168. from iron, 160. from iron and alimainium, 159. from iron, aluminium, and chromium, 156, 157. from manganese and cobalt, 166. from zinc, 145. Nickelous, hydroxide, properties of, 61. oxide, 61. reduction to nickel, 63. Nitrates, determination, by the Kjeldahl method, 288. in water, 418. of nitrogen in, 283, 288, Ex. 54,291. of NA in, 107. jeduction of, by Kjeldahl method, 283. Nitrie acid, determination of, by means of nitron, 106. in nitrates, 107. fuming, oxidation of sulphur com- pounds with, 103. pure compounds of, 32. reagent, preparation of, 505. Nitric acid, removal of, in sulphur de- terminations, 104. specific gravitjr, table of, 520. Nitrites, determination of, in water, 416. Nitrogen, determination, 105. by the Kjeldahl method, 283. in gases, 470. in milk, Ex. 63, 289. in potassium nitrate by the Kjeldahl method, Ex. 54, 291. in water, 411. organic, conversion of, into ammo, nium sulphate, 283. Nitrometer, Lunge's, use of, in stand- ardization of permanganate solu- tions, 314, 316. Nitron, determination of nitric acid by means of, 106. Nitrous acid, determination of, by iodo- metric methods, 343. Normal, acids, 247. hydrochloric, 247. percentage given by the num- ber of cubic centimeters used, 249. sulphuric, 247. chlorine in water, 432. oxidizing solutions, 307. reducing solutions, 307. solutions, 247. Note-book, method of keeping, 3. Noyes, Wm. A., on coal analysis, 390. Nutrient agar-agar for bacterial count, preparation of, Ex. 72, 430. Nutrient gelatin, adjusting acidity of, Ex. 72, 429. preparation of, Ex. 72, 428. Odor, numerical standards for, 410. determination of, in water, 410. Ohm's law, 203. Oil analysis, 454. Oils, determination of fatty acids ofj 455. of flash point of, 452. of viscosity of, 453. identification of, 4,56. mineral, detection of, 454. saponification of, 454. See also fats and oils. Oil. testers, use of, 453. Operations, general, 21. Ore-crushers, 154. Ores, determination of, lead in, 223. zinc in, 225. pulverizing of, 153. taking samples of, for analysis,153. Organic, acids, ignition of salts of, 71. 546 INDEX. Organic, acids, titration of, 255. of base in salts of, 253. matter, containing boric acid, de- composition of, 295. in water, determination of, 419. removal of, 54. Organic sulphides, produced by solu- tion of steel in acids, 370. Orsat apparatus, for gas analysis, 481. manipulation of, 483. Oxalates, electrolytic decomposition of, 201. solubility of, 121. Oxalic acid, anhydrous, preparation of, 260. standardization of acids with, 260. crystallized, standardization of acids with, 2596. pure, preparation of, 2596. determination of available oxygen in pyrolusite by, 322, 325. pure compounds of, 32. pure, testing and preparation of, 34, 2596. separation of arsenic and anti- mony from tin by, 134. solution, preparation of, 420. standard solution of, 303. use of, in determining carbon dioxide in air, 303. standardization of permanganate solutions by, 313. Oxidation, and reduction equations, balanc- ing of, 497. and reduction, methods, 306. titrations, calculation of, 502. definition of, 306. indicators for, 307. Oxides, metallic, properties of, 50. Oxidizing, solutions, normal, 307. substances, 306. analysis of, by iodometric methods, 342. Oxygen, available, determination of, in manganese ores, 322, Ex. 58, 325; in pyrolusite, Ex. 58, 325; in potassium dichromate, 307 ; in potassium permanganate, 307. basis for computing atomic weights, 496. determination of, in iUuminating- gas, 479. determination of, in water, 421. ' gaseous, absorbents for, 462. pure, not absorbed by phospho- rus. 463. Oxygen, pure, method of obtaining stream of, 378. method of producing, 468. required for water, determination of, 419. standardization of permanganate by measurement ol, 314, 316. table of solubility in fresh and sea water, 629. Palladium sponge, combustion of hy- drogen by, 469. Palladous iodide, properties of, 96. Pans of balance, 7. Paper pulp, use of, in filtration, 27, 54. Parr calorimeter, 397. chemical reactions involved in, 397. construction of the instrument, 399. correction factors for, 402. determination, of calorific value of coal by means of, Ex. 71, 403. of sulphur in coal by means of, S94. sodium peroxide for, 398. use of accelerators in, 401. water equivalent of, 400. Pemberton's alkalimetric method of tit- rating phosphates, 360. Penfield's method of determining water, 43. Percentage given by number of cubic centimeters of acid used, 249, 502. Persulphuric acid, analysis of, by iodo- metric methods, 343. Pettenkofer's method of determining carbon dioxide in water, 298. in gases, 301. Phenacetolin, use of, as indicator, 251. Phenol, use of, in the Kjeldahl diges- tion solution, 283. Phenol-sulphuric acid, solution of, for determination of nitrates, 418. Phenolphthalein, effect of carbon diox- ide on, 250. use of, as indicator, 251, 254. Phosphate precipitates, ignition of, 87. Phosphates, titration of, with permanganate solution, 362 with uranium solution, 357. Phosphine, formation of, in solution of iron and steel, 364. Phosphomolybdate, direct weighing of, 119. precautions in precipitation of, 118. . precipitation of, 119. separation of phosphoric acid as, 118. titration of, 360. by Pemberton's alkalimetric . method, 360. INDEX. 547 Phosphomolybdate, titration of, with permanganate solution, 362. washing of, 119. Phosphoric acid, detennination of, 117, 357. precipitation of, as magnesium ammonium phosphate, 117, 120, 361. pure compounds of, 32. separation of sihcic and arsenic acids from, 118. titration of, by Pemberton's meth- od, 360. by the uranium method, 357. with permanganate solution, 363. weighing as phosphomolybdate, 119. Phosphorous acid, reduction of mercury by, 93. Phosphorus, condition in which present in iron and steel, 364. determination, in iron and steel, 372. in manganese - phosphorus - bronze, 150. solution of iron in nitric acid and separation of silica for, 372. yellow, absorption of oxygen by, 463. moulding of, into sticks, 464. oxidation of, with potassium per- manganate, 373. pentoxide, drying properties of, 44, 46. method of filling U-tubes with, 47. use of, in Kjeldahl digestion, 283. pure, oxygen not absorbed by, 463. separation from titanium, 373. Physical, examination of water, 408. tests of fats and oils, 448. Phytosterol and cholesterol, detection of, in fats and oils, 451. melting-points of, 452. Pipettes, 227. double absorption for gases, 475. method of filling, 475. explosion, 476. for combustion with hot platinum wire, 477. method of using, 231. simple absorption, for gases, 474. Platino-chloride, ammonium, proper- ties of, 106. potassium, properties of, 90. Platinum, anodes, 205. Platinum, boat, for filtration of carbon on, 377. cone, use of, 24. crucible, damaged by ignition of phosphates, 88. cylinders, as electrodes, 206. dish, precipitating iron in, 57. dishes, as cathodes, 205. gauze cylinders as electrodes, 207. lead peroxide on, dissolving of, 223. tin deposits on, dissolving of, 221. utensils, rules for using and clean- ing, 38. zinc deposits on, dissolving of, 224. Polenske value, of fats and oils, 441. determination of, 442. of butter-fat, table of, 459; Policeman, 55. Post of balance, 7. Potash alum, calculation of formula of, 494. determination of aluminium in, Ex. 10, 64. percentage of Al^O^ in, 57. pure, preparation of, 34. Potassium, acid iodate, purity of, 35. standardization of sodium- thiosulphate solution with, 338, 341 . bromide, pure, testing of purity of, 35. chromate, fusion of nitrates with, 107. use of, as indicator, 350, 353. cyanide, determination of cyano- gen in, with silver nitrate, 351, Ex. 68, 354. fusion of stannic phosphate with, 150. solution, electrolytic deposi- tion of silver from, 217. determination, as platino-chloride, 90. by Lindo-Gladding method, 91. in feldspar, 196. in presence of sodium, 353,491 . in salts of volatile acids, 76. dichromate, indicator for, 308. oxidizing power of, 307. properties of, 330. pure, preparation of, 35, 330. reagent, preparation of, 508. solution, standardization of, by ferrous ammonium sul- phate, 331, 3.32. by iron wire, 330, 332. standard solution of, 330. conditions of use of, 330. preparation of, Ex. 61, 332. 548 INDEX. Potassium, dichromate, standardiza- tion, of acids with, 260. of sodium thiosulphate solution witti, 337. ferricyanide, as indicator, 308. in uranium titration, 357. hydroxide, absorption of carbon dioxide by, 464. reagent, preparation of, 506. standard alcoholic solution of, 281, 437. iodide, pure, testing of, 35. magnesium sulphate, preparation of, 29, Ex. 5, 36._ nitrate, determination of nitrogen in, by the Kjeldahl method, Ex. 54, 291. purity of, 35. standard solution of, 418. permanganate, decomposition of, by manganese dioxide, 308. determination of manganese in, Ex. 22, 82. impurities in, 309. its own indicator, 308. oxidizing power of, 306, 307. pure, testing of, 35. permanganate solution, alkaline for water analysis, 412. determining manganese by, 323. preparation of, 309, Ex. 56, 315. removal of manganese dioxide from, 309, 315. stability of, 309. standardization of, 309, 315. by direct measurement of oxygen, 314, 316. by ferrous ammonium sulphate, 313, 316. by oxalates, 313, 316. by oxalic acid, 313. by pure iron wire, 309, 315. by standard iron solu- tions, 310. standardization of sodium thiosulphate solution w'ith, 337. titration of phosphoric acid with, 362. platino-chloride, properties of, 90. pure compounds of, 31. separation from sodium, 187, 196. sulphate, pure, testing of, 35. use of, in the Kjeldahl diges- tion, 293. tetroxalate, preparation and stand- ardization of acids with, 260. pure, preparation of, 35. Potassium tetroxalate, standardization of permanganate solutions by, 313, 316. Potential, electrolytic, separation of metals by, 199. of electrical charge on ions, 199. Precipitates, contamination of, with soluble salts, 22. digestion of, 23. finely divided, treatment of, 23. ignition of, 56. transferring to the filter-paper, 54. washing of, 26. by deoantation, 26. without filtration, 92. weighable, properties of, 50. Precipitation, by change of solvent, 30. by double decomposition, 30 complete, theory of, 21. excess of reagent in, 22. methods, theory of indicators for, 349. of double salts, 29. Preparation, of potassium magnesium sulphate, Ex. 5, 36. of pure copper sulphate, Ex. 4, 36. of pure sodiimni chloride, Ex. 6, 37. Pressure, regulation of, in gas analysis, 471. Primary cells, 203. Problems, 245, 246, 257, 277, 278, 334, 356, 485. Proximate analysis of coal, 390, Ex. 70, 392. Pure salts, preparation of, 26. Pycnometers, determination of specific gravity of oils by, 449'. use of, 274. Pyrite, analysis of, Ex. 36, 172. in coal, 392. Pyrogallol, absorption of oxygen by, 462. Pyrolusite, commercial uses of, 322. determination, of available oxygen in, 322, Ex. 58, 325, Ex. 65, 345. of manganese in, by Volhard's method, 323, Ex. 59, 326. Quantitative analysis, neatness, necessary for, 3. object of, 1. patience and honesty necessary for, 3. required by industries, 6. skill and knowledge required for, 2. utility and importance of, 6. Reagents, preparation of, 505. Records, care in keeping, 3. INDEX. 549 Recrystallization of salts, 27. Reducing, agents, analysis of, by iodo- metric methods, 342. solutions, 307. substances, 306. Reduction and oxidation, equations, balancing of, 497. methods, 306. titrations, calculation of, 502. Reichert and Reichert-Meissl, value of fats and oils, 440. determination of, 440. table of, 459. Results, calculation of, 4. Retainers in water determinations, 40. Rider of balance, 8. Rocks, pulverizing of, 163. taking samples of, for analysis, 153. Rose crucible, use of, 81. Rose's metal, analysis of, Ex. 27, 129. determination, of bismuth in, 130. of copper in, 130. of lead in, 129. Rosin in shellac, determination of, by A. C. Langmuir's method, 446. Rosolic acid, use of, as indicator, 251 . Rossetti, table of density of water, 527. Rotation of anode or cathode, 209, 217, 221,225. Roth's ether separation of iron, 160. Rules for using balance, 20. Salicylic acid, use of, in the Kjeldahl di- gestion, 283, 288. Salt, common, purification of, 30, Ex. 6,37. Salts, calculation of formula of, 493. drying of, 28. pure, preparation of, 26. by recrystallization, 27. reason for analysis of, 121. weighing of, 41. Sample, drying of, 156. for analysis, selection and prepara- tion of, 152, 347. of air, collection of, 301, 305. of flue gases, taking of, 484. average, taking of, 484. of illummating-gas, collection of, 478. of iron and steel, method of ob- taining, 365. of water, taking of, 408. Sand-bath, use of, 42. Sanitary analysis of water, 406. Saponification of oils, 454. value of fats and oils, definition of, 439. Scale-forming ingredients of water, 434. SchrBtter apparatus, determination of carbon dioxide by, 109. sources of error in use of, 110. Selenium, determination of, 105. Sensibility of balance, 9. definition of, 10. determination of, Ex. 1,15. Separation of elements, difficulty of complete, 121. Sewage, contamination of water by, 407. Sewaschen, apparatus for determining carbon dioxide in air, 301. Shellac, determination of rosin by A. C. Langmuir's method, 446. Sieves, brass, 155. Sifting of powdered samples, 155. Silica, determination of, in chromite, 164. in dolomite, 181. in feldspar, 193. in ferric oxide, 59. in precipitates, 69. in pyrite, 175. in silicates, 193. in smaltite, 169. in water, 435. fusion of nitrates with, 107. precipitates, fusion of, 366. volatilization with hydrofluoric acid, 59, 181, 193, 365. Silicates, decomposition of, 188. by the method of J. Lawrence Smith, 190. determination, of alkalies in, 190. of iron, aluminium, and titan- ium in, 189. of silica in, 189. fusion of, with alkali carbonates, 188, 192. with boric oxide, 190. with calcium carbonate anc> ammonium chloride, 190. Silicic acid, dehydration of, 189. separation of, from phosphoric acid, 118. Silicon, condition in which present in iron and steel, 364. determination of, in iron and steel, 366 after solution of the iron in hydrochloric acid, 366. Drown's method for, 366. tetrachloride, formation of, in solution of iron and steel, 364. Silver, bromide, properties of, 95. chloride, as indicator, 351. properties of, 91. chromate, use of, as indicator, 350. 550 INDEX. Silver, coin, analysis of, Ex. 25, 123. determination of silver in, Ex. 69, 355. division of, volumetrically, 124. electrolytic analysis of, Ex. 43, 218. cyanide, determination of hydro- cyanic acid as, 100. determination, as chloride, 91. as sulphide, 78. by Volhard's method, 355. in a coin, Ex. 69, 355. in alloys, 355. in a silver coin, Ex. 25, 123, Ex. 43, 218. in silver nitrate, Ex. 24, 93a, Ex. 42, 218. electrolytic deposition of, 201, 217, 218. determination of, in silver nitrate, Ex. 42, 218. iodide, properties of, 95. metallic, purity of, 35. nitrate, determination, of potassium and sodium by, 353. of silver in, Ex. 24, 93a, Ex. 42, 218. indirect determination by means of, 353. purity of, 35. reagent, preparation of, 508. solution, preparation of, 424. standard solution of, 352, 354. titration of chlorides with, 350, 351. of cyanides with, 351, 354. nitrite, preparation of, 417. pure compounds of, 31. separation, from copper, 123, 214. from lead, 123. from lead and copper, 217. from mercury, 123. from other metals, 123. from zinc, 145. volumetric determination of, 226, 349. Smaltite, analysis of, Ex. 35, 169. Smith and Riche, method of determin- ing tin, 224. Smith, J. Lawrence, method of, for de- composing silicates, 190, 195. Soda-lime, absorption of carbon diox- ide by, 114. Soda reagent, preparation and use of, 269, 270. Sodium, ammonium phosphate, pure, test- ing of, 35. afseBitej solution, 336. Sodium, arsenite, standard solution of, 389. titration, of bleaching powder with, 347. of permanganic acid by, 389. bicarbonate and carbonate, analy- sis of mixtures of, 265, 267. bicarbonate, conversion to carbon- ate, 258, 2G4. carbonate, determination of, in caustic soda, Ex. 49, 267. and hydroxide, analysis of mixtures of, 265, 267. calculation of titration of, with standard acid, 248. ignition and weighing of, 71. precipitation of metals by, 68. pure, preparation of, 258. purification of, 30, 259. solution, determination of carbon dioxide in water by, 296. for water analysis, 412. standardization of acids with, 258, 264. testing for impurities, 258. chloride, purification of, 30. preparation of pure, Ex. 6, 37. determination, as carbonate, 71. in feldspar, 196. in presence of potassium, 353, 491. in salts of organic acids, 71. of volatile acids, 76. hydroxide, and carbonate analysis of mixtures of, 265, Ex. 49, 267. calculation of titration of, with standard acid, 248. determination of, in caustic soda, Ex. 49, 267. fifth-normal solution, prepa- ration of, Ex. 52, 282. precipitation of metals by, 60. reagent, preparation of, 506. solution, determination of car- bon dioxide in water by, 296. standard solutions of, 279. removal of carbon diox- ide from, 280. iodate, st'nd rdization of sodium thiosulphate solution with, 338. nitrite, stindird solution of, 417. oxalate (Surensen's), 259. standardization of potassium permanganate by, 316. peroxide, for the Parr calorimeter, 398; fusion of sulphur com- pounds with, 102. phosphate, pure, testing of, 35. INDEX. 551 Sodhun, pure compounds of, 31. separation of from potassium, 187, 196. sulphide, reagent, preparation of, 606. solution, preparation of, 137. sulphite, determinatioii of sulphur dioxide in, Ex. 66, 346. extraction of sulphur by, 80. thiosulphate solution, 336, 339. preparation and standardiza- tion of, Ex. 64, 339, 443. standardization, by arsenious oxide, 339, 341. by barium thiosulphate, 338, 340. by potassium dichro- mate, 337. by potassium perman- ganate, 337. by resublimed iodine, 337, 341. by the iodates of sodium and potassium, 338, 341. Solder, soft, analysis of, Ex. 26, 128. determination, of arsenic in, 129. of iron in, 129. of lead in, 128. of tin in, 128. of zinc in, 129. Solids, total, in water, determination of, 426. Solutions, calculation of standardization of, 499; standard, 226, 247. Sorensen's sodium oxalate, 259. Sources of error, in water determina- tions, 40. Specific gravity, definition of, 272. determination of, 272. by means of, hydrometers, 275. pycnometers, 274. the Westphal balance, 275. of fats and oils, 448. standardization of acids by, 261. standard temperatures for, 273. Specific heat, of metals, 401. of water, 396. Specific temperature, reaction of oils and fats, 449. table of, 461. Standard, acid, most desirable strength of, 256. percentage given by the num- ber of cubic centimeters used, 249, 502. acids, 258. Standard alcoholic caustic-potash solu- tions, 281, 437. alkalies, 279. alkaline solutions not permanent, 279. ammonia solutions, 282. barium-hydroxide solutions, 281. for set of weights, 18. hydrochloric acid, preparation of, Ex. 48, 263. liters in use, 232. sodium-hydroxide solution, 279, 282. preparation of, Ex. 52, 282. removal of carbon dioxide from, 280. solutions, 226, 247. calculation of standardization of, 499. steels for carbon determinations, 383. temperature for specific gravity, 273. for volumetric work, 231. Stands, electrolytic, 205, 207. Stannic oxide, impurities present in, 125. purification of, 125, 128. separation of tin as, 125. Stannic phosphate, decomposition of, 150. Stannous chloride, analysis of, by iodo- metric methods, 342. determination of tin in, 327, Ex. 60, 328. reagent, preparation of, 507. reduction of iron by, 319. Starch iodide paper, as indicator, 348. preparation of, 348. Starch solution, as indicator, 308, 339. method of preparing, 339. Steel, analysis of, 364. determination of manganese in, 385. by Deshay's method, 388. by Ford's method, 385. by Volhard's method, 387. of phosphorus in, 372. of silicon in, 365. of sulphur in, 367. by evolution method, 369. by solution in nitric acid, 367. of total carbon in, 374. method of obtaining sample of, for analysis, 365. standard, for carbon determina- tions, 383. 552 INDEX. Steel, solution of, for carbon determi- nations, 384. in potassium cupric-chloride solution, 376. Sterilization of culture medium for bac- terial count, Ex. 72, 429. Stirrups of balance, 7. Stoicluometry, 486. Stoking, effect of, on flue gases, 484. Storage, batteries, 204. cells, 203. Strontium, carbonate, determination of stron- tium in, Ex. 14, 66. ignition of, 65, 71. percentage of strontium in, 66. pure, preparation of, 33. weighing of, 66. determination, as sulphate, 74. in salts of organic acid, 71. in salts of volatile acids, 77. in strontium carbonate, Ex. 14, 66. precipitation, by ammonium car- bonate, 64. separation, as nitrate from cal- cium, 179. as sulphate from magnesium and the alkalies, 177. from barium and calcium, 176. from barium as cliromate, 178. from the alkali metals, 176. sulphate, properties of, 74. pure compounds of, 31. Suction, use of, in filtration, 24. Sulphanilic acid, solution of, for deter- mination of nitrites, 417. Sulphates, determination of, in water, 435a. insoluble, decomposition of, 103. solubility of, 121. Sulphides, analysis of, by iodometric methods, 342. determination of sulphur in, 101, 1J04. drying of, 79. ignition of, in hydrogen, 81. Sulphur, absorption of, by cadmium solu- tion, 372. by lead solution, 371 . available for sulphuric-acid manu- facture, determination of, 172. condition in which present, in iron and steel, 364. in coal, 392. crude, determination of sulphur in, 104. determination of, 101. by fusion with alkali nitrates and carbonates, 101, 174. Sulphur, determination of, by fusion with sodium peroxide, 102. by method of Bishop and Allen, 173. by oxidation, with aqua regia, 171, 172. with chlorine, 104. with fuming nitric acid, 103; with liquid bromine, 103, 104; with potassium chlorate, 103, 104. in coal, 392, 393. by means of the Parr calorimeter, 394. Eschka's method for, 392, 393. In iron and steel, 367. after solution in nitric acid, 367. in pig iron, 368. ill pyrite, 172. in reagents, 368. in smaltite, 171. in steel by evolution method, 369. apparatus for, 370. solution of iron for, 371. dioxide, absorption of, by lead per- oxide, 379. determination of, in sodiimi sulphite, Ex. 66, 346. evolution methods for, in steel, 369. iodometric determination of, 369. removal from sulphides, 79. weighing as cadmium sulphide, 372. Sulphuric acid, concentrated, drying properties of, 44, 46. method of filling U-tubes with, 46. fuming, absorption of iUuminants by, 466. normal, 247. pure compounds of, 32. reagent, preparation of, 605. specific gravity, table of, 523. standard, fifth-normal, prepara- tion of, Ex. 51, 271. titration of barium hydroxide with, 303, 305. Sulphurous acid, oxidation of, 306. reduction of iron by, 320. Synthetic equations, balancing of, 496. Table of Polenske values of butter-fat, 459. Tagliabue's oil tester, 453. Technical analysis, 364. Tellurium, determination of, 105. Temperature, regulation of, in gas analysis, 471. INDEX. 553 Temperature, standard, for volumetric work, 231. Tetrozalate of potassium, preparation of, 260. standardization of acids with, 260. Thallous iodide, properties of, 97. Theoretical percentage, calculation of, 491. Thermal units, defrnition of, 395. Thompson's glycerine method oi titrat- ing boric acid, 293. Thomson, R. T., classification of indi- cators, 250. table of basicity of acids with indi- cators, 252. Time, economizing, 124. Tin, deposits on platinum, dissolving ofif, 221. determination of, in brass or bronze, 146. in Britannia metal, 144. in salts of volatile acids, 71. in smaltite, 169. in soft solder, 128. in type-metal, 142. in Wood's metal, 131. electrolytic, determination of , 221. by ammonium-oxalate method of Classen, 221 . separation from antimony 137. , ' precipitation as sulphide, 78. pure compounds of, 31. separation, as stannic oxide, 125. from antimony, by H. Rose's method, 136. by means of metallic iron, 136, 141. from arsenic and antimony, 134. F. W. Clarke's method for, 134, 144. from copper, 215. from zinc, 145. volumetric determination of, 327. weighing as oxide, 80. Titanium, determination in feldspar, 194. in silicates, 189. separation of, from silica and phos- phorus, 373. Turbidity, determination of, in water, 408. numerical standards for, 409. Turmeric, use as indicator, 251. Twaddell scale, 276. Type-metal, analysis of, Ex. 29, 140. Units of volume, definition of, 232. Uranium, method of titrating phosphoric acid, 357. indicator for, 357. solution, titration of phosphates with, 359. standardization of, with cal- cium phosphate, 359. standard solution, preparation of, 358. U-tube, making impervious connections with, 47. Marshand, 46. method of filling, with calcium chloride, 46. with phosphorus pent- oxide, 47. with sulphuric acid, 46. Vacuo, reducing weights to, 11, 234. Valence, charge on ions proportional to, 198. Valve, Bunsen, 311. Vapor tension, of water, 526. Viscosity of oils, determination of, 453. Volatile combustible matter, determina- tion of, in coal, 391, 393. Volhard's method of determining man- ganese, 323, 327, 387. calculation of results of, 324. determination of amount of error in, 323. interfering metals, 324. method of adding perman- ganate solution in, 324. method of determining silver, 355. Voltage of current, reduction of, 203. separation of metals by, 199. Volt-meter, 205. Volume, units of, 232. Volumetric apparatus, 227 calibration of, 233. cleaning mixture for, 240. standard temperature for, 231. determinations; calculation of, 499. methods, 226. advantage of, 2. definition of, 1, 226. Wash-bottle, hot-water, 55 Washing precipitates, 26. by decantation, 26. without filtration, 92. Water, albuminoid ammonia in, determi- nation of, 411, 416. alkalies in determination of, 435. ammonia-free, preparation of, 412. analysis of, 406. calculation of results of, 435&. 554 INDEX. Water, analysis of, for use in boilers, 434. literature of, 435!). sanitary interpretation of re- sults of, 431. bacteria present in, 427. bacterial count in, 428, Ex. 72, 428. bacteriological examination of, 437. B. typhus in, 427. calcium in, determination of, 435. calibration, by weighing of, 233, 236. chemical examination of, 410. chlorides in, determination of, 424. color, determination of, 409. conditions in which it is held, 39. confining liquid for gases, 471 . density of, table of, 527. determination, 39. in barium chloride, Ex. 8, 44. in copper sulphate, Ex. 7, 41. in magnesium sulphate, Ex. 9, 47. of carbon dioxide in, 296. by Pettenkofer's method, 298; in bottled car- bonated waters, 300. of dissolved oxygen in, 421. of free mineral acid in, 297. sources of error in, 40. direct weighing of, 47. expansion of, 232. free ammonia in, determination of, 411, 413. free mineral acids in, determina^ tion of, 435 . ground, characteristics of, 433. hardness, determination of, 268, 434, Ex. 50, 269. hygroscopic, determination of, 40. iron and aluminium in, determina- tion of, 435 J. magnesium in, determination of, 435. motor, use of, for rotation of elec- trodes, 210. nitrates in, determination of, 418. nitrites in, determination of, 416. nitrogen in, determination of, 411. normal chlorine of, 432. numerical standards for odor of, 410. odor, detemiination of, 410. of constitution, determination of, 40. of crystallization, determination of, 40. organic matter in, determination of, 419, 421. oxygen, required, determination of, 419. Water, Penfield's method of determin- ing, 43. physical examination of, 408. .sanitary analysis of, 406. scale-forming ingredients of, 434. silica in, determination of, 435. source of impurities in, 407.' specific heat of, 396. sulphates in, determinationof, 435a. surface, characteristics of, 433. taking sample of, 408. temperature of, 408. test for B. coli in, 4?7, Ex. 73, 431. total solids in, determination of, 426. turbidity, determination of, 408. vapor tension of, 526. weighing substances containing, 19. WeighaWe precipitates, properties of, 50. Weighing, a salt, 41. bottle, 19. errors in, 14. hygroscopic substances, 19. independent of force of gravity, 9. limit of error in, 5. method of substitution for, 12. methods of, 7, 12. substances containing water, 19. tube, 49. Weights, calibration of, Ex. 3, 16. constant, 41. factor, method of obtaining, 492. fractional, 8. reducing to vacuo, 11. standard for, 18. Westphal balance, use of, -271, 275. Whipple, Geo. C, characteristics of pol- luted surface-waters, 433. Whipple, Geo. C. and Whipple, M. C, table of solubility of oxygen in fresh and sea water, 529. Wijs' iodine-monochloride solution, 443. Wilfarth's modification of the Kjeldahl method, 287. Winkler modified gas burette, 472 Wood's metal, analysis of, Ex. 28, 131. Zeiss butyro-refractometer, use of, 449. Zero-point of balance, 10. determination of, 15. variations in, 11. Zinc, absorption of hydrogen sulphide by alkaline solution of, 369. ammonium phosphate, ignition of, 86. precipitation of, 85. INDEX. 555 Zinc, ammonium sulphate, detennina- tion of zinc in, Ex. 16, 70. percentage of zinc in, 70. carbonate, properties and ignition of, 69. deposits on platinum, difficulty of dissolving, 224. determination, as phosphate, 84. as sulphide, 81. in brass or bronze, 148. in German silver, 148. in ores, 225. in salts of volatile and organic acids, 71. in soft solder, 129. in Wood's metal, 132. in zinc ammonium sulphate, Ex. 16, 70. electrolytic^ determination of, 224.- Zinc, electrolytic, separation, from copper, 214. from lead, 222. oxide, neutralization of manganese solution with, 388. properties of, 70. purification of, 388. precipitation with sodium carbon- ate, 68, 69. pure compounds of, 31. reduction of iron solutions by, 311^ 319. separation, as sulphide, 145. as zincate, 145. from iron, 132. from iron and aluminium, 159. from iron, aluminium, and chromium, 156, 157. sulphate, pure, preparation of, 35 titration of iron in, 312. LITERATURE OF THE CHEMICAL INDUSTRIES On our shelves is the most complete stock of technical, industrial, engineering and scientific books in the United States. The technical liter- ature of every trade is well represented, as is also the literature relating to the various sciences, including books useful for reference as well as those fitted for students' use as textbooks. A large number of these we publish and for an ever increasing number we are the sole agents. ALL INQUIRIES MADE OP US ARE CHEER- FULLY AND CAREFULLY ANSWERED AND COMPLETE CATALOGS AS WELL AS SPECIAL LISTS SENT FREE ON REQUEST D. VAN NOSTRAND COMPANY Publishers and Booksellers 8 WARREN STREET NEW YORK