UNIVERSITY OF CALIFORNIA 
 AT LOS ANGELES
 
 last date star 
 
 SOUTHERN BRANCH, 
 
 UNIVERSITY OF CALIFORNIA, 
 
 LIBRARY, 
 
 LOS ANGELES, CALIF.
 
 WORKS OF 
 
 PROF. HENRY FAY 
 
 PUBLISHED BY 
 
 JOHN WILEY & SONS, Inc. 
 
 Microscopic Examination of Steel. 
 
 iv+8z pages, 6 by 9, many plates. Cloth, $1.50 net. 
 
 An Advanced Course in Quantitative Analysis with Explana- 
 tory Notes. 
 
 v+ui pages, 6 by 9. Cloth, $1.50 net.
 
 An Advanced Course 
 
 IN 
 
 Quantitative Analysis 
 
 WITH 
 
 Explanatory Notes 
 
 HENRY FAY, Ph.D., D.Sc. 
 
 Professor of Analytical Chemistry and Metallography in the Massachusetts 
 Institute of Technology 
 
 SECOND EDITION REVISED 
 
 NEW YORK 
 
 JOHN WILEY & SONS, INC. 
 
 LONDON: CHAPMAN & HALL, LIMITED 
 1922 
 
 49108
 
 COPYRIGHT, 1917, 1922, 
 
 BY 
 HENRY FAY 
 
 Stanhope preos 
 
 F. H.GILSON COMPANT 
 BOSTON, U.S.A.
 
 rot 
 
 PREFACE TO SECOND EDITION 
 
 IN offering a new edition every effort has been made to so correct 
 
 the text in detail as to make it more certain for the student to get 
 
 I accurate results. The only additions have been in the "Analysis 
 
 | of Duralumin" which has been selected not only on account of 
 
 3 its importance as a type of aluminum alloys but because of the 
 
 opportunity to introduce some new principles. 
 
 Much consideration has been given to the various electrometric 
 ^ methods and it has been a great temptation to introduce them 
 t but they were omitted on account of their highly specialized nature 
 and the desire to keep the volume small. 
 
 The author wishes to express his appreciation of the help re- 
 ceived from his associates of the Division of Analytical Chemistry. 
 
 HENRY FAY. 
 September, 1922. 
 
 iii
 
 PREFACE TO FIRST EDITION 
 
 THIS book is intended primarily for students of the Massachu- 
 setts Institute of Technology, and represents the work planned for 
 those who have finished the excellent introductory course in 
 "Quantitative Analysis" by Professor Henry P. Talbot. Assump- 
 tion has been made that the student is familiar with the use of the 
 balance, the principles of volumetric analysis, and of stoichiometry, 
 and no attempt has been made to enlarge upon these subjects. The 
 material offered is chosen to illustrate principles and to train the 
 student in manipulation. In certain cases a student is offered a 
 choice of method, as in the determination of sulphur in pyrite, and 
 in other cases where the principle differs, as in the determination 
 of manganese in steel, it is desirable that the student acquaint 
 himself with each method. 
 
 The student is earnestly requested to read and understand each 
 method before beginning any laboratory work. Adequate prepa- 
 ration of apparatus should also be made. 
 
 In preparing this volume the author wishes to express his appre- 
 ciation of the helpful suggestions of Professor R. S. Williams, and 
 of the kindly services of Mr. L. F. Hamilton. 
 
 HENRY FAY. 
 
 MASSACHUSETTS INSTITUTE OF TECHNOLOGY, 
 May, 1917.
 
 CONTENTS 
 
 PART I 
 
 MINERAL ANALYSIS 
 
 PAGE 
 SAMPLING FOR ANALYSIS i 
 
 DETERMINATION OF SILICA IN SILICATES 4 
 
 Decomposable Silicate (Laumontite) 
 
 Refractory Silicate (Feldspar) 
 DETERMINATION OF POTASSIUM AND SODIUM IN SILICATES 9 
 
 J. Lawrence Smith Method. 
 
 Perchloric Acid Method. 
 ANALYSIS OF SPATHIC IRON ORE 15 
 
 Determination of Iron, Aluminum, Manganese, Calcium, and Magnesium. 
 DETERMINATION OF SULPHUR IN PYRITE 29 
 
 Lunge Method; Allen and Bishop Method; Fresenius Method; 
 
 Sodium Peroxide Method. 
 DETERMINATION OF TITANIUM IN TITANIUM IRON ORE 39 
 
 Method of Gooch; Method of Barneby and Isham. 
 IODOMETRIC DETERMINATION OF COPPER 43 ]/ 
 
 Method of Low. 
 PROXIMATE ANALYSIS OF COAL .... 47 
 
 Determination of Moisture, Ash, Volatile Matter, Fixed Carbon, 
 Sulphur. 
 
 PART II 
 METAL ANALYSIS 
 
 ANALYSIS OF PHOSPHOR-BRONZE 50 / 
 
 Determination of Copper, Lead, Tin, Phosphorus. 
 DETERMINATION OF CARBON IN STEEL 59 
 
 Direct Combustion Method. 
 DETERMINATION OF MANGANESE IN STEEL 65 
 
 Volhard Method; Williams Method; Bismuthate Method. 
 DETERMINATION OF PHOSPHORUS IN IRON AND STEEL 75 / 
 
 Reduction Method; Alkalimetric Method. 
 DETERMINATION OF SULPHUR IN STEEL 83 
 
 Evolution Method. 
 
 vii
 
 viii CONTENTS 
 
 PAGE 
 
 DETERMINATION OF SULPHUR IN CAST IRON 87 
 
 Bamber Method. 
 
 DETERMINATION OF SILICON m CAST IRON 89 
 
 Drown Method. 
 
 DETERMINATION OF COPPER IN STEEL 91 
 
 DETERMINATION OF NICKEL nsr STEEL 92 
 
 Brunck's Method. 
 
 DETERMINATION OF CHROMIUM IN STEEL 94 
 
 Barba Method; Cain Method. 
 
 DETERMINATION OF TUNGSTEN IN STEEL 98 
 
 DETERMINATION OF VANADIUM IN STEEL 100 
 
 ANALYSIS OF DURALUMIN 103 
 
 LOGARITHM TABLES 108 
 
 TABLE OF ATOMIC WEIGHTS 112
 
 QUANTITATIVE ANALYSIS 
 
 SAMPLING FOR ANALYSIS 
 
 THE selection and preparation of a sample for analysis, 
 whether it be an ore, rock, mineral, or metal, is just as im- 
 portant as the method of analysis itself. In a brief dis- 
 cussion of the subject it is very difficult to lay down general 
 rules for making the samples owing to the varying con- 
 ditions, such as size of shipment, or the character of the ore 
 or metal. The desired end is to obtain a sample which will 
 be representative of a whole shipment, whether it be a car 
 load, a train load, a cargo, or a large or small shipment of 
 metals in their varied forms. Many of the controversies 
 between producer and consumer may be traced to inequality 
 of sample. The sampling by the producer and consumer 
 in the case of ores is done usually under very different con- 
 ditions and it is quite remarkable that there should be such 
 close agreement rather than that there should be much 
 divergence. 
 
 Inasmuch as the principle of mineral or rock sampling is 
 in general the same, the sampling of a car of ore will be 
 taken as typical. Immediately on receipt of the car, samples 
 should be taken by means of a small shovel or trowel from 
 various parts of the car. The best method of getting even 
 distribution is to take a one- or two-pound sample from a 
 large number of places, the car being divided so as to obtain 
 the samples at equal distances from the centre and ends, as 
 well as between; or the car may be sampled along a zig-zag 
 line running from one corner to the diagonally opposite corner, 
 crossing the car at one- third and two- thirds of the length. 
 This method must be followed not only at the surface of the 
 car, but also midway between top and bottom and near the
 
 2 QUANTITATIVE ANALYSIS 
 
 bottom of the car. When lumps are encountered, small 
 portions of each lump must be chipped off and an equal 
 portion of fine must be taken adjacent to the lump. 
 
 The total sample is then mixed, piled in a cone and alternate 
 quarters rejected. The material thus obtained is further 
 subdivided by crushing or grinding, and again coned and 
 quartered, and this process continued until a one- or two- 
 pound sample which will pass through a loo-mesh sieve is 
 obtained. In rock or mineral sampling, it is absolutely 
 necessary to crush or grind all of the quarters taken and not 
 to reject any portion on account of extreme hardness or 
 softness. This would defeat the end which is aimed at. 
 
 Precaution should also be taken when removing a sample 
 from the general supply bottle as there is strong tendency 
 for the coarser particles to find their way to the bottom of 
 the container. The bottle should be well shaken, or prefer- 
 ably rotated on a wheel, or the contents spread out and mixed 
 on a large piece of glazed paper. Unexpected variations may 
 arise from year to year from neglect of this precaution. 
 
 The sampling of ores is made difficult by the uneven char- 
 acter of coarse and fine, hard and soft, material, and of minerals 
 of high and low specific gravity. The degree of fineness to 
 which a mineral or ore is ground is also an important factor. 
 Many analyses are affected due to failure to take this fact 
 into consideration. In the determination of water, sulphur 
 in pyrite, or ferrous Iron in silicates large variations may 
 occur between different analysts because of the differences in 
 the fineness of the samples used. In fine grinding oxidation 
 of sulphur or iron may occur, or there may be loss of water. 
 It is advisable in such cases to make determinations on both 
 fine and coarse material. 
 
 The difficulty of sampling metals is principally due to the 
 tendency to segregation in the ingot, or cast form. This 
 tendency is due to the fact that during solidification selec- 
 tive crystallization takes place and thus produces a melt 
 richer in one or more constituents which solidify at a lower 
 temperature. When a metal passes from the liquid to the
 
 SAMPLING FOR ANALYSIS 3 
 
 solid state, the outside skin solidifies first on account of the 
 more rapid conduction of heat from the surface, and thus 
 leaves a molten interior. This molten mass gives up crystals 
 to those already formed and there is a constant enriching of 
 the material in the melt. In steel, carbide and phosphide 
 of iron are thus concentrated and are always segregated 
 in that portion of an ingot last to solidify, viz.: in the upper 
 central portion of the ingot. 
 
 Again, compounds of lower or higher specific gravity than 
 the main mass of the melt may form and thus rise or fall in 
 the still molten mass. A good example of this type of segre- 
 gation is in the case of sulphur in steel which is always com- 
 bined with manganese as manganese sulphide and being of 
 lower specific gravity than the steel itself rises to the surface 
 of the melt. 
 
 Due regard to segregation must be made in the sampling of 
 any cast metal. In products which are forged from the ingot 
 the segregation persists in the forged material irrespective of 
 shape. Rigid specifications for steel usually call for the rejection 
 of definite portions of the top (usually 25 per cent) and bottom 
 (usually 5 per cent) of the ingot, but even in spite of these 
 precautions, differences are invariably found. It is never 
 safe to assume that any metal is uniform and it is much better 
 to assume that it lacks uniformity and proceed accordingly. 
 
 In the sampling of a steel rail, for instance, it would be 
 foolish to assume that a single sample would represent the 
 whole unless that single sample were obtained by boring a 
 hole entirely through the rail from top to bottom. Even so 
 this method would not be so satisfactory as taking samples 
 from a number of holes distributed over the head, foot, and 
 web, or better by planing over the whole surface. In all 
 cases, as much of the surface as possible should be represented 
 in the sample. 
 
 If possible, a macroscopic survey should be made before 
 sampling any steel if the material is likely to be a subject of 
 controversy. To this end, a cross section of the piece is fairly 
 smoothly polished and then etched either with an 8 per cent
 
 4 QUANTITATIVE ANALYSIS 
 
 solution of copper ammonium chloride 1 or a 6 per cent solution 
 of iodine in alcohol. In the former case segregated carbide 
 and phosphide appear darker than the surrounding metal, and 
 in the latter case phosphide in particular appears lighter 
 colored. Interpretation of results thus obtained should always 
 be left to those experienced in this work. 
 
 It is needless to add that the tool with which metal samples 
 are taken should be free from oil. 
 
 DETERMINATION OF SILICA IN A DECOMPOSABLE 
 SILICATE 
 
 LAUMONTITE 
 
 Weigh approximately 0.5 gram of the finely ground material 
 into a 300 cc. porcelain casserole. Add about 25 cc. of water, 
 and stir until the silicate is completely wet. Now add in 
 small portions 25 cc. hydrochloric acid (sp. gr. 1.12), heating 
 the casserole and stirring the contents during the addition of 
 the acid, and continuing the heating until the silicate is com- 
 pletely decomposed or until no gritty residue is left. Evapo- 
 rate the solution to dryness, stirring frequently if there is any 
 tendency toward the formation of a jelly-like mass. When 
 completely dry, add just enough concentrated hydrochloric 
 acid to moisten the residue, and evaporate again. Moisten 
 the residue again with concentrated hydrochloric acid, warm 
 gently, then dilute with 200 cc. of water, and heat to boiling. 
 Filter off the silica as soon as it settles, without attempting 
 to remove that which adheres to the wall of the casserole, and 
 wash with hot dilute hydrochloric acid (one part acid, sp. gr. 
 1. 1 2, to three parts water) until the filter paper is free from 
 iron stain. Finish the washing with hot water, and wash 
 until free from chlorides. 
 
 Evaporate the filtrate and washings to dryness in the same 
 casserole, and treat exactly as in the first evaporation, receiv- 
 ing the recovered silica on a second filter paper, and removing 
 the silica adhering to the walls of the casserole by means of 
 
 1 Heyn. International Soc. for Testing Materials, Brussels Congress, 1906.
 
 DETERMINATION OF SILICA 5 
 
 a rubber "policeman." Wash as before, and then transfer 
 the partially dried filter papers to a weighed platinum crucible. 
 Place the crucible on its side and, by means of a low flame 
 placed under the front of the crucible, smoke off the paper at 
 a low heat. When the volatile products are removed, place 
 the full oxidizing flame of the Tirrill burner at the rear of the 
 crucible, and heat until the residue is white. Finish the 
 ignition, with the crucible covered, over a Meker burner or 
 blast lamp, and heat to constant weight. 
 
 When constant weight has been obtained, add to the cru- 
 cible one or two drops of v/ater, and then 3 to 4 cc. of hydro- 
 fluoric acid. Add one or two drops of dilute sulphuric acid, 
 and heat gently in a hood with good draught, until all of the 
 silica has volatilized. If there is considerable residue it may 
 be necessary to repeat the treatment with hydrofluoric acid. 
 When all of the silica has volatilized, heat over the Tirrill 
 burner, and finally over a Meker burner or blast lamp for a 
 few minutes. The loss in weight represents the amount of 
 SiC>2 present. 
 
 DETERMINATION OF SILICA IN A REFRACTORY 
 SILICATE 
 
 PELDSPAR 
 
 Mix thoroughly in a platinum crucible approximately 0.50 
 gram of finely ground feldspar with about six times its weight 
 of pure, dry sodium carbonate. Place the covered crucible 
 over a small flame and gradually raise the flame to full heat 
 and continue heating for 20-30 minutes, or until effervescence 
 has ceased. Now place the crucible in the oxidizing flame of 
 a Meker burner or blast lamp, directing the flame at an angle, 
 if the blast lamp is used, against the side and bottom of the 
 crucible, and continue heating for five minutes. At the end 
 of this time shut off the gas supply suddenly and allow the air 
 blast to play upon the covered crucible until cool. Place the 
 crucible top side down in a porcelain casserole and gently tap 
 until the solidified mass is removed. If the Meker burner is
 
 6 QUANTITATIVE ANALYSIS 
 
 used the best method of removing the melt is to insert into the 
 molten mass the coiled end of a platinum wire and allow the 
 mass to solidify around the wire. When cool, reheat gently 
 with a small flame so as to soften the edges and at the same 
 time put the wire under tension. If properly done this should 
 remove the melt as a whole. Wash with hot water the material 
 adhering to the cover and crucible into the casserole and then 
 proceed with the decomposition and determination of the silica 
 exactly as described in the preceding method. 
 
 Notes. i. The ease with which any silicate is decomposed 
 is determined by the ratio of base to acid. In the more basic 
 silicates, hydrochloric acid will decompose them readily. Such 
 silicates are referred to as soluble or decomposable. With 
 increase in the ratio of silica to base the decomposition with 
 acids takes place less readily, and when the silicate is distinctly 
 acidic in character there is little, if any, action by dilute acids. 
 The latter type of silicate is called refractory, or undecompos- 
 able. Such silicates may, however, be rendered decomposable 
 by converting them into basic silicates by fusion with sodium 
 carbonate, lead oxide, or other similar basic compounds. In 
 this way the proportion of base to silica is so raised as to make 
 them easily decomposable by dilute acids. 
 
 Most of the artificial silicates, such as the various metal- 
 lurgical slags, Portland cement, etc., are decomposable. 
 
 Certain silicates occurring as minerals are insoluble in their 
 natural state, but, when brought to the fusion temperature, 
 are converted, after cooling, without change of composition, 
 into soluble silicates. 
 
 2. When strong acid is brought in contact with decompos- 
 able silicates silicic acid is separated in gelatinous form which 
 encloses particles of undecomposed silicate. On evaporation 
 this jelly-like mass dries around the unattacked particles and 
 protects them from the action of the acid. In order to avoid 
 this difficulty, the silicate is first moistened with water, and 
 small amounts of acid are added from time to time so as not to 
 have it in excess until after the decomposition has taken place. 
 By the cautious addition of acid the whole of the silicate may 
 be brought into solution and no silica will separate until the 
 solution has been concentrated somewhat.
 
 DETERMINATION OF SILICA 7 
 
 Stirring and breaking up of the jelly, when once formed, aids 
 the evaporation and also prevents spattering. 
 
 3. The recovery of silica by a second evaporation after the 
 filtration of the major portion seems to be absolutely neces- 
 sary. Formerly it was the custom to recover the silica passing 
 into the nitrate after a single evaporation by evaporating the 
 precipitate of iron and alumina produced by ammonia with 
 either hydrochloric or sulphuric acids, or to correct the ignited 
 oxides of iron and alumina for silica by fusion with potassium 
 pyrosulphate. The double evaporation method, with inter- 
 vening nitration, seems to be more efficient and less cumber- 
 some. 
 
 The necessity of the recovery of the silica passing into the 
 filtrate after a single evaporation was early recognized by 
 Bunsen, 1 later by Ludwig, 2 and by Meineke, 3 and more re- 
 cently by Cameron, 4 and Hillebrand. 5 
 
 Invariably small amounts are recovered by the second 
 evaporation, and consequently the operation should never 
 be neglected. This process has the distinct advantage over 
 other methods in that no silica is introduced through the use of 
 ammonia, which always carries from small to large amounts. 
 
 4. The temperature of drying the siliceous residue has been 
 the subject of considerable investigation and much difference 
 of opinion. 
 
 According to the work of Gilbert 6 and of Hillebrand, 7 dry- 
 ing temperatures above that of the steam bath have little 
 advantage and are sometimes the cause of considerable silica 
 going into the filtrate on account of recombination of silica 
 at the higher temperatures. Furthermore, silica so obtained 
 frequently contains large amounts of impurity. Blount 8 
 recommends a temperature of 200 for the determination of 
 silica in Portland cement, but this is advisable only hi certain 
 special cases and should not be adopted for general work. 
 
 Am. Chem. Pharm., 61, 265. 
 
 Z. anal. Chem., 9, 321. 
 
 Repert. anal. Chem., 7, 215. 
 
 Chem. News, 69, 171. 
 
 Bulletin 422, U. S. Geol. Survey, p. 91. 
 
 Tech. Quart., 3, 61. 
 
 /. Am. Chem. Soc., 24, 262. 
 
 /. Soc. Chem. Ind., 21, 1217.
 
 QUANTITATIVE ANALYSIS 
 
 5. The final temperature of ignition for- silica, as well as for 
 other oxides, such as A^Os, Fe 2 O 3 , TiO2, Sn02, etc., must 
 be high. It is claimed by many investigators that this heat 
 is necessary to deprive the silica of the last traces of water, 
 although the evidence is not conclusive that moisture alone is 
 the cause of variation in weight. It has been shown by experi- 
 ments conducted in this laboratory that in many cases sodium 
 or potassium salts are the cause and that it is necessary to 
 have a prolonged high temperature to remove the traces of 
 alkali which had been occluded by the silica. 
 
 When igniting to constant weight, it must not be forgotten 
 that some platinum crucibles lose weight steadily, and the rate 
 of loss should therefore be determined for such a crucible. 
 
 6. Silica, as determined by evaporation to dryness, is never 
 absolutely pure, but is always contaminated by traces of 
 aluminum, iron, calcium, or magnesium. The weight must 
 always be corrected by volatilizing the silica as fluoride and 
 weighing the residue after heating as before. This residue, 
 although frequently small, should never be neglected. 
 
 7. Sodium carbonate is invariably contaminated with 
 silica and other impurities. For analytical purposes it should 
 either be purified by recrystallization, or the amounts of silica, 
 etc., should be carefully determined. One should never be 
 deceived by a label and should prove to his own satisfaction 
 the presence or absence of impurities. If a sample of pure 
 sodium carbonate be prepared or the amount of silica be de- 
 termined, such a sample should be carefully guarded and used 
 only for silicate analyses. 
 
 Sodium bicarbonate has been recommended as a flux and 
 has much to commend its use, but it is very seldom employed. 
 
 8. Lead oxide, 1 bismuth oxide, 2 and boric oxide 3 have been 
 proposed for the decomposition of refractory silicates. By 
 the use of these substances there is not only the distinct ad- 
 vantage that all of the flux may be removed lead and bis- 
 muth by means of hydrogen sulphide, and boron by means of 
 methyl alcohol thus preventing adhesion of sodium salts, 
 but sodium and potassium may be determined in the 
 
 1 Berthier, Ann. de Chim. el de Phys. (2), 17, 28. 
 
 1 Hempel, Z. anal. Chem., 20, 496. 
 
 * Jannasch and Heidenreich, Z. anorg. Chem., 12, 208.
 
 DETERMINATION OF POTASSIUM AND SODIUM 9 
 
 sample. These methods have much to commend them, but 
 they have not been extensively used. The principal objection 
 to the first two is the easy reducibility of lead and bismuth 
 either by the reducing flame of the burner or by the presence 
 of organic matter. This objection is not valid if ordinary 
 precautions are taken. 
 
 Hillebrand has urged two objections against the boric oxide 
 method, viz.: the deposition of boric acid due to the hydrolysis 
 of the volatile boric ether, and the use of the oxygen flame when 
 the alkalis are to be determined. Another very important 
 objection is the difficulty in preparing the flux in pure and 
 finely divided condition. Fused boric oxide is almost as 
 difficult to handle as some of the hardest minerals. 
 
 9. It is evident that the methods given are not applicable 
 to the determination of silica in minerals which also contain 
 fluorine. For a method consult Bulletin 700, U. S. Geological 
 Survey. 
 
 DETERMINATION OF POTASSIUM AND SODIUM IN 
 SILICATES 
 
 METHOD OF J. LAWRENCE SMITH 
 
 Weigh into a clean, dry, porcelain mortar 0.5-1.0 gram 
 of finely ground feldspar. Weigh an equal quantity of dry 
 ammonium chloride into the mortar. Place the mortar on a 
 piece of glazed paper and grind together thoroughly with the 
 pestle until the feldspar and ammonium chloride are thoroughly 
 mixed. Weigh out eight parts of calcium carbonate, free from 
 alkali, and transfer gradually about three-quarters of this 
 amount to the mortar, mixing it with the feldspar. Transfer 
 the mixture to a platinum crucible, and use the remaining 
 carbonate to clean out the mortar, placing this on top of 
 the main portion, as a cover. Place the lid on the crucible, 
 and heat so gently for fifteen minutes that the ammonium 
 chloride is not visibly volatilized, but that the odor of ammo- 
 nia is distinctly perceptible. Then heat for forty-five min- 
 utes, so that the lower third of the crucible is red. The 
 proper degree of heat can be obtained if the crucible is placed
 
 10 QUANTITATIVE ANALYSIS 
 
 just over a flame which has been lowered to about three- 
 quarters of an inch in height. 
 
 When cool detach the sintered mass from the crucible 
 and place it in a casserole. Pour on it about 150 cc. of hot 
 water and allow it to stand until the mass has softened. 
 Then cautiously crush and grind until no lumps remain. Boil 
 the water for a few minutes, allow the residue to settle, and 
 decant through a filter into a large casserole. Repeat the 
 boiling and decantation, using about 50 cc. of water each 
 time until the wash water is free from chlorides, finally throw- 
 ing the insoluble residue onto the filter and washing with hot 
 water. As the process of washing free from chlorides is a 
 tedious one, it is advisable not to begin testing until at least 
 500 cc. of wash water have been used, and then by using very 
 small quantities and comparing with a blank containing the 
 same quantity of reagents. 
 
 Add to the filtrate a slight excess of ammonium carbonate 
 solution, a drop or two of ammonia, and then concentrate to 
 about 100 cc. 
 
 Filter the calcium carbonate on a small filter paper, and 
 receive the filtrate and washings in a platinum dish. Evapo- 
 rate the solution to dryness, and then cautiously expel the 
 ammonium salts, taking care not to allow the bottom of the 
 dish to become red. When cool dissolve in a small quantity 
 of water, add one or two drops of ammonia, and ammonium 
 oxalate drop by drop until no further precipitation takes place. 
 Digest on a water bath for a quarter of an hour, and again 
 filter through a small filter. Evaporate the filtrate to dryness, 
 and again cautiously ignite to remove ammonium salts. 
 
 The residue, which should be nearly white, is ready for 
 weighing after drying in a desiccator. It shouM consist of the 
 chlorides of sodium and potassium, unless lithium is present. 
 
 To separate the sodium and potassium proceed as follows: 
 Dissolve the weighed residue in a small quantity of water. 
 The solution should show no evidence of insoluble matter. 
 Add one or two drops of hydrochloric acid (sp. gr. 1.12) 
 and about one and one-half times the calculated amount of
 
 DETERMINATION OF POTASSIUM AND SODIUM II 
 
 hydrochlorplatinic acid, assuming the total weight of chlorides 
 to be sodium chloride. Evaporate the solution over a beaker 
 of hot water, taking care not to allow the water to boil. The 
 evaporation should continue until a crystallized crust has 
 formed over the surface and only i or 2 cc. of solution are 
 left. When removed from the water bath the whole mass 
 should be dry. Cover the residue with a liberal amount of 
 absolute alcohol, crush the salts with a glass rod, and stir at 
 frequent intervals for a half hour. Decant the liquid through 
 a weighed Gooch crucible, wash thoroughly with alcohol until 
 the washings run through colorless, and a drop evaporated on 
 a piece of clean platinum leaves no residue. Dry to constant 
 weight at 110. 
 
 The precipitate should consist of a heavy yellow powder 
 composed of octahedral crystals, having the composition rep- 
 resented by the formula K 2 PtCl 6 . It should be entirely free 
 from the orange-colored sodium salt NasPtCle^H^O, and from 
 crystals of sodium chloride. 
 
 Notes. i. The reaction upon which the method is founded 
 may be represented as follows: 
 
 aKAlSisOs + 6CaCO 3 + 2NH 4 C1 = 
 
 6CaSiO 3 + 6CO 2 + A1 2 O 3 + 2KC1 + 2NH 3 + H 2 O. 
 
 In addition to this reaction there is also, of course, con- 
 siderable decomposition of the calcium carbonate to calcium 
 oxide and some formation of calcium chloride, both of. which 
 appear in the nitrate in the water extract. The completeness 
 of the reaction is dependent upon the fineness of the sample 
 and the care exercised in mixing the sample with ammonium 
 chloride. The heat necessary to complete the reaction is not 
 great, if sufficient care has been taken to have a finely ground 
 sample; but it is preferable, if possible, in order to ensure 
 complete decomposition, to use the long platinum crucible, 
 closed by a cap, as recommended by J. Lawrence Smith. 
 
 In case the silicate contains much ferrous iron, the mass, 
 instead of sintering, will fuse to a hard cake. In cases of 
 this kind it is necessary to increase the amount of calcium 
 carbonate. 
 
 Silicates may also be decomposed by means of sulphuric
 
 12 QUANTITATIVE ANALYSIS 
 
 and hydrofluoric acid, but this method has the disadvantage 
 that it is necessary to remove iron, aluminum, magnesium, 
 etc. Each of these substances may carry down more or less 
 sodium and potassium. In the J. Lawrence Smith method 
 these substances are removed in the beginning of the opera- 
 tion, being precipitated by the calcium hydroxide. 
 
 In case it is necessary to remove magnesium from a solution 
 in which sodium and potassium are to be determined, it is 
 undesirable to use the phosphate separation on account of 
 the difficulty of the removal of the excess of the reagent. 
 It is much better to convert the magnesium, sodium, and 
 potassium into chlorides, and then to heat with mercuric 
 oxide. 
 
 HgO + MgCl 2 = MgO + HgCl 2 . 
 
 The excess of mercuric oxide and mercuric chloride can 
 be volatilized, and the potassium and sodium chlorides can be 
 separated from magnesium oxide by means of water. 
 
 The alkalies may also be determined when the silicate is 
 decomposed by means of lead oxide, boric acid, bismuth 
 oxide, etc. 
 
 2. One of the defects of the method is the fact that the 
 calcium carbonate invariably contains small amounts of 
 alkali, principally the sodium salt; but by making a blank 
 determination this may be corrected. The amount of im- 
 purity should be placed on the label of the bottle of cal- 
 cium carbonate, and this sample should be preserved for this 
 determination only. The ammonium chloride may be purified 
 by sublimation. 
 
 3. In addition to the removal of the calcium it may also 
 be necessary, in case the rock contained sulphur, to remove 
 small amounts of sulphate. This would cause a contamina- 
 tion of the potassium chlorplatinate on account of the insolu- 
 bility of sodium sulphate in alcohol. Its removal may be 
 accomplished by the addition of a drop or two of barium 
 chloride, and the removal of the excess of the latter by means 
 of ammonium carbonate. 
 
 The removal of the calcium by means of ammonium car- 
 bonate has the distinct advantage that the products of its 
 hydrolysis 
 
 (NH4) 2 C0 3 + 2H 2 ^ 2NH40H + H 2 O
 
 DETERMINATION OF POTASSIUM AND SODIUM 13 
 
 are more or less completely removed in the process of evapo- 
 ration. With ammonium oxalate, however, there is more or 
 less accumulation of oxalic acid, which upon heating gives 
 considerable carbonaceous material which is difficult to remove 
 from the solution. 
 
 4. The separation of potassium from sodium is based on 
 the fact that potassium chlorplatinate is insoluble in alcohol, 
 while the sodium chlorplatinate and excess hydrochlorplatinic 
 acid are soluble. As there is some danger of hydrolysis 
 of the potassium chlorplatinate, according to the reaction, 
 
 K 2 PtCl 6 + H 2 ^ K 2 PtCl 5 OH + HC1, 
 
 it is necessary to make the separation in the presence of a 
 drop or two of hydrochloric acid to prevent this. After the 
 addition of hydrochlorplatinic acid it is advisable to protect 
 the solution from dust, which would cause a partial reduction 
 to metallic platinum. To confirm the result the potassium 
 chlorplatinate may be reduced by heating with dry oxalic 
 acid, and the metallic platinum may be weighed, or the 
 chlorine in the potassium chloride may be determined. The 
 reaction taking place is shown in the equation, 
 K 2 PtCl 6 + (COOH) 2 = 2KC1 + Pt + 
 2HC1 + 2C0 2 + C1 2 . 
 
 This method may also be applied to the nitrate containing the 
 sodium, finally converting the sodium into sodium sulphate. 
 
 5. Instead of weighing the potassium chlorplatinate on a 
 Gooch crucible it may be filtered on paper, washed in the 
 usual way with alcohol, and after the washing is complete 
 it may be dissolved in hot water and the solution allowed 
 to run into a weighed platinum dish. After evaporation to 
 dryness this may be weighed. 
 
 In all cases precaution must be taken to protect the solu- 
 tion containing the hydrochlorplatinic acid from ammonia. 
 Otherwise there will be formation of ammonium chlorplatinate. 
 
 THE PERCHLORATE METHOD FOR POTASSIUM 
 
 After weighing the chlorides of sodium and potassium, 
 obtained as in the preceding process, finish the separation as 
 follows: Dissolve the mixed chlorides in 20 cc. of water, and
 
 14 QUANTITATIVE ANALYSIS 
 
 add sufficient 20 per cent solution of perchloric acid to convert 
 both sodium and potassium into the perchlorates and to have 
 a slight excess. Evaporate to dryness. Dissolve in 10 cc. 
 of water and add a small amount of perchloric acid. Evapo- 
 rate again to dryness and repeat this process until dense white 
 fumes of perchloric acid appear. Allow to cool, and take up 
 with 25-30 cc. of 97 per cent alcohol, containing 0.20 per cent 
 perchloric acid (i cc. of 20 per cent solution to 100 cc. of 97 
 per cent alcohol). Break up the residue with a stirring rod, 
 decant the clear liquid through a Gooch crucible containing an 
 asbestos mat which has been washed with 0.20 per cent per- 
 chloric acid in alcohol. Wash several times by decantation 
 with alcoholic perchloric acid solution, transfer the residue to 
 the filter, and wash several times on the filter. Dry the 
 crucible at 120 for one hour, cool, and weigh as KC1O 4 . 
 
 Notes. i. Potassium perchlorate is insoluble in 97 per 
 cent alcohol containing perchloric acid, and the sodium salt 
 is soluble. This method gives satisfactory results, and has 
 the distinct advantage over the hydrochlorplatinic acid 
 method that it is very much cheaper. 
 
 The method has been studied by Wense, 1 Caspari, 2 and 
 Baxter 3 and has generally been accepted as accurate, having 
 been adopted as an official method 4 by the fertilizer chemists 
 of Germany. Scholl 5 has applied the method to the deter- 
 mination of potassium in water analysis. 
 
 2. The method is not accurate in the presence of sulphate 
 and ammonium ions. The former can be removed by means 
 of barium chloride and the latter by heating. 
 
 3. The perchloric acid solution containing alcohol should 
 never be evaporated. 
 
 Dangerous explosions may result. There is, however, no 
 danger at the ordinary temperature. 
 
 1 Z. angew. Chem., 4, 691 (1891); 5, 233 (1892). 
 
 2 Z. angew. Chem., 6, 68 (1893). 
 
 3 J. Am. Chem. Soc., 39, 249. 
 
 4 Fifth Internal. Congr. of Appl. Chem., I, 216 (1903); 4, 940 (1903). 
 6 /. Am. Chem. Soc., 36, 2085.
 
 ANALYSIS OF SPATHIC IRON ORE 15 
 
 4. The precipitate is much less soluble in alcohol containing 
 a small amount of perchloric acid than in alcohol alone. 
 
 ANALYSIS OF SPATHIC IRON ORE 
 DETERMINATION OF SiO 2 , MnO, CaO, AND MgO 
 
 Weigh into No. i lipped beakers two portions of the ore 
 of approximately 0.5 gram each, and dissolve in 15 cc. hydro- 
 chloric acid (sp. gr. 1.12), heating to aid the solution. Filter 
 off the insoluble residue, and wash first with dilute hydro- 
 chloric acid and afterward with hot water. Ignite the paper 
 and residue at a low temperature, and weigh approximately 
 within 5 milligrams. Cover with six times the weight of 
 sodium carbonate and fuse until action ceases. Add a few 
 cubic centimeters of water to the crucible and warm, then 
 wash the contents of the crucible into the original acid solu- 
 tion. Evaporate to dryness on the steam table, moisten 
 the dried mass with i or 2 cc. of hydrochloric acid (sp. gr. 1.20), 
 and again evaporate to dryness. Place in a hot closet and 
 keep at a temperature of no C. for one-half hour to de- 
 hydrate the silica. 
 
 Moisten the residue with 5 cc. of strong hydrochloric acid 
 and warm. Then add 50 cc. of water and heat until solution 
 is complete. Filter immediately, wash the silica with dilute 
 hydrochloric acid (one part acid, sp. gr. 1.12, to three parts 
 of water) until the filter shows no traces of iron, and then 
 finally with hot water. Evaporate the filtrate for the recov- 
 ery of traces of silica, as was done in the silicate analysis. 
 Ignite both filters in a platinum crucible and weigh. Treat 
 the ignited silica with one or two drops of water, enough 
 hydrofluoric acid to dissolve it, and two drops of sulphuric 
 acid (i : i). Evaporate to dryness in a hydrofluoric acid 
 burner, ignite, and weigh again. The difference in weight 
 represents the silica. The weight of the residue after the hy- 
 drofluoric acid treatment should not be more than i or 2 milli- 
 grams at the most. 
 
 Concentrate the filtrate from the silica, adding bromine
 
 1 6 QUANTITATIVE ANALYSIS 
 
 water to oxidize the iron to the ferric state. Evaporate 
 almost to dryness, or until the ferric chloride will completely 
 crystallize when allowed to cool. This operation must be 
 carried on with caution. It is necessary to remove the excess 
 of hydrochloric acid, but if the evaporation is allowed to go 
 too far the iron salts will become baked on the bottom of the 
 vessel, and it will then be necessary to dissolve in acid and 
 re-evaporate. When cold add to the ferric chloride 20 cc. 
 of water; solution should be complete. If any silica has 
 escaped dehydration it will show itself at this point, and it 
 will be necessary to filter it off and repeat the evaporation of 
 the filtrate. Transfer the solution to a small beaker and add, 
 from a burette or dropper, sodium carbonate solution until a 
 precipitate forms which will dissolve only in two or three 
 minutes stirring. This point is indicated by the solution 
 assuming an intense red color, and by the ferric hydroxide 
 going more and more slowly into solution. If an excess of 
 sodium carbonate should have been added, it is necessary to 
 add very dilute hydrochloric acid drop by drop, with constant 
 stirring, until the solution clears again ; then to bring the solu- 
 tion back to the proper point with sodium carbonate. Having 
 obtained a solution properly neutralized, add 10 cc. of acetic 
 acid (sp. gr. 1.04). 
 
 Meanwhile heat to boiling one liter of distilled water for 
 each determination in actual progress, and weigh out as many 
 portions of 3 grams each of sodium acetate. Transfer the 
 solution of the iron ore, together with the liter of hot water, 
 to a 1500 cc. round-bottom flask, and add one portion of the 
 sodium acetate dissolved in a little hot water. Bring the so- 
 lution to boiling and continue the boiling for three minutes. 
 If there is any tendency to froth this may be overcome by 
 pricking the bubbles as they form in the neck of the flask 
 with a long, glass stirring rod. Remove the flask from the 
 lamp and allow the precipitate to settle. The supernatant 
 liquid should be colorless and clear. 
 
 Filter, while still hot, through a 24 cm. fluted paper. Wash 
 once by decantation with 250 cc. of hot water. Dissolve
 
 ANALYSIS OF SPATHIC IRON ORE 17 
 
 the precipitate in warm, dilute hydrochloric acid (sp. gr. 
 1.12), and wash the filter paper until free from iron. Evapo- 
 rate, with the addition of a little bromine water, until a crust 
 forms, and repeat the basic acetate separation as described 
 above, this time washing the whole of the precipitate onto 
 the filter paper and adding this filtrate to the first one. 
 
 The combined filtrates from the two basic acetate separa- 
 tions are evaporated to about 400 cc., and bromine water is 
 added in excess. The solution should be kept hot for an 
 hour, and at the end of this time there still should be a slight 
 odor of bromine. Boil off the excess of bromine and then 
 filter the manganese dioxide, washing with hot water. Dis- 
 solve any of the precipitate which may have adhered to the 
 beaker in warm, dilute hydrochloric acid to which a few cubic 
 centimeters of sulphurous acid have been added. Pour this 
 solution through the paper holding the manganese dioxide, 
 and wash the paper with hot water until entirely free from 
 chlorides. Dilute the filtrate to 100 cc., add 10 cc. of di- 
 sodium phosphate, 10 cc. of ammonium chloride, and just 
 enough ammonia to produce a permanent precipitate. Bring 
 the solution to boiling, and boil until the small amount of 
 precipitate appears to have become crystalline. Then add 
 ammonia slowly until it is present in slight excess, and continue 
 to boil, with constant stirring, until the whole of the manga- 
 nous ammonium phosphate has become silky and crystalline. 
 Allow to stand until cold, filter, and wash with a solution of 
 ammonium nitrate (10 grams NHiNO 3 in 100 cc. H 2 O) which 
 has been made slightly alkaline with ammonia. Dry the 
 precipitate and paper, place in a platinum crucible, and ignite 
 at as low a temperature as possible, taking the same pre- 
 cautions as is usual with the corresponding magnesium 
 ammonium phosphate. The paper should be charred at as 
 low a temperature as possible, and under no circumstances 
 should it be allowed to burn with flame. After it is com- 
 pletely charred the bottom part of the inclined crucible is 
 heated in the oxidizing flame of a Tirrill burner until the 
 precipitate is white.
 
 1 8 QUANTITATIVE ANALYSIS 
 
 In the filtrate from the manganese dioxide boil off the 
 bromine, evaporate to about 400 cc., then add to the boiling 
 solution ammonia until slightly alkaline, and ammonium oxa- 
 late drop by drop until all of the calcium has been precipi- 
 tated, finally adding 5-10 cc. in excess. Boil for several 
 minutes, and then allow to stand in a warm place until the 
 precipitated calcium oxalate has completely settled. Decant 
 the solution through a filter paper hi such a way that most 
 of the precipitate remains hi the beaker. Add a few cubic 
 centimeters of ammonium oxalate to the filtrate to see if 
 precipitation has been complete. Now dissolve the calcium 
 oxalate in the beaker with hydrochloric acid, and pour the 
 solution through the filter, receiving it in a No. 4 beaker 
 and washing with hot water until the washings amount to 
 250 cc. Heat the solution to boiling, make slightly am- 
 moniacal, add 5 cc. of ammonium oxalate, and allow to stand 
 until the precipitate has completely settled. Pour the clear 
 supernatant liquid through a filter, wash the precipitate twice 
 by decantation with hot water, and then transfer the whole 
 precipitate to the filter and wash until the washings show 
 no test for chloride. Dry the precipitate and paper, place 
 in a platinum crucible, and smoke off the paper with a low 
 flame. After the carbon has all been burned, heat the 
 covered crucible with the full flame of a Tirrill burner and 
 finally over the blast lamp to constant weight. 
 
 The combined filtrates from the calcium precipitation are 
 evaporated to a volume of 400 cc. If during this evaporation 
 there should be a separation of a white precipitate it will be 
 necessary to evaporate to dryness with concentrated hydro- 
 chloric acid. If, however, there is no precipitation add 
 ammonia in excess and di-sodium phosphate, with constant 
 stirring, until precipitation is complete. Allow to stand in 
 the cold (ice water) for several hours, decant the solution 
 through a filter, and wash the precipitate onto the filter 
 with dilute ammonia (three parts H 2 O, one part NE^OH, 
 sp. gr. 0.96), washing several times with this wash water. 
 Now dissolve the precipitate in a small amount of dilute
 
 ANALYSIS OF SPATHIC IRON ORE 19 
 
 hydrochloric acid, and wash the paper several times with 
 water. To this solution add 5 cc. di-sodium phosphate, and 
 then neutralize with ammonia, adding it slowly, with stirring, 
 until present in very slight excess. After standing 10 minutes, 
 add 10 cc. strong ammonia, sp. gr. 0.90. Allow to stand 
 several hours in ice water, filter, wash with dilute ammonia, 
 and continue until a few drops of the washing, acidified with 
 nitric acid, shows no turbidity with silver nitrate. Dry the 
 precipitate and paper, and smoke off the filter paper at as low 
 a temperature as possible, being careful not to allow the paper 
 to burn with flame. When charred raise the temperature of 
 heating to the full flame of the Tirrill burner and continue 
 until the precipitate is perfectly white. 
 
 DETERMINATION OF COMBINED OXIDES, Fe 2 3 , A1 2 3 , P 2 O 5 
 
 For the determination of ferric oxide, alumina, and phos- 
 phoric anhydride weigh two portions of approximately 0.2 
 gram each and proceed exactly as in the determination of 
 manganese, calcium, and magnesium, except that the silica 
 which is removed must be treated with hydrofluoric acid, and 
 the residue obtained by this operation must be added to the 
 solution in which iron and alumina are to be determined. 
 To do this, fuse with a small amount of sodium carbonate and 
 dissolve in a small amount of hydrochloric acid. 
 
 Instead of making a double basic acetate separation, as 
 was done in the previous case, the second precipitation is 
 made by means of ammonia, and the filtration is made through 
 a closely fitting paper. In washing, care should be taken to 
 thoroughly wash the top of the filter. The filtrate from the 
 basic acetate separation should always be evaporated and 
 examined for iron. 
 
 Dissolve the basic acetate precipitate on the filter, and 
 also any which may have adhered to the flask, in hot, dilute 
 nitric acid (sp. gr. 1.20) and wash the paper free from iron. 
 While still warm add ammonia in slight excess, allow the 
 precipitate to settle, decant through a filter paper, wash 
 twice by decantation with hot water, and then transfer the
 
 20 QUANTITATIVE ANALYSIS 
 
 precipitate to the paper and wash several times on the paper 
 with hot, dilute ammonium nitrate solution. Place the 
 washed and dried precipitate in a weighed platinum crucible 
 and smoke off the filter paper at a low temperature. This is 
 best accomplished by inclining the crucible on a triangle and 
 placing a low flame under the front end of it. The flame 
 should be so low that the volatile products will smoke off, 
 but will not ignite. When the paper is completely charred, 
 ignite over the full flame for fifteen minutes and then over 
 the blast lamp for five minutes. During this last heating the 
 crucible remains in an inclined position, with the cover par- 
 tially covering the mouth, but the oxidizing flame is directed 
 toward the bottom of the crucible. Heat to constant weight. 
 The oxides thus obtained will always contain traces of 
 silica and should be corrected for it. This is done by fusing 
 the oxides with twelve times their weight of acid potassium 
 sulphate until the fusion becomes perfectly clear. After the 
 fusion has become cold it is dissolved in water and the silica 
 is filtered off, washed, ignited, and weighed. This weight 
 subtracted from the weight obtained above gives the true 
 weight of the combined oxides. 
 
 DETERMINATION OF IRON IN SPATHIC IRON ORE 
 ZIMMERMANN-REINHARDT METHOD 
 
 Weigh three portions of thoroughly cleaned iron wire, from 
 0.15-0.25 gram, into 150 cc. Erlenmeyer flasks, and dissolve 
 in 20 cc. of hydrochloric acid (sp. gr. 1.12) and 20 cc. of water. 
 When the solution is complete, add strong potassium perman- 
 ganate solution drop by drop until the carbonaceous material 
 is completely oxidized and the solution has shown a very 
 marked change of color. Place the flasks on a piece of white 
 paper, and reduce while still hot with the smallest possible 
 volume of freshly-prepared stannous chloride which will pro- 
 duce a colorless solution. If too much has been added it may 
 be corrected by adding potassium permanganate solution until 
 a slight yellow coloration appears, and then adding just enough 
 stannous chloride to reduce the small amount of ferric chloride
 
 ANALYSIS OF SPATHIC IRON ORE 21 
 
 present. Cool the solution by holding the flask in running 
 water, and when cold add rapidly 10 cc. of mercuric chloride 
 (50 grams per liter) . This should produce a very small amount 
 of white, silky mercurous chloride. If there is a large amount 
 formed, or if it is at all dark-colored or flocculent, it is not 
 advisable to try to finish the titration. Wash this solution 
 into 400 cc. of cold water to which has been added 20 cc. of 
 manganous sulphate solution (20 grams MnSO 4 , 220 cc. H 2 O, 
 40 cc. H 2 SO 4 sp. gr. 1.84), and titrate to pink color with 
 potassium permanganate (3 grams per liter) solution which 
 has been filtered through asbestos to remove manganese 
 dioxide. 
 
 Potassium permanganate which is to be used in this as well 
 as in other determinations may be standardized accurately 
 against sodium oxalate as follows: Dissolve 0.25-0.30 gram of 
 sodium oxalate in 250 cc. of hot water in a 400 cc. beaker, and 
 add 10 cc. of (i : i) sulphuric acid. Titrate with permanga- 
 nate, stir vigorously, and add the last 0.5-1.0 cc. drop by drop, 
 allowing each drop to be decolorized before adding another. 
 Estimate the excess of permanganate necessary to color the 
 solution by comparing with a similar volume of water and 
 acid. The temperature of the solution must be maintained 
 throughout the titration between 70-90 C. 
 
 To determine iron in the spathic iron ore weigh two portions 
 of approximately 0.3 gram each into porcelain crucibles. 
 Roast the ore for ten minutes to destroy organic matter, 
 and then dissolve in a small quantity, 15-20 cc., of hydrochloric 
 acid. If there is an insoluble residue it will be necessary to 
 fuse it with a very small quantity of sodium carbonate, take it 
 up in hydrochloric acid, and add it to the main portion. When 
 the solution is complete, reduce with stannous chloride, and 
 then proceed as in the standardization of potassium perman- 
 ganate against iron wire. 
 
 Notes. i. Spathic iron ore is essentially ferrous carbon- 
 ate, but there is always found with it varying amounts of 
 organic matter, silica, alumina, manganese, calcium, mag- 
 nesium, phosphorus, and sulphur, with traces of titanium.
 
 22 QUANTITATIVE ANALYSIS 
 
 There is invariably an insoluble residue left after treating with 
 hydrochloric acid, which may contain some or all of these 
 elements. After fusion of the residue with sodium carbonate, 
 solution in hydrochloric acid takes place easily. 
 
 A definite amount of sodium carbonate is prescribed, so as 
 to prevent the introduction of an unnecessarily large excess 
 of sodium salts, as would probably be the case if the amount 
 were left to the judgment of the worker. 
 
 Four portions are weighed for analysis, because if large 
 enough portions were taken to determine accurately the 
 manganese, calcium, and magnesium there would be difficulty 
 in handling the very bulky precipitate of iron and alumina; 
 consequently smaller portions have been taken for the separate 
 determination of these constituents. 
 
 2. The double evaporation of the solution with hydro- 
 chloric acid is necessary in order to dehydrate completely 
 all of the silica. This was shown to be true by Hillebrand 
 in the case of silicates, and is even more true here on account 
 of the solvent effect of ferric chloride. This solvent effect is 
 very noticeable if the hydrochloric acid solution is allowed 
 to stand any length of time before filtration. 
 
 It is necessary to moisten the dry residue with hydrochloric 
 acid before adding water, because of the strong tendency of 
 ferric and aluminum salts to hydrolyze and form basic salts, 
 which would be filtered off and weighed with the silica. 
 Even when the procedure is carefully followed there is always 
 a slight residue left after the hydrofluoric acid treatment. 
 This residue contains some iron and alumina, and traces of ti- 
 tanium, the latter often being in combination with phosphoric 
 acid. 
 
 The treatment with hydrofluoric acid is based on the 
 reaction, 
 
 SiO 2 + 4HF = SiF 4 + 2H 2 0, 
 
 but as there is always the tendency for the silicon tetrafluo- 
 ride to react with water according to this reaction, 
 
 3 SiF 4 -f 4H 2 O = 2H 2 SiF 6 + Si(OH) 4 , 
 
 it is necessary to have some sulphuric acid present to remove 
 the water as fast as it is formed. Furthermore, certain other
 
 ANALYSIS OF SPATHIC IRON ORE 23 
 
 fluorides, such as those of iron and titanium, are volatilized 
 in the absence of sulphuric acid. 
 
 3. The oxidation of the iron to the ferric state is abso- 
 lutely necessary to the success of the operation. The 
 presence of ferrous iron would not only render the separation 
 incomplete, but it may also cause the formation of the so- 
 called "brick-dust" precipitate, which contains ferrous iron 
 and is extremely difficult to handle. Other oxidizing agents 
 than bromine may be used, but this has been selected, not 
 only on account of its efficiency, but also on account of the 
 ease with which the excess may be removed. If nitric acid 
 is used as the oxidizing agent there sometimes forms, on 
 evaporation, a dark-colored gummy mass, which can only be 
 converted into the crystallized ferric chloride by repeated 
 evaporation with concentrated hydrochloric acid. 
 
 The careful evaporation of the excess of hydrochloric acid 
 is to remove the excess of oxidizing agent and to prevent an 
 unnecessary addition of a large quantity of sodium carbonate 
 when neutralizing the ferric chloride solution. 
 
 4. Ferric hydroxide possesses, to a very high degree, the 
 property of forming a colloidal solution. When sodium car- 
 bonate solution is added to the ferric chloride the ferric 
 hydroxide which is first formed gradually goes into solution 
 in the colloidal state, the increasing amount present being 
 indicated by the solution assuming the deep-red color charac- 
 teristic of the undissociated iron salts. As the colloidal con- 
 dition is easily destroyed by heat and the presence of an 
 electrolyte, it is essential to keep the solution cool until ready 
 for use, and under no circumstances should the sodium acetate 
 be added to it before dilution. 
 
 The older methods for the separation of iron from man- 
 ganese were based on the decomposition of this colloidal 
 solution by means of heat in the presence of an electrolyte 
 (Na 2 SO4), which caused a separation of the colloidal ferric 
 hydroxide and the formation of a basic iron salt. 
 
 5. The separation of iron and aluminum from the bivalent 
 metals is based upon the relative ease with which the differ- 
 ent acetates hydrolyze. The hydrolysis of ferric acetate may 
 be expressed in this way: 
 
 Fe(CH3C0 2 ) 3 + 2 H 2 ^ Fe(OH) 2 .CH3CO 2
 
 24 QUANTITATIVE ANALYSIS 
 
 Since hydrolysis is dependent upon the dissociation of 
 water into hydroxyl and hydrogen ions, and as the dissocia- 
 tion is greater at the boiling temperature, the hydrolysis of 
 ferric and aluminum acetates takes place more rapidly and 
 completely at this temperature. Manganese acetate would 
 also hydrolyze somewhat under the same conditions, espe- 
 cially in neutral solution, and it is therefore necessary to 
 diminish the dissociation of water to such an extent that the 
 ferric and aluminum acetates may hydrolyze, but that the 
 manganese acetate may not. This is accomplished by 
 the addition of the w r eak acetic acid; and in order that it, 
 and also the acetic acid formed in the reaction, may not 
 furnish too great a concentration of hydrogen ions, and thus 
 force the water completely back into the undissociated state, 
 the concentration of hydrogen ions is diminished by the addi- 
 tion of a salt containing an ion in common, viz., sodium 
 acetate. Consequently the strength of the acid is weakened. 
 
 If too much acetic acid is used the concentration of the 
 hydroxyl ion is diminished to such an extent that the ferric 
 and aluminum acetates cannot hydrolyze completely. On 
 the other hand, if too little acetic acid or, what is equivalent 
 to the same thing, too much sodium acetate has been used, 
 the concentration of the hydroxyl ion may be so great that the 
 manganese acetate will also hydrolyze and be precipitated 
 with the iron and alumina. 
 
 As the reaction involving the hydrolysis is a reversible one, 
 it is evident that in order to force the reaction completely 
 in one direction it is necessary to make the nitration while 
 the solution is still hot. Otherwise some of the precipitate 
 redissolves on cooling. 
 
 6. If the process has been properly carried out the nitrate 
 should be clear and colorless, and free from iron. On evapo- 
 ration any iron which was not precipitated will separate. 
 This, however, should not be confused with the hydrated 
 manganese dioxide, some of which invariably separates 
 during the process of evaporation, the manganous acetate 
 being oxidized by the air. 
 
 The best method of deciding whether this precipitate is 
 iron or manganese is to filter it off, dissolve in a small 
 amount of hydrochloric acid, and to make a basic acetate
 
 ANALYSIS OF SPATHIC IRON ORE 25 
 
 separation in small volume. If ammonia and ammonium 
 chloride are used the separation may not only be incomplete, 
 but there may be subsequent difficulty in precipitating the 
 manganese in the main nitrate on account of the presence 
 of ammonium salts. 
 
 7. The basic acetate process is the most accurate method 
 for the separation of iron from manganese. For precise work 
 this method should always be used, in spite of some of its 
 disadvantages. It is also used for the separation of alu- 
 minum from the bivalent metals, but for the separation of 
 chromium from the bivalent metals the barium carbonate 
 process is to be preferred. Chromium acetate is incom- 
 pletely hydrolyzed except in the presence of a very large 
 excess of iron. 
 
 In any case where this separation is used phosphoric acid 
 and titanium hydroxide will be precipitated with the trivalent 
 elements. 
 
 If it is desired to determine the iron in this precipitate 
 instead of making a determination in a separate sample, it 
 may be done by the process described on page 20. 
 
 8. The double separation by means of the basic acetate 
 process is necessary when manganese is to be determined, 
 as there is always some manganese carried down in the first 
 precipitation. This may be partially due to occlusion, but 
 it is also undoubtedly true that in this weakly acid solution 
 some of the manganese is oxidized to the dioxide and carried 
 down in this form. 
 
 9. When only iron and aluminum are to be determined, 
 it is believed that a single basic acetate precipitation, followed 
 by a separation with ammonium hydroxide, suffices to give a 
 pure precipitate. The solution in nitric acid must take place 
 before the precipitate has had time to dry. If this precaution 
 is neglected it requires considerably more hot nitric acid to 
 dissolve the precipitate, and in this process the filter paper 
 is more or less attacked. The solution will, therefore, con- 
 tain enough organic matter to hold the iron and aluminum in 
 solution when ammonia is added. If the recently precipi- 
 tated hydroxides are dissolved in nitric acid there will be no 
 difficulty in making the ammonia precipitation. 
 
 The use of nitric acid has the distinct advantage that it
 
 26 QUANTITATIVE ANALYSIS 
 
 does away with the tedious washing to remove chlorides; 
 and furthermore, it is easier to keep the iron oxidized to its 
 highest state during the ignition. 
 
 Although some of the aluminum hydroxide is probably 
 dissolved in the colloidal form when the final washing is 
 made with pure water, it is believed that this may be largely 
 overcome by making the final washing with dilute ammonium 
 nitrate solution. 
 
 10. Bromine separates manganese dioxide from the ace- 
 tate solution according to this reaction: 
 
 It would seem much simpler to ignite the manganese dioxide 
 and weigh the oxide MrisO^ but this process is open to two 
 serious objections. In the first place, manganese dioxide, 
 when precipitated from solutions containing sodium salts, 
 occludes considerable quantities of sodium; then the com- 
 position of the oxide formed on ignition is not always 
 represented by the formula MngCX Consequently it is 
 very much more accurate to use the excellent process de- 
 vised by Gibbs, 1 involving the separation of the manganese 
 as manganese ammonium phosphate, MnNH4PO 4 , and its con- 
 version by heat into the pyrophosphate Mn 2 P 2 Oy. Gooch 
 and Austen 2 have carefully studied the best conditions for 
 obtaining a pure precipitate. The precipitate which first 
 forms is the tri-manganic phosphate Mns(P04) 2 , but this is 
 converted by heating with an excess of the precipitant and 
 ammonium chloride into the crystalline manganese ammonium 
 phosphate. 
 
 The same care which is recommended in Note 13 for the 
 ignition of magnesium ammonium phosphate should be used 
 in this case. If these precautions are neglected there will 
 be some reduction of the precipitate and consequent alloying 
 of the crucible. 
 
 ii. The double precipitation of the calcium oxalate is 
 absolutely necessary in order to purify it from other salts 
 which it carries down with it. If there is considerable 
 aluminum in the ore it is necessary to ignite the first calcium 
 oxalate precipitate, dissolve the oxides in hydrochloric acid, 
 and to add ammonia in very slight excess to the solution, 
 
 1 Am. J. Sci. [2], 44, 216. * Am. J. Set. [4], 6, 233.
 
 ANALYSIS OF SPATHIC IRON ORE 27 
 
 in order to throw out traces of aluminum hydroxide, which 
 has reached this point in the analysis on account of its having 
 been dissolved in the colloidal form when washing the basic 
 acetate precipitate with water. 
 
 Richards 1 has shown that magnesium oxalate is carried 
 down with calcium oxalate, and that the amount of magnesium 
 oxalate occluded is proportional to the amount of undis- 
 sociated magnesium salt in solution. A double precipitation is 
 therefore always, necessary, not only to get a pure precipitate of 
 calcium, but in order to get all of the magnesium into the 
 filtrate. 
 
 It is also well known that sodium salts are occluded by 
 calcium oxalate. 
 
 12. The first precipitation of magnesium is made simply 
 with the idea of removing it completely from the rest of the 
 solution rather than to get a normal precipitate of the com- 
 position represented by the formula MgNHjPO^ Neubauer z 
 has shown that complete precipitation takes place, even in 
 the presence of a large amount of ammonium salts. He has 
 shown, and Gooch and Austen 3 have independently confirmed 
 his results, that a normal precipitate can only be obtained 
 in the absence of ammonium salts and ammonia. To ensure 
 a pure precipitate, the solution must be as nearly neutral as 
 possible, as free as possible from ammonium salts, and 
 ammonia must be added after the addition of the phosphate 
 solution. The large excess of ammonia, as prescribed in 
 most books, is distinctly disadvantageous, except in the 
 presence of an excess of ammonium oxalate. Reidenbach 4 has 
 shown it to be necessary in this case. If, therefore, the first 
 precipitation is made in the presence of an excess of ammonia, 
 the second precipitation can be made under proper conditions. 
 
 13. The ignition of magnesium ammonium phosphate 
 must be done with extreme caution. If the initial heat is 
 applied too rapidly the ammonia, which is one of the products 
 of the reaction, 
 
 2MgNH4P0 4 = MgzPsOi + 2NH 3 + H 2 O, 
 is decomposed, and the hydrogen then attacks the phosphate, 
 reducing it and causing the phosphorus to alloy with the 
 
 1 Richards, McCaffrey and Parker: Z. anorg. Chcm., 28, 71. 
 
 2 Z. angew. Chem. 1896, 435. 3 Am. J. Sci. [3], 7, 187. 
 4 Inaug. Dissertation, Munich, 1910.
 
 28 QUANTITATIVE ANALYSIS 
 
 platinum. Heraeus, 1 who has studied this, has shown that 
 this decomposition and reduction take place below 900 C. 
 
 14. The Zimmermann-Reinhardt method for the determina- 
 tion of iron has a very distinct advantage because of the pos- 
 sibility of using hydrochloric acid, in which most of the ores 
 are soluble. The reactions upon which the determination is 
 based are as follows: 
 
 2FeCl 3 + SnCl 2 = 2FeCl 2 + SnCU. 
 
 SnCl 2 + 2HgCl 2 = SnCl 4 + 2HgCl. 
 2 KMn0 4 + ioFeC! 2 + i6HCl = 
 
 ioFeC! 3 -f 2MnCl 2 + 2KC1 + 8HO 2 . 
 
 These reactions are practically the same as in the bichromate 
 process for the determination of iron, but the use of potas- 
 sium permanganate obviates the use of an outside indicator. 
 
 By using proper care, just enough stannous chloride can 
 be put in to reduce the iron and to have a very slight excess. 
 If the operation is properly carried out, the precipitate of 
 mercurous chloride is very small in amount and is silky in 
 appearance. If platinum salts have been introduced into 
 the analysis, through an acid potassium sulphate fusion, or 
 through the solution in platinum of an ignited precipitate, 
 it is difficult, and sometimes impossible, to tell when enough 
 stannous chloride has been introduced into the solution. 
 When most of the ferric iron has been reduced, then the 
 platinum begins to be reduced to the platinous form, and, 
 this salt being highly colored, the solution, instead of be- 
 coming colorless, as would ordinarily be the case, assumes 
 the red tint of the platinum salt. 
 
 15. If too small an amount of hydrochloric acid has been 
 used the stannic chloride, which is one of the products of the 
 reaction, will hydrolyze and give products which interfere 
 with the sharpness of the end point. If too much hydro- 
 chloric acid is present it will react with the potassium per- 
 manganate, giving high results. Baxter and Frevert 2 have 
 shown that the high results are due to the formation of 
 hypochlorous acid, which escapes from the solution without 
 giving up its oxygen, thus causing an addition of a greater 
 amount of permanganate than would correspond to the actual 
 
 1 Z. angew. Chem., 15, 917. 
 
 2 Am. Chem. J., 34, 109.
 
 DETERMINATION OF SULPHUR IN PYRITE 29 
 
 amount of iron present. This reaction is completely cor- 
 rected by the addition of a manganous salt, as has been shown 
 by Kessler 1 and by Zimmermann. 2 The mechanism of this 
 reaction is not completely understood, but the evidence points 
 toward the formation of a manganic compound, which reacts 
 with the ferrous iron. In other words, the manganese salt 
 catalyzes the reaction between potassium permanganate and 
 ferrous chloride. On the other hand, in a series of interesting 
 experiments on the oxidation of iron, some evidence has been 
 found that iron is capable of forming a peroxide of the formula 
 Fe 2 0a, which reacts with the ferrous salt. This view, however, 
 would not explain the fact that a manganese salt will acceler- 
 ate the reaction between potassium permanganate and oxalic 
 acid, causing the reaction to take place in cold solution, while 
 iron salts have no effect upon this reaction. 3 
 
 DETERMINATION OF SULPHUR IN PYRITE 
 
 I. AVAILABLE SULPHUR LUNGE METHOD 
 
 Spread out on the bottom of a 200 cc. casserole 0.4 to 
 0.5 gram of pyrite which has been ground to extreme fine- 
 ness. Pour over this 20 cc. of a freshly made mixture of 
 three volumes nitric acid (sp. gr. 1.42) and one volume hydro- 
 chloric acid (sp. gr. 1.20). Cover the casserole with a watch 
 glass and warm gently until action commences; then remove 
 from the heat until action slackens. Start the action again 
 from time to time and remove from the heat when the oxides 
 of nitrogen begin to be evolved. In this way continue the 
 heating until all action has apparently ceased. Complete 
 disintegration should be accomplished in five to ten minutes. 
 If sulphur separates it is probable that the reaction was 
 allowed to proceed too violently, or that the pyrite was not 
 ground finely enough. In case sulphur has separated it is 
 preferable to start with a new sample rather than to try to 
 oxidize the separated sulphur by the addition of potassium 
 chlorate. 
 
 1 Z. anal. Chem., i, 329. 2 Berichte, 14, 779. 
 
 3 For discussion and literature upon this reaction consult Barneby, /. Am. 
 Chem. Soc., 36, 1429.
 
 30 QUANTITATIVE ANALYSIS 
 
 When disintegration is complete add 5 cc. of sodium car- 
 bonate solution, and evaporate to dryness on the steam 
 bath. Pour over the dry mass 5 cc. of hydrochloric acid 
 (sp. gr. i. 20) and again evaporate to dryness. Moisten the 
 residue with i cc. of concentrated hydrochloric acid, warm 
 gently for a moment, then dilute with 100 cc. of hot water. 
 Filter, wash several times with hot dilute hydrochloric acid, 
 and finally with hot water. 
 
 Pour the filtrate into a slight excess of ammonia and keep 
 the mixture hot for ten or fifteen minutes. Filter through 
 a 4-inch Swedish paper, wash first by decantation, and then, 
 throwing the precipitate onto the paper, wash thoroughly 
 with hot water. 
 
 If the filtrate and washings exceed 250 cc., acidify with 
 hydrochloric acid and concentrate to approximately this vol- 
 ume. To the boiling, slightly acid solution add a hot solu- 
 tion of 5 per cent barium chloride from a dropper, allowing 
 the drops to run down the inside of the beaker until a slight 
 excess has been added. Allow to stand in a warm place for 
 half an hour, filter, washing by decantation with hot water 
 until the washings are free from acid, and then complete the 
 washing on the filter until free from chlorides. 
 
 When the filter has drained as much as possible from 
 wash water, fold it over the barium sulphate, and place it 
 in the bottom of a weighed platinum crucible. Incline the 
 crucible on its side, with the lid partially covering the front, 
 and place under the front end a flame so small that it will 
 dry the paper so slowly that there is no danger of spatter- 
 ing, and that it will char the paper without having it take fire. 
 When the paper is completely charred, place under the back 
 of the crucible the full oxidizing flame of a Tirrill burner, 
 and heat thirty minutes. As some of the barium sulphate 
 is invariably reduced to sulphide, it is necessary to allow free 
 access of air into the crucible, and to continue the heating 
 until constant weight is obtained.
 
 DETERMINATION OF SULPHUR IN PYRITE 31 
 
 H. AVAILABLE SULPHUR METHOD OF ALLEN AND 
 BISHOP 1 
 
 About 0.5 gram of the dried ore is put in a dry 300 cc. 
 Pyrex beaker, 10 cc. of a mixture of 2 parts by volume liquid 
 bromine and 3 parts carbon tetrachloride are added and 
 the beaker is covered with a watch glass. After standing 
 15 minutes at room temperature with occasional gentle 
 shaking, 10 cc. nitric acid (sp. gr. 1.4) are added and the 
 mixture is allowed to stand fifteen minutes longer at room 
 temperature with occasional shaking. 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 volatilized. It is then placed within 
 the rings of the bath and the solution evaporated to dryness, 
 the cover glass being raised above the rim of the beaker by 
 means of riders made of bent glass rods. Ten cc. hydro- 
 chloric acid (sp. gr. 1.2) are next added, and, after shaking 
 to mix thoroughly, the solution is evaporated to dryness, 
 still keeping the beaker covered as in the former evaporation. 
 When completely dry the silica is dehydrated by heating in 
 an air bath at 100 C. for several hours. 
 
 One cc. of hydrochloric acid (sp. gr. 1.2) are now added to 
 the dehydrated mass, followed after five minutes by 50 cc. 
 of hot water. The cover, sides of beaker, and riders are care- 
 fully washed down with hot water and after the removal of the 
 riders the cover is replaced. The mixture is heated for five 
 minutes to insure complete solution of all sulphate. After 
 the solution has partially cooled by standing for about five 
 minutes, o.i gram of powdered aluminum is added and 
 the beaker gently stirred until the iron color has disappeared, 
 showing complete reduction. It is now advisable to cool 
 the solution to prevent possible mechanical loss of material 
 in the form of mist when filtering, due to the action of warm 
 hyd r ochloric acid on the excess of aluminum powder. After 
 cooling, the cover glass and sides of the beaker are washed 
 
 1 Eighth International Congress of Applied Chemistry, I, 33; /. Ind. Eng. 
 Chem.,8, 1167; /. Ind. Eng. Chem., 11, 45-
 
 32 QUANTITATIVE ANALYSIS 
 
 down and the solution is filtered through an n cm. filter paper 
 into a No. 7 beaker and the residue washed thoroughly with 
 hot water. To the filtrate are added 3 cc. more of hydrochloric 
 acid (sp. gr. 1.2) and sufficient cold water to dilute to 650 cc. 
 After stirring to mix well, the beaker is covered with a large 
 cover glass and the sulphate precipitated by adding through 
 a special form of " precipitating cup" (discharging at the rate 
 of 5 cc. per minute) 50 cc. of 5 per cent barium chloride solu- 
 tion. The solution is not stirred while the barium chloride 
 is being added but after all is in, the supernatant liquor is 
 mixed by gentle stirring. 
 
 After the barium sulphate has settled it is filtered through 
 a tared Gooch crucible, using suction. The Gooch crucible 
 used has a capacity of 35 cc., with perforated bottom attached, 
 and with a moderately thick asbestos mat. The clear liquid 
 is siphoned through the crucible, the precipitate is transferred 
 to the filter by a stream from a wash bottle, and the beaker 
 washed six times with cold water. The crucible is then 
 ignited slowly, placing it first on an asbestos board over a 
 flame for 25 minutes so as to expel the water gradually, or it 
 may be given this preliminary drying by placing it in the steam 
 oven for a few hours if more convenient. The crucible is 
 placed upon the lid of a platinum crucible and is then heated 
 with the full flame of a Tirrill burner for thirty minutes, cooled 
 in a desiccator, and weighed. 
 
 III. TOTAL SULPHUR FRESENIUS METHOD 
 
 Mix 0.4 to 0.5 gram finely ground pyrite with ten parts of a 
 mixture of four parts sodium carbonate and one part potassium 
 nitrate in a platinum crucible. 
 
 Cut an opening in an asbestos board (at least 4 inches 
 square) sufficiently large to allow the lower two-thirds of the 
 crucible to project below the board. Place the covered cru- 
 cible in this opening, heat gently at first, and finally raise 
 the temperature enough to bring the mass to fusion. Any 
 unnecessary increase in the heat should be avoided on account 
 of the action of the oxidizing mixture on the platinum. Keep
 
 DETERMINATION OF SULPHUR IN PYRITE 33 
 
 the mass in fusion fifteen minutes, stirring, if necessary, with 
 a stout platinum wire to prevent the fusion from caking on the 
 sides of the crucible. 
 
 When the action has apparently ceased, place the crucible 
 in a porcelain casserole, and extract the residue several times 
 with boiling water. Decant the solution through a filter 
 paper and then boil the residue with 20 cc. of a 10 per cent 
 solution of sodium carbonate. Finally, wash the residue 
 thoroughly with a i per cent solution of sodium carbonate. 
 
 Make the filtrate distinctly acid and add 5 cc. of hydro- 
 chloric acid (sp. gr. 1.20) in excess. Boil to expel the carbon 
 dioxide and evaporate to dryness. Moisten the residue with 
 5 cc. concentrated hydrochloric acid and again evaporate to 
 dryness. Then warm the residue gently with 2 cc. of hydro- 
 chloric acid (sp. gr. 1.20) and add 100 cc. of hot water; filter, 
 and wash the silica. 
 
 In the filtrate, which should amount to about 250 cc., 
 the sulphuric acid is precipitated in boiling solution by hot 
 barium chloride, as in the Lunge method, and the barium 
 sulphate should be filtered, washed, and ignited in the same 
 
 IV. TOTAL SULPHUR SODIUM PEROXIDE METHOD : 
 
 Mix thoroughly about 0.5 gram of finely ground pyrite with 
 5 grams sodium peroxide and 4 grams sodium carbonate in an 
 iron crucible. Place the crucible in an opening cut in an 
 asbestos board, heat gently for ten minutes and then raise 
 to the full heat of a Tirrill burner for 25 minutes. Allow the 
 crucible to cool, place in a casserole containing 150 cc. of water, 
 digest until the sodium salts are dissolved. Remove the 
 crucible, wash with hot water, and to the solution add 5 cc. 
 of hydrochloric acid (sp. gr. 1.20) which has been saturated 
 with bromine. Heat to boiling, filter the ferric hydroxide, 
 and wash free from sulphate. Acidify the filtrate with hydro- 
 chloric acid, evaporate to dryness to remove silica, take up in 
 
 1 Hempel, Z. anorg. Chem., 3, 193; Clark, /. Chem. Soc., 63, 1079; Glaser, 
 Chem. Zlg., 18, 1448.
 
 34 QUANTITATIVE ANALYSIS 
 
 2 cc. of hydrochloric acid (sp. gr. 1.20) and add 100 cc. of hot 
 water; filter and wash. 
 
 In the filtrate, precipitate the sulphate and proceed exactly 
 as in the Lunge method. 
 
 Notes. i. The very great commercial importance of 
 pyrite makes the accurate estimation of sulphur one of the 
 most important determinations in the whole field of analyt- 
 ical chemistry. For the determination of available and total 
 sulphur, the methods of Lunge and Fresenius, respectively, 
 are undoubtedly the most accurate. The Lunge method 
 determines only that portion of the total sulphur which is 
 available for the manufacture of sulphuric acid; the Frese- 
 nius method determines all of the sulphur irrespective of the 
 form in which it exists. Very many other methods have 
 been proposed, but none of them has met with such general 
 favor as these two, and they can at this time be considered 
 as standards for this determination. 
 
 2. The Fresenius method has the disadvantage that the 
 oxidizing flux attacks the platinum crucible badly and, further- 
 more, it is always necessary to completely destroy the nitrate 
 before precipitating the sulphate ion. It has been repeatedly 
 demonstrated that the sodium peroxide method gives entirely 
 satisfactory results, and it has the advantage of being more 
 rapid than the Fresenius method. Occasionally the water 
 extract of the fusion shows the presence of manganate or per- 
 manganate and this may be destroyed by the addition of a few 
 drops of alcohol. 
 
 3. If the reaction in the Lunge method is allowed to 
 begin too vigorously there is invariably a separation of sul- 
 phur, and, while it is possible to oxidize the separated sulphur, 
 it is only done at the expense of much time and after the 
 addition of considerable potassium chlorate. 
 
 The Lunge method cannot be used for the decomposition 
 of many of the native sulphides on account of the strong 
 tendency of some of them to liberate their sulphur in the 
 free state. By using pure fuming nitric acid, or by reversing 
 the proportions of nitric and hydrochloric acids, the decom- 
 position and oxidation of most sulphides may be accomplished 
 successfully.
 
 DETERMINATION OF SULPHUR IN PYRITE 35 
 
 The Allen and Bishop method has the advantage over the 
 Lunge method in that it efficiently takes care of separated 
 sulphur. The carbon tetrachloride dissolves the sulphur 
 immediately on its separation and it is then oxidized by the 
 bromine also held in solution by this solvent. 
 
 4. It is necessary in all cases, before precipitation of 
 barium sulphate, to remove completely all traces of nitric or 
 nitrous acids by repeated evaporation with hydrochloric acid. 
 It was shown as early as 1842 by Mitscherlich that barium 
 sulphate carried down many salts from solution; on the other 
 hand, its complete precipitation is inhibited by the presence 
 of certain salts. In either case, when an accurate determi- 
 nation of sulphur is to be made, the salts which interfere 
 with the precipitation must be removed before barium chlo- 
 ride is added to the solution. Among the more common 
 substances which are carried down with barium sulphate are 
 the chloride, nitrate, and chlorate of barium, sodium, and 
 potassium; copper, zinc, iron, and aluminum salts. Some 
 interesting experiments have been made by Hulett and 
 Duschak, 1 showing the amounts of barium chloride actually 
 carried down with the barium sulphate. They have also 
 devised a very simple method by which corrections can be 
 made for this error. The complete precipitation of all of 
 the sulphuric acid in solution is interfered with by the pres- 
 ence of chromium salts, metaphosphoric acid, etc. 
 
 5. The precipitation of barium sulphate in the presence 
 of ferric iron has received attention from a large number of 
 investigators. It has been known for a long time that low 
 results are obtained when barium sulphate is precipitated 
 from solutions containing ferric iron, notwithstanding the 
 fact that the precipitate is contaminated with ferric oxide. 
 In the process of ignition sulphur trioxide is lost, and the 
 amount lost more than counterbalances the weight of the 
 ferric oxide. The old method of fusing the ignited precipi- 
 tate with an alkaline carbonate if it showed a red color, led to 
 inaccurate results, although insuring a purer precipitate. Some 
 of the sulphuric acid was always lost during the process of 
 ignition, the amount depending upon the tune and temperature 
 of heating. 
 
 Various remedies have been proposed hi order to overcome 
 1 Z. anorg. Chem., 40, 196.
 
 36 QUANTITATIVE ANALYSIS 
 
 the difficulty of the precipitation of barium sulphate in the 
 presence of iron salts. These remedies may be classed under 
 three heads: 
 
 (1) Removal of the iron before precipitation. 
 
 (2) Conversion of the ferric ion into a complex ion. 
 
 (3) Conversion of the ferric ion to the ferrous ion. 
 
 The Fresenius method is the best representative of the 
 first principle. In this case the iron is completely removed 
 as the oxide, and, after the subsequent removal of nitrous 
 and nitric acids by evaporation with hydrochloric acid, the 
 precipitation may be carried out under the most favorable 
 conditions. 
 
 In the Lunge method this same result is reached by pre- 
 cipitation of the iron by means of ammonia. Many chemists 
 have claimed that this precipitation of ferric hydroxide always 
 gave inaccurate results on account of the retention of basic 
 iron sulphate in the ferric hydroxide precipitate. Lunge has 
 repeatedly expressed the firm conviction, however, that if 
 the method is carried out as he directs there will be no loss 
 on this account. Gladding l modified the method so as to 
 obviate the tedious washing of the ferric hydroxide precipi- 
 tate. He recommended redissolving the ferric hydroxide in 
 hydrochloric acid and then adding barium chloride to this 
 solution to recover the small amount of sulphuric acid carried 
 down. 
 
 The most interesting modification of the Lunge method 
 involving the same principle is that of Kiister and Thiel. 2 
 They advised the precipitation of the ferric ion by an excess 
 of ammonia and the addition of barium chloride to this solu- 
 tion, dissolving the ferric hydroxide after the complete pre- 
 cipitation of the barium sulphate. This method overcomes 
 the objection raised by Gladding, in that it removes the iron 
 from solution, and furthermore it does away with a trouble- 
 some filtration. The results obtained by this method are un- 
 doubtedly higher than those obtained by the Lunge method. 
 Whether or not this is due to a certain counterbalancing of 
 errors remains to be seen after further experiments have 
 
 1 /. Amer. Chem. Soc., 16, 401. 
 
 2 Z. anorg. Chem., 19, 98.
 
 DETERMINATION OF SULPHUR IN PYRITE 37 
 
 been tried. It is highly probable that the occlusion of barium 
 chloride by the precipitation of the barium sulphate in the 
 ammoniacal solution would be considerable. Lunge l has 
 admitted and shown by a series of experiments that his own 
 method gives accurate results only because of a counter- 
 balancing of errors. 
 
 The conversion of the ferric ion into a complex ion was 
 first suggested by Jannasch and Richards, 2 who used for this 
 purpose formic, acetic, and citric acids, and this was extended 
 by Kiister and Thiel, 3 who used oxalic and tartaric acids; 
 but in neither case were the results entirely satisfactory, as 
 the precipitates were invariably colored by occluded iron. 
 
 The conversion of the ferric into the ferrous ion also 
 originated with Jannasch and Richards, 4 but the results by 
 this method have not been entirely satisfactory. When 
 zinc and hydrochloric acid are used for the reduction, the 
 zinc in solution is always carried down by the barium sul- 
 phate. In the hands of some chemists this method has been 
 more satisfactory, although it has not come into general use. 
 
 No completely satisfactory explanation has yet been offered 
 as to why barium sulphate carries down iron salts nor why 
 solutions containing iron salts will retain some barium sul- 
 phate. Some evidence has been offered by Jannasch and 
 Richards 5 for the existence of a barium ferric sulphate, 
 and this idea has been subjected to experimental study by 
 Schneider, 6 who came to the conclusion that a solid solu- 
 tion of low concentration was formed. On the other hand, 
 Ostwald 7 has called attention to the analogy with chromium 
 salts, which on heating form a complex ion containing the 
 SOi group, which does not react with barium ions. Kiister 
 and Thiel 8 developed this idea and assumed that some such 
 complex as Ba[Fe(SO 4 ) 2 ]2 is formed, but they admit from 
 the data at hand that it is impossible to say what the 
 nature of the complex is. It seems probable that the true 
 explanation will be that barium sulphate is capable of form- 
 ing a solid solution with some product of the hydrolysis of 
 ferric sulphate. 
 
 1 Z. angew. Cltem., 1904, 913; 1905, 1921. 2 /. prak. Chem., 147, 325. 
 
 3 Z. anorg. 'Chem., 19, 97. * J- prak. Chem., 147, 322. 
 
 6 /. prak. Chem., 147, 332. Z. phys. Chem., 10, 425. 
 
 7 Z. phys. Chem., 29, 340. 8 Z. anorg. Chem., 25, 322. 
 
 49108
 
 38 QUANTITATIVE ANALYSIS 
 
 Smith 1 reaches the conclusion that barium sulphate precipi- 
 tates thrown down in presence of ferric iron carry with them 
 a hydrated complex sulphato-ferriate of barium or other cation 
 which is present in intimate mechanical mixture with the 
 barium sulphate. 
 
 6. The explanation of the occlusion of barium chloride 
 is very much more satisfactory. Hulett and Duschak have 
 carefully determined the amounts of barium chloride carried 
 down by barium sulphate under varying conditions, and have 
 offered an explanation of this occlusion which seems to be 
 satisfactory. They have found that barium sulphate carefully 
 washed and dried at 140 contains not only barium chloride, 
 but also chlorine in some form of combination which splits 
 off hydrochloric acid at higher temperatures. Reasoning by 
 analogy, they assume that barium chloride may, like lead and 
 mercuric chlorides, dissociate not directly into Ba and Cl 
 ions, but the dissociation may take place in this way: 
 
 Bad' * Ba" + Cl'. 
 
 In like manner sulphuric acid also dissociates: 
 H 2 SO 4 <^HS0 4 ' +H* 
 HS0 4 +S0 4 " +H\ 
 
 On the assumption of such partial dissociation, Bad* ions 
 could combine with HSO' 4 ions to form the salt BaCl'HS0 4 , 
 which, upon being sufficiently heated, would split off hydro- 
 chloric acid and leave neutral BaSO 4 . Hulett and Duschak 
 also assumed the possibility of such reactions as the following 
 taking place: 
 
 I. .BaCT+SCV'^SO,. -' 
 
 II. 
 
 TTQ/"V 
 
 When the salt _ / Ba is heated it should give off sul- 
 HSO 4 * 
 
 phuric acid and leave barium sulphate; when heated together 
 
 with more than the equivalent of SO 4 , hydrochloric 
 
 Bad S 
 
 acid should form; and when heated with less than the equiv- 
 
 alent, then we should expect both hydrochloric and sulphuric 
 
 1 /. Amer. Cltem. .b'oc., 39, 1152.
 
 DETERMINATION OF TITANIUM IN TITANIUM IRON ORE 39 
 
 acids. This condition was actually realized by Folin, 1 who 
 studied the conditions under which either one or the other 
 of these two salts should form. He also suggested the possi- 
 bility of formation of salts of the character represented by 
 
 the formula ..-.,.,_ x Ba in which M represents any metal. 
 M o(J4 / 
 
 This would account for the carrying down of potassium when 
 barium sulphate is precipitated in its presence. 
 
 7. Too long and too high heating of the fusion in the 
 Fresenius method will cause considerable attack upon the 
 platinum crucible, and consequently the introduction of plati- 
 num into the analysis. If other constituents than sulphur are 
 to be determined, this fact must be taken into consideration, 
 and the platinum removed accordingly. 
 
 The completeness of the reaction may be judged by the 
 appearance of the residue after fusion. It should have 
 the color of rouge. If black, it will contain ferrous sulphide. 
 
 The use of coal gas for heating the crucible is almost 
 unavoidable in a large laboratory, but in small laboratories 
 the use of an alcohol lamp is strongly advised. The deflec- 
 tion of the products of combustion by means of an asbestos 
 board partially overcomes the difficulty. 
 
 That barium sulphate is easily reduced to sulphide during 
 the process of ignition is not sufficiently recognized. As 
 much care should be used in the ignition of this precipitate 
 as in the ignition of magnesium ammonium phosphate. 
 
 It should also be remembered that under no circumstances 
 is it permissible to heat barium sulphate over the blast lamp. 
 Above 900 it loses sulphur trioxide, and the residue becomes 
 alkaline, owing to loss of sulphur trioxide. 
 
 DETERMINATION OF TITANIUM IN TITANIUM 
 IRON ORE 
 
 I. METHOD OF GOOCH 2 
 
 Fuse 0.4-0.6 gram of finely-ground ore with 6 to 8 times 
 its weight of sodium carbonate until action ceases. Extract 
 the mass with hot water, and decant the solution through a 
 
 1 J. Biol. Chem., I, 131. 
 
 2 Chemical News, 52, 55, 68.
 
 40 QUANTITATIVE ANALYSIS 
 
 filter. Boil the residue with 25 cc. of sodium carbonate solu- 
 tion, filter, and then wash the residue on the filter paper 
 several times with dilute sodium carbonate solution. Place 
 the filter and residue in a platinum crucible, and ignite at a low 
 temperature until the filter paper is burned. Fuse with 12 to 
 15 parts of dry potassium pyrosulphate for one-half hour. 
 The temperature of the fusion should be so regulated that the 
 mass is kept in the molten condition, but sulphur trioxide 
 should escape only when the lid of the crucible is removed. 
 Cool, and remove the fusion from the crucible by means of a 
 platinum wire. Suspend the fusion in 200 cc. of cold water, 
 to which has been added 100 cc. of sulphurous acid, and allow 
 to stand in a cool place until solution is complete. 
 
 Filter if necessary. To the solution add 125 cc. of acetic 
 acid (sp. gr. 1.04), and dilute to 800 cc. in a liter beaker. Add 
 20 grams of ammonium acetate dissolved in a small amount of 
 water and boil from 3 to 5 minutes, adding, just before the 
 boiling point is reached, an additional 25 cc. of sulphurous 
 acid. Allow to stand in a warm place for one-half hour and 
 then filter, by means of a siphon, through a 9 cm. paper. 
 
 Wash the precipitate with five per cent acetic acid solution 
 until most of the sulphate has been removed, and then ignite 
 the paper and precipitate at a low temperature. Fuse with 
 potassium pyrosulphate again. Proceed exactly as before, 
 finally igniting to constant weight and weighing the precipitate 
 asTiOa. 
 
 Notes. i. Titanium iron ore is essentially ferrous titanate 
 of the composition represented by the formula FeTiO 3 , but it 
 rarely occurs in this form except in isolated crystals. In the 
 massive form it approximates this composition, but it is 
 always contaminated with silica, phosphorus, alumina, man- 
 ganese, calcium and magnesium. It is of small importance 
 at the present time as an iron ore, but it is the source from 
 which ferro-titanium, an increasingly useful product, is 
 obtained. 
 
 2. As some of the titanium ores are very refractory and do 
 not yield to treatment with acids, special methods have been
 
 DETERMINATION OF TITANIUM IN TITANIUM IRON ORE 41 
 
 devised, not only for getting the ores into solution, but also 
 for some of the separations necessary. Even if the ore proved 
 to be acid soluble, silica cannot be removed by the usual 
 process of evaporation to dryness, on account of the formation 
 of insoluble compounds containing phosphorus and titanium. 
 It is, therefore, customary to remove silica and phosphorus 
 together by fusing the ore with sodium carbonate. The 
 silicate and phosphate of sodium are leached out and removed 
 from the insoluble sodium titanate, ferric oxide, and carbon- 
 ates of the alkaline earths. Some of the aluminum goes into 
 the nitrate, and part remains in the residue. 
 
 3. Ignition of the sodium titanate residue must be made 
 at a low temperature, as it becomes less soluble on prolonged 
 or high heating. 
 
 4. Fusion of any substance with potassium pyrosulphate 
 requires considerable skill, patience and caution. The pre- 
 liminary heating should be carried out slowly and at a low 
 temperature, on account of the tendency to froth over, which 
 always results in the loss of the determination, and sometimes 
 in serious burns. To avoid as far as possible the frothing, 
 which is most vigorous during the first part of the heating, 
 potassium pyrosulphate should be used, and not the acid 
 potassium sulphate. The former may easily be made by 
 heating the latter to fusion for a very short time. Prolonged 
 heating leads to the formation of the normal sulphate which 
 expands on solidification and may crack the crucible. 
 
 It is an inexcusable mistake to try to hurry the solution of 
 the sulphate fusion by heating. Hydrolysis invariably takes 
 place, with the formation of a colloidal solution. 
 
 5. The separation of titanium from iron, aluminum and 
 manganese is based upon the same principle involved in the 
 separation of iron and aluminum from manganese, as given 
 under the Analysis of Spathic Iron Ore. 1 The iron is pre- 
 vented from hydrolyzing by being kept in the ferrous condi- 
 tion, and the aluminum and manganese are prevented by 
 having a hydrogen ion concentration so great that they will 
 not hydrolyze, but titanium may hydrolyze completely, and 
 precipitate. 
 
 Titanium will hydrolyze completely on continued boiling 
 from strong acid solutions, but it usually separates in such a 
 1 Page 23, note 5.
 
 42 QUANTITATIVE ANALYSIS 
 
 form as to make its filtration difficult, if not impossible. If 
 the directions are carefully followed, titanium hydroxide 
 should separate quickly and in a flocculent form which filters 
 with comparative rapidity. 
 
 6. Separation of iron and titanium by this method is usually 
 complete, although a second precipitation is necessary. Other 
 methods for the separation of these two elements have been 
 recommended, the most promising of which depends upon the 
 solubility of ferric chloride in ether. 
 
 7. Colorimetric estimation of titanium, 1 when present in 
 small amounts, is not only rapid, but very accurate. The 
 method is based upon the reaction which takes place between 
 titanium salts and hydrogen peroxide. The intense yellow 
 color produced is compared with standards of known titanium 
 content. 
 
 H. METHOD OF BARNEBY AND ISHAM 2 
 
 The weighed sample, 0.4-0.6 gram, is moistened in a 
 platinum crucible with a few drops of water, 5-10 drops of 
 concentrated sulphuric acid and i cc. of hydrofluoric acid are 
 added. The crucible is heated as in the removal of silica until 
 the complete removal of sulphur trioxide. Five grams of 
 sodium carbonate and a little sodium nitrate are added and 
 the mixture fused for 30 minutes. After cooling, the crucible 
 is placed in a beaker, covered with hot water, and heated until 
 the melt is disintegrated. The residue is filtered and washed 
 with hot water. The filter is perforated and washed with 
 hydrochloric acid (sp. gr. i.n) into a clean beaker, and any 
 residue in the crucible is dissolved and added to this solution. 
 The beaker is heated until solution is complete and until the 
 volume is reduced to 15-20 cc. After cooling the solution is 
 transferred to a separatory funnel, the beaker being washed 
 with hydrochloric acid (sp. gr. i.n). An equal volume of 
 ether, which has been shaken with concentrated hydrochloric 
 acid, is added, the funnel inverted and shaken, releasing the 
 pressure at intervals by opening the stop-cock, and allowing 
 
 1 U. S. Geol. Survey, Bulletin 422, 128. 
 
 2 J. Am. Chem. Soc., 32, 957.
 
 IODIMETRIC DETERMINATION OF COPPER 43 
 
 the liquid carried up by the escaping vapor to run back before 
 closing the stop-cock. The funnel is allowed to stand in an 
 upright position for 10 minutes, and then the aqueous layer is 
 drawn off into another separatory funnel. The ether extract is 
 rinsed twice by shaking with 5-10 cc. portions of hydrochloric 
 acid and the washings added to the aqueous solution. The 
 extraction with ether is repeated until the ether layer no longer 
 shows the greenish tinge of dissolved ferric chloride. The 
 aqueous solution is then transferred to a beaker, 10 cc. of con- 
 centrated sulphuric acid are added and the solution evaporated 
 to fumes of sulphur trioxide. The cooled solution is diluted 
 to about 100 cc. and nearly neutralized with ammonia. Two 
 grams of ammonium bisulphite are added and the solution 
 warmed for one-half hour. Ten grams of ammonium acetate 
 and 5-10 cc. of glacial acetic acid are added and the solution 
 boiled for 15 minutes. Filter, wash with dilute acetic acid, 
 ignite and weigh as TiCV 
 
 Notes. i. The method is based on the removal of silica 
 with hydrofluoric acid, the removal of phosphates, sulphates, 
 and aluminates by fusion with sodium carbonate and extrac- 
 tion with hot water, and the removal of the iron by extraction 
 of the ferric chloride with ether. The titanium is then hydro- 
 lyzed as in the preceding method hi the presence of an acetate 
 and acetic acid. 
 
 IODIMETRIC DETERMINATION OF COPPER 
 
 MODIFIED METHOD OF A. H. LOW 1 
 
 Standardization of Sodium Thiosulphate Solution. Prepare 
 a solution of sodium thiosulphate by dissolving about twenty- 
 five grams of the salt in one liter of water, which has been 
 recently boiled and cooled out of contact with the air. Weigh 
 accurately two portions of about 0.25 gm. of pure, clean copper 
 wire into 250 cc. Erlenmeyer flasks, and dissolve in a mixture 
 of 5 cc. concentrated nitric acid and 25 cc. of water. When 
 solution is complete, heat to boiling, add 5 cc. of bromine water 
 
 1 J. Am. Chem. Soc., 24, 1082.
 
 44 QUANTITATIVE ANALYSIS 
 
 continuing the boiling until the bromine is expelled. Allow 
 to cool, add a slight excess of strong ammonia, and boil the 
 solution until the excess has been driven off, adding water 
 occasionally to maintain the volume. This point is indicated 
 by the change in color of the solution, or by a slight precipi- 
 tation of basic copper salt. Add a few cubic centimeters in 
 excess of acetic acid and boil if necessary to dissolve any basic 
 copper salt. Cool to room temperature, add three grams of 
 potassium iodide, dissolved in a small amount of water, and 
 titrate at once with the sodium thiosulphate solution. When 
 the brown color of the iodine has become weak, add enough 
 starch solution to give a distinct blue color, and finish the 
 titration-by adding sodium thiosulphate slowly until the color 
 due to the free iodine has completely disappeared. 
 
 Work slowly toward the end point and stop short of complete 
 decolorization, and then continue only when the liquid after 
 standing a minute or two, still persists in a tinge of color. 
 
 Procedure. To 0.25-0.50 gram of finely ground ore weighed 
 into a 250 cc. Erlenmeyer flask, add 6 cc. of nitric acid (sp. gr. 
 1.42) and heat gently nearly to dryness. Add 5 cc. of strong 
 hydrochloric acid and heat again. As soon as the incrusted 
 matter has dissolved, add 7 cc. of concentrated sulphuric acid 
 and heat until the sulphuric acid fumes freely. Cool, and add 
 25 cc. of water. Then heat until any anhydrous ferric sul- 
 phate is dissolved, and filter to remove insoluble sulphates and 
 silica. Wash the flask and filter paper until the volume of the 
 filtrate amounts to about 75 cc., receiving it in a No. 2 beaker. 
 Place on its edge in the beaker a piece of sheet aluminum, 
 2 in. x 6 in., bent into triangular shape. Cover the beaker and 
 boil gently for seven to ten minutes, which will be sufficient to 
 precipitate all the copper provided the solution does not ex- 
 ceed 75 cc. Avoid evaporating to very small bulk. The 
 aluminum should now appear clean, the copper being de- 
 tached or loosely adhering. Remove from the heat and wash 
 down the cover and sides of the beaker with hydrogen sulphide 
 water. Decant the liquid through a filter and then without 
 delay transfer, by means of a jet of hydrogen sulphide water
 
 IODIMETRIC DETERMINATION OF COPPER 45 
 
 from a wash bottle, the copper to the filter, leaving the foil as 
 clean as possible in the beaker. Wash the copper and filter 
 thoroughly with hydrogen sulphide water, being careful not 
 to allow the filter to stand empty until the washing is finished. 
 Place the original flask under the funnel and pour over the 
 aluminum in the beaker 5 cc. of a mixture of equal parts of con- 
 centrated nitric acid and water. Heat just to boiling and pour 
 the hot acid very slowly upon the filter, lifting the fold if 
 necessary. Now, before washing, pour 5 cc. of bromine water 
 into the filter and wash the beaker and filter with hot water. 
 Finally, remove the filter and wash any residue upon it into 
 the flask. If the bromine was not sufficient to give a slight 
 tinge to the filtrate more of it must be added. Boil the nitrate, 
 which does not exceed 75 cc. to expel the excess of bromine but 
 do not concentrate to small volume. Remove from the heat 
 and add a slight excess of strong ammonia (usually 7 cc.). 
 Boil off the excess of ammonia and add 3 or 4 cc. of strong 
 acetic acid. Cool to room temperature, add three grams of 
 potassium iodide, dissolved in a small amount of water and 
 titrate with thiosulphate as in the standardization. 
 
 Notes. i. The principle of this method is based upon the 
 fact that when potassium iodide is added to a weakly acid 
 solution of a copper salt iodine is liberated according to this 
 equation: 
 
 2 (CH 3 COO) 2 Cu + 4 KI = Cu 2 I 2 + 4 CH 3 COOK + 1 2 . 
 The iodine is then made to react with standardized sodium 
 thiosulphate according to the reaction: 
 
 2Na 2 S20 3 + I 2 = aNal + Na 2 S,O 6 . 
 
 2. Copper ores contain varying amounts of silica, sulphur, 
 arsenic, antimony, lead, silver and iron, and the method is so 
 designed as to remove or render these elements inert. 
 
 Nitric acid is the best solvent for sulphide ores and in order 
 to dissolve oxides of iron, remove silver, lead and silica hydro- 
 chloric and sulphuric acids are used simultaneously. Com- 
 plete solution of anhydrous ferric sulphate must take place 
 previous to the filtration of silica, silver chloride and lead 
 sulphate.
 
 46 QUANTITATIVE ANALYSIS 
 
 3. Aluminum foil precipitates copper from its solution 
 quickly and completely provided the proper concentration of 
 acid has been obtained. In addition to copper, lead, silver, 
 bismuth, arsenic and antimony may be precipitated if present. 
 Lead should, however, have been almost completely removed 
 as insoluble sulphate, and silver as chloride. The presence of 
 lead and bismuth have no effect upon the analysis except for 
 a slight change of color due to the presence of their slightly 
 colored iodides. 
 
 Arsenic and antimony would interfere if allowed to remain 
 in the lower state of oxidation by reacting with the liberated 
 iodine. Consequently the nitric acid solution is treated with 
 bromine to oxidize these elements to the higher state of oxida- 
 tion. Bromine also oxidizes any nitrous acid which might be 
 in the solution as the result of the reaction between copper and 
 nitric acid. 
 
 4. Hydrogen sulphide water is used not only to precipitate 
 traces of copper left in solution, but also to prevent oxidation 
 of finely divided copper. 
 
 5. The addition of ammonia and acetic acid is for the 
 purpose of regulating the hydrogen ion concentration. In the 
 presence of strong acids there is danger of hydriodic acid 
 reducing arsenic and antimonic acids with the liberation of 
 free iodine. This reaction takes place so very slowly in the 
 weak acid solution that the effect is negligible. 
 
 The presence of much ammonium acetate interferes with 
 the sharpness of the end point, due to the oxidizing action of 
 iodine upon the ammonia from the hydrolyzed ammonium 
 acetate. 
 
 6. The amount of potassium iodide added is in excess of 
 that required for 0.50 gram of copper and consequently should 
 be enough to take care of all ores on the weight of sample 
 specified. The reaction proceeds slowly unless an excess is 
 present. An excess beyond a certain limit has no effect upon 
 the analysis, and this excess must be controlled in accordance 
 with the volume of the solution. 
 
 The reaction takes place in the presence of sulphuric and 
 hydrochloric acids as well as in acetic acid solution, but on 
 account of the greater hydrogen ion concentration in the two 
 former acids, the volume of the solution must be more care-
 
 PROXIMATE ANALYSIS OF COAL 47 
 
 fully regulated in order to prevent reaction between the acid 
 and potassium iodide. Gooch and Heath 1 have studied the 
 effect of concentration upon this reaction for sulphuric, hydro- 
 chloric, nitric and acetic acids. 
 
 7. Standardized solutions of sodium thiosulphate and iodine 
 are both fairly stable if properly prepared from pure materials 
 and protected from light and heat. Water free from carbon 
 dioxide should be used for the solution of the sodium thio- 
 sulphate, otherwise there is separation of free sulphur. Both 
 solutions should, however, be frequently checked against each 
 other and against some standard. Copper is used in this case 
 as the standard as it is easily obtained in pure form, and further 
 it reproduces all of the conditions obtained during the course 
 of the analysis. In this connection it is well to refer to the 
 method given on page 83, for the standardization of sodium 
 thiosulphate. 
 
 PROXIMATE ANALYSIS OF COAL 2 
 
 Moisture. A one gram air-dried sample is weighed into a 
 shallow porcelain capsule J inch deep and if inches in diameter, 
 and heated for one hour at 105 in a constant temperature 
 oven, through which a current of air, dried by passing through 
 sulphuric acid, is passing at a rate to change the total volume 
 of air in the oven two to four times a minute. The covered 
 capsule is cooled in a desiccator over sulphuric acid and 
 weighed. The loss in weight is called "moisture at 105." 
 
 Ash. The determination is made upon the same sample 
 in which moisture has been determined. The capsule con- 
 taining the sample is heated at a low temperature obtained by 
 placing it above the tip of a flame turned down to 2 or 3 inches 
 in height. The sample should be stirred frequently with a 
 stout nichrome wire. After much of the carbon has been 
 burned off the temperature should be gradually raised to 
 about 750 C. and the heat continued at this temperature for 
 
 1 Am. J. Set. [4], 24, 67. 
 
 2 Technical Paper 76, Bureau of Mines 1914; Am. Soc. for Testing Materials 
 (1914); Bureau of Mines Bulletin, 116; Technical Paper 133, Bureau of Mines, 
 1917.
 
 4 QUANTITATIVE ANALYSIS 
 
 30 minutes, or until all of the carbon is burned off. Cool in 
 a desiccator and weigh. The residue represents the "un- 
 corrected ash." 
 
 Volatile Matter. One gram of coal is weighed into a plati- 
 num crucible which is provided with a tightly fitting per- 
 forated nickel cover. The crucible is placed in the flame of a 
 Meker burner and heated for seven minutes at 950 C. After 
 the more rapid early discharge of the volatile matter, tap the 
 crucible cover so as to seal the cover and crucible more per- 
 fectly. The temperature should be determined by means of 
 a thermo-couple placed through the cover, and touching 
 lightly upon the bottom of the crucible. Cool the crucible in 
 a desiccator and weigh. The loss in weight minus the moisture 
 at 105 C. represents the volatile matter. 
 
 Fixed Carbon. The fixed carbon is obtained by subtracting 
 the sum of the percentages of moisture, ash and volatile matter 
 from 100. 
 
 Sulphur. Mix thoroughly on glazed paper i gram of coal 
 and 3 grams of Eschka mixture (two parts of magnesium oxide 
 and one part of sodium carbonate), transfer to a porcelain 
 crucible, or a platinum crucible, and cover with i gram of 
 Eschka mixture. Heat the crucible at a low temperature for 
 30 minutes with an alcohol flame or in an electrically heated 
 muffle furnace, and then raise the temperature to about 925 C. 
 for one hour or until all of the black particles of carbon dis- 
 appear. 
 
 Transfer to a 300 cc. beaker and digest with 100 cc. of hot 
 water for one-half hour. Filter, wash by decantation several 
 times and finally transfer the residue to the filter paper and 
 wash. Treat the filtrate amounting to about 250 cc. with 
 10 cc. saturated bromine water, acidify slightly with hydro- 
 chloric acid and boil off the excess of bromine. To the 
 boiling solution add hot dilute barium chloride and proceed 
 as in the determination of sulphur in pyrite. 
 
 Notes. i. The proximate analysis of coal originated in 
 response to a demand for a rapid, fairly accurate method for 
 classifying coals for industrial purposes. The results give
 
 PROXIMATE ANALYSIS OF COAL 49 
 
 important data as to the commercial value, and may be 
 obtained in very much shorter time than the results of an 
 ultimate analysis. 
 
 2. The sampling of coal, whether at the mine or at the place 
 of destination, is a very important matter, and the subject is 
 too extensive to be dealt with here. The student should read 
 on this subject the papers referred to on page 47. 
 
 3. The loss of weight at 105 includes not only that moisture 
 which is adsorbed on the surface, but also some of the water 
 which is an essential part of the coal itself. Samples may have 
 been taken from a wet face of the mine or from a car which had 
 been standing in the rain or snow, and the amount of this 
 moisture varies with the weather conditions. It is, therefore, 
 desirable to establish nearly normal conditions by air-drying 
 the sample. Coal exposed to circulating air at ordinary tem- 
 peratures will either give up or take on moisture depending 
 upon the fineness of the sample and the humidity of the air. 
 A point of equilibrium is always reached which varies with 
 the size of the lumps of coal exposed, the time of exposure 
 and the humidity of the air. 
 
 The inherent moisture or water of constitution varies fairly 
 constantly with the district from which the coal is taken. The 
 coal from the more important veins of the Appalachian field 
 holds about 2-4 per cent of moisture; that from the western 
 part of this field, in Ohio, varies from 4-10 per cent; and that 
 from the Indiana and Illinois field varies from 8-17 per cent. 
 
 The commercial value of a coal naturally decreases with the 
 increase in the amount of moisture. 
 
 4. The ash of a coal is the incombustible residue left after 
 all of the carbonaceous material has been burned. It consists 
 of compounds of silica, alumina, calcium, magnesium, iron, 
 and the alkali metals, and is derived from the original vegetable 
 matter, or from the impurities formed during the formation 
 of the coal bed. 
 
 The amount of ash in a coal influences its commercial value 
 greatly, as the efficiency of combustion is influenced not only 
 by the amount but also by the character of the ash. The 
 fusibility of the ash is also an important factor, and there is a 
 close relationship between the melting point of the ash and the 
 tendency to clinker. 
 
 1 Am. Soc.for Testing Materials, Standards, 1918, 673.
 
 50 QUANTITATIVE ANALYSIS 
 
 5. The volatile matter and fixed carbon represent respec- 
 tively, the gaseous and solid material which are capable of 
 being utilized by heating in a closed space. The former 
 consists of carbon monoxide, hydrogen, methane and higher 
 hydrocarbons, and the non-combustible gases, carbon dioxide 
 and. water vapor. The composition of the volatile matter 
 varies greatly with the source of the coal, and the amount 
 varies with the temperature of heating. The temperature of 
 heating, 950, is purely arbitrary and is a convenient one from 
 the analyst's point of view. There is greater variation in 
 results when heating at low temperatures than at the tempera- 
 ture selected. The fixed carbon and ash represent roughly the 
 amount of coke which may be obtained from any coal. 
 
 6. Sulphur 1 is present in coal as pyrite or marcasite, as 
 sulphate of iron, aluminum or calcium, and as an organic 
 compound. The presence of much pyrite increases the 
 tendency of the ash to clinker, and thus decreases the com- 
 mercial value of the coal. The sulphur present as an organic 
 compound is without much influence except that it lowers the 
 heating value. 
 
 7. Eschka's mixture is especially valuable in the oxidation 
 of sulphur in coal, as more energetic oxidizing mixtures would 
 oxidize the carbon with explosive violence. Oxidation of 
 sulphur in coal by this method is a rather slow process, but a 
 large number of determinations can be carried on simultane- 
 ously without much attention. 
 
 The Eschka mixture may also be applied to the oxidation of 
 sulphur in mineral sulphides and to the determination of sul- 
 phur in iron and steel. 
 
 ANALYSIS OF PHOSPHOR-BRONZE 
 
 DETERMINATION OF TIN 
 
 Treat 0.5 gram of the borings in a No. i beaker with 15 cc, 
 HNOs (sp. gr. i . 20) . Evaporate the solution on the water bath 
 or steam table until the residue is just dry, and remove from the 
 heat as soon as this stage is reached. Prolonged heating of 
 this residue must be avoided. Treat the residue with a mix- 
 ture of 10 cc. HNOs (sp. gr. 1.42) and 50 cc. water, dividing the 
 acid into three portions and boiling and decanting after each 
 
 1 Technical Paper, 254, Bureau of Mines.
 
 ANALYSIS OF PHOSPHOR-BRONZE 51 
 
 addition through a 9 cm. hardened filter paper; finally com- 
 plete the washing by boiling and decanting with distilled water 
 until the metastannic acid is apparently free from copper. 
 The first portion of filtrate should be carefully examined for 
 metastannic acid, refiltered if necessary, and each successive 
 clear filtrate should be removed from below the funnel before 
 more wash water is added, as the precipitate is likely to run 
 through the filter. Preserve the filtrate and washings for the 
 determination of lead and copper. Leave as much of the meta- 
 stannic acid as possible in the beaker during the washing. 
 When the washing is completed and the filter has drained, 
 remove the filter from the funnel, spread it out upon a watch 
 glass, hold it vertically above the beaker containing the 
 residue, and wash off all the visible residue into the beaker, 
 limiting the amount of water as far as practicable. To dis- 
 solve the small amount of tin remaining in the filter, spread 
 on the bottom of a beaker and pour over it 25 cc. of yellow 
 ammonium sulphide solution and warm gently (covered) for 
 15 minutes. Meanwhile add 75 cc. of the sulphide solution 
 to the metastannic acid in the beaker. Pour into this the 
 sulphide from the filter and wash the filter three times with hot 
 water. If the filter is perfectly white after washing out the 
 sulphide, it may be discarded; otherwise it should be pre- 
 served and treated as directed below. Cover the beaker 
 containing the sulphide solution and digest the solution on 
 the steam table at a low heat for about two hours with occa- 
 sional stirring. Filter off the copper and lead sulphides, 
 receiving the filtrate in a 600 cc. beaker and wash with 100 cc. 
 of dilute ammonium sulphide (i volume ammonium mono- 
 sulphide and 4 volumes of water). Complete the washing 
 with water. 
 
 Treat the filter with 15 cc. of hot HN0 3 (i part HN0 3 sp. 
 gr. i .42 to 4 parts water) passing it through the filter until the 
 sulphides are dissolved. If the filter treated with ammonium 
 sulphide was preserved, pass the same hot nitric acid solution 
 through that filter and wash both filters completely, adding 
 the filtrates and washings after neutralization with ammonia
 
 52 QUANTITATIVE ANALYSIS 
 
 to the copper and lead solution. Smoke off the filters in 
 a weighed porcelain crucible and ignite finally at a high 
 temperature, weighing the residue as SnO 2 . This weight is 
 to be added to that obtained below. 
 
 Dilute the ammonium sulphide solution to 500 cc., make 
 distinctly acid with dilute acetic acid, but avoid a large excess. 
 Cover the beaker and allow it to stand in a warm place for 
 three to four hours, or over night. Decant the liquid through 
 a filter, pour over the precipitate 150 cc. water, again decant 
 and pour over it 150 cc. of dilute ammonium nitrate solution 
 (2 gms. per 100 cc. H^O acidified with a few drops of acetic 
 acid). Again decant and transfer the precipitate to the filter, 
 completing the washing with the nitrate solution. Transfer 
 the moist filter and precipitate to a weighed porcelain crucible. 
 Dry and smoke off the filter over a low flame, then gradually 
 raise the temperature until the filter burns and the sulphur 
 takes fire. Continue the gentle heating until there is no odor 
 of sulphur dioxide, and then increase the heat gradually until 
 the highest heat of the Tirrill is reached. A short heating 
 over the blast lamp is advisable before final weighing. Ignite 
 to constant weight. Weigh as Sn0 2 and report as metallic 
 tin. 
 
 Notes. i. The phosphor bronzes are essentially alloys of 
 copper, tin, and phosphorus, but they may also contain lead 
 and traces of other elements. They are characterized by their 
 excellent qualities as a mechanical constructive material, the 
 small effect which rise of temperature has upon the physical 
 properties, and their desirable properties for bearings for high 
 speed machinery and fairly heavy loads. For bearings they 
 possess a low coefficient of friction and are hard enough to 
 resist abrasion. They may be classified according to their 
 mechanical properties into two groups: (i) the malleable phos- 
 phor bronzes; and (2) the cast phosphor bronzes. 
 
 The malleable phosphor bronzes are those in which the 
 quantities of tin and phosphorus are relatively small, the upper 
 limits of which may be taken as 6 per cent of tin and 0.3 per 
 cent of phosphorus. The malleability is due to the fact that, 
 with small amounts of phosphorus, copper phosphide,
 
 ANALYSIS OF PHOSPHOR-BRONZE 53 
 
 forms a homogeneous solid solution, and, further, with larger 
 amounts of phosphorus, the tin promotes the separation of 
 finely divided and widely disseminated copper-copper phos- 
 phide eutectic. Such alloys are useful for wire, gear wheels, 
 pinions, bushings, worm wheel rims, etc. 
 
 The cast phosphor bronzes are particularly suited for bear- 
 ings. The composition of such alloys varies widely, but the 
 upper limits of tin and phosphorus may be placed respectively 
 at 12 and 1.5 per cent. Lead is also a common constituent of 
 these bearing metals and may be found sometimes to the extent 
 of 10 per cent. These alloys show on microscopic examination 
 the essential structure of a good bearing metal, viz.: a hard 
 crystalline constituent embedded in a softer plastic matrix. 
 The load is carried by the harder material while the plastic 
 material provides for continuous lubrication. The presence 
 of lead increases the amount of plastic substance. 
 
 2. In addition to the elements copper, tin, phosphorus and 
 lead, bronzes may also contain antimony, arsenic, iron, alu- 
 minum, but, inasmuch as most of the commercial bronzes con- 
 tain these elements only in traces, a special method has not 
 been given for their separation. If it is found that antimony 
 is present and it is desired to separate and determine it, 
 reference should be made for details to Treadwell-Hall, 
 Quantitative Analysis. The principle of the separation is 
 based on the insolubility of antimony sulphide in oxalic acid 
 solution while stannic salts are not precipitated. Copper and 
 lead are also held in small amounts in the residue and are left 
 as sulphides after treatment of the antimony sulphide with 
 ammonium polysulphide. 
 
 3. Evaporation to hard dryness of the metastannic acid 
 causes it to lose water, and results in making it quite insoluble 
 in ammonium polysulphide, and consequently considerable 
 care should be exercised at this point. It is not even neces- 
 sary to evaporate to complete dryness, as the subsequent 
 boiling with dilute nitric acid will cause the complete separa- 
 tion of the metastannic acid. 
 
 4. Metastannic acid is not only finely divided and will 
 pass through the finest filter paper, but it also has a strong 
 tendency to become colloidal. Boiling the residue with sepa- 
 rate portions of dilute nitric acid retards the formation of
 
 54 QUANTITATIVE ANALYSIS 
 
 the colloid and also removes small amounts of copper and lead, 
 which are carried down with it. 1 
 
 5. Stannic sulphide is thrown out of its sulpho-stannate 
 solution with acetic acid, rather than with hydrochloric or 
 sulphuric acids, because of its being a weaker acid. There is, 
 therefore, less danger of re-dissolving the sulphide with a 
 slight excess of acid. 
 
 As stannic sulphide is slightly soluble in solutions containing 
 hydrogen sulphide in excess, it is essential to allow it to stand 
 in a warm place until all hydrogen sulphide is removed. If 
 this precaution is neglected, tin will appear in the nitrate. It 
 is always advisable to test this nitrate for tin, passing hydrogen 
 sulphide into the boiling solution. 
 
 6. The stannic sulphide, if washed with water, becomes 
 colloidal on the removal of the electrolytes present and passes 
 through the filter. It is therefore necessary to wash with an 
 electrolyte to prevent this. Ammonium nitrate is chosen 
 because it leaves no residue on ignition, and, further, it helps 
 appreciably in burning the filter paper. 
 
 7. It is essential to heat the stannic sulphide at a low 
 temperature at first because of the volatility of the sulphide. 
 The burning of the sulphur furnishes enough heat. As soon 
 as the sulphur is burned off the temperature may be raised 
 without danger. Like silica and titanic oxide, stannic oxide 
 loses its last traces of moisture only at the temperature of the 
 blast. The color of this ignited oxide darkens slowly at the 
 higher temperatures and usually varies in color from a light 
 to a dark gray. 
 
 It is advisable in the ignition of any sulphide to oxide to 
 heat with dry ammonium carbonate to remove traces of 
 sulphate which might have formed during the ignition. 
 
 DETERMINATION OF COPPER AND LEAD 
 
 The filtrate from the metastannic acid should not exceed in 
 volume 200-250 cc. To determine these two elements the 
 solution is subjected to electrolysis until both have been com- 
 pletely precipitated, the copper as metal on the cathode, the 
 
 1 This and other colloids may be conveniently coagulated by stirring into the 
 solution some white of egg, which on heating will coagulate and bring down with 
 it the substance which it is desired to filter.
 
 ANALYSIS OF PHOSPHOR-BRONZE 55 
 
 lead as hydrated peroxide on the anode. For this purpose, 
 clean two platinum electrodes, preferably of gauze, by immers- 
 ing for some time in hot dilute nitric acid, washing with water 
 and then with alcohol. Dry at 130 and weigh. Place the 
 electrodes in the solution making the smaller the cathode, the 
 larger the anode and begin the electrolysis with a current of 
 0.25 ampere. Continue this strength current from 4-6 hours, 1 
 and then reduce to o.i ampere. After 18-20 hours the elec- 
 trolysis should be complete. The level of the liquid in the 
 beaker should be raised by about one-eighth of an inch by 
 the addition of water. After half an hour an observation 
 should be made to see whether any more copper has been 
 precipitated on the clean surface of the wire of the electrode. 
 If the copper has been completely deposited, the wire of the 
 cathode will remain perfectly bright just under the surface 
 of the liquid. If any copper shows, this process must be 
 repeated until it has all been removed from solution. When 
 complete the solution should be siphoned off and the elec- 
 trodes carefully washed with distilled water, the current being 
 allowed to pass through the solution during the process of 
 washing. Before disconnecting the electrodes insert a wire 
 between the binding posts so as not to cut out any other 
 solutions in the same circuit. Immerse the copper electrode 
 in water, wash with alcohol, dry at 100. After washing the 
 lead peroxide in water, dry at 180 C. 
 
 The lead peroxide does not adhere as firmly to the electrode 
 as the copper. A better deposit may be obtained by increas- 
 ing the surface area of the electrode, and by having a matt 
 surface. If any peroxide should become detached, the solu- 
 tion is decanted away from it, it is washed with water, and 
 then by means of alcohol washed on to a weighed watch glass 
 and the excess of alcohol is evaporated off. In this way small 
 quantities of lead peroxide can be conveniently collected and 
 weighed. 
 
 1 If gauze electrodes are available the current may be increased to 4-5 
 amperes so as to bring down the copper and lead dioxide very much more 
 rapidly, finishing the electrolysis in the course of several hours.
 
 .56 QUANTITATIVE ANALYSIS 
 
 Notes. i. In nitric acid solution of copper and lead salts, 
 the former is deposited on the cathode as strongly adhering 
 metallic copper, and the latter on the anode as hydrated lead 
 dioxide. Various views have been expressed to account for 
 the deposition of lead as peroxide on the anode, and it is not 
 certain whether lead nitrate hydrolyzes, giving rise to PbO 2 ~ 
 ion, or whether a similar ion is formed from the decomposition 
 of Pb(N0 3 ) 4 . 
 
 2. Lead may be accurately separated from copper by taking 
 advantage of the insolubility of lead sulphate in dilute sul- 
 phuric acid, but the simultaneous separation of the metals 
 electrolytically produces results just as accurate, and for 
 commercial work very much more convenient. The separa- 
 tion must take place in nitric acid solution and with a con- 
 centration of from 5 to 10 per cent by volume of the strong 
 acid. This separation is not applicable in presence of meta- 
 stannic acid, as both copper and lead dioxide are contaminated 
 with it. 
 
 3. The time for electrolysis seems excessive, but where a 
 large number of determinations are to be carried out from day 
 to day this is not an important factor. Increasing the current 
 density is out of the question because of the spongy, non- 
 adherent deposits. Stirring the solution helps some, but the 
 most efficient aid to rapid electrolysis is in the use of the 
 rotating cathode or anode. By this process the solution is 
 not only stirred, thus constantly bringing fresh solution to the 
 electrode, but it is possible to obtain adherent deposits with 
 high current density. The rotating cathode was proposed by 
 Gooch and Medway, 1 and the rotating anode by Exner. 2 In 
 each case the electrode was rotated at high speed by means of 
 a small motor. Frary 3 proposed in place of the motor to place 
 the solution in a solenoid and thus cause a rotation of the 
 solution itself. Winkler 4 proposed the use of platinum gauze 
 electrodes, but the advantages were not recognized until 
 Stoddard had called attention to their use, and showed that 
 there is economy of platinum and time, as well as avoidance of 
 
 1 Am.J.Sci. (4), 15, 320. 
 
 2 /. Am. Chem. Soc., 25, 896. 
 
 3 Zeit. Electrochem, 1907, 308. 
 
 4 Ber., 32, 2192.
 
 ANALYSIS OF PHOSPHOR-BRONZE 57 
 
 the more complicated apparatus. Stoddard 1 has obtained 
 excellent results by using a gauze cathode. He operated at 
 higher temperatures and used heavier currents than would be 
 possible with plate electrodes, and reduced the time of opera- 
 tion to the limit of the rotating apparatus. 
 
 4. Washing, drying and weighing of the deposited copper 
 must be carried out without unnecessary delay, in order to 
 prevent oxidation. By careful ignition lead dioxide may be 
 quantitatively converted into lead monoxide. 
 
 DETERMINATION OF PHOSPHORUS^ IN PHOSPHOR-BRONZE 
 
 Weigh one gram of the borings into a No. 2 beaker and dis- 
 solve in 20 cc. of aqua regia, made by mixing just previous to 
 use equal volumes of the concentrated acids. Have the 
 beaker covered with a watch glass, and after solution is com- 
 plete heat nearly to the boiling point of the solution for fifteen 
 minutes. After cooling add 25 cc. of water, and then just 
 sufficient ammonia (sp. gr. 0.90) to re-dissolve all of the copper 
 hydroxide and to produce a deep blue colored solution. Now 
 add 50 cc. colorless ammonium sulphide. This should be 
 enough to precipitate the sulphides, and the supernatant 
 liquid should show no blue color. If it does, more ammonium 
 sulphide should be added. Digest at a temperature near the 
 boiling point of the solution for fifteen minutes. Allow the 
 precipitated sulphides of copper and lead to settle, filter into 
 a 300 cc. Erlenmeyer flask, decanting the clear liquid carefully 
 from the precipitate, and finally throwing the precipitate upon 
 the filter. When the filter has drained put the filter and 
 precipitate back into the beaker, add 50 cc. of ammonium 
 sulphide wash water (one part colorless ammonium sulphide 
 to three parts water) warm and stir occasionally for ten 
 minutes; then pour the whole upon another filter, wash with 
 50 cc. of ammonium sulphide wash water and allow it to drain 
 completely. The total volume should not be over 250 cc., 
 but it is not necessary to evaporate in case this volume is 
 
 1 /. Am. Ghent. Soc., 31, 385. 
 
 2 Dudley and Pease, Eng. and R. R. Journ., March, 1894.
 
 58 QUANTITATIVE ANALYSIS 
 
 slightly exceeded. Add to the filtrate 10 cc. of magnesia 
 mixture and shake. Place the flask in ice water and allow 
 to stand with occasional shaking for two hours. Filter off 
 the precipitate of magnesium ammonium phosphate upon a 
 small filter and wash with ammonia water (one part 0.96 sp. 
 gr. ammonia to three parts water) until nearly free from 
 sulphide. Pour 10 cc. hydrochloric acid (one part HC1 sp. 
 gr. i. 20 to four parts water) into the flask, taking care that all 
 of the precipitate adhering to the walls of the flask is dissolved, 
 and then pour the acid through the filter, allowing the solution 
 to run into a No. i beaker. Wash the flask and filter with 
 10 cc. more of the same acid. Add 3 cc. of magnesia mixture 
 to the nitrate, and then add ammonia slowly with constant 
 stirring until present in slight excess. Add 10 cc. ammonia 
 sp. gr. 0.90, allow to stand in ice water for two hours, stirring 
 the solution occasionally. Filter, and wash the precipitate 
 with ammonia water until free from chlorides, and ignite with 
 the usual precautions, weighing as 
 
 Notes. i . The principle of this determination is dependent 
 upon the oxidation of phosphorus, existing in the alloy as 
 phosphide, by means of aqua regia; separation of lead and 
 copper as sulphides by means of ammonium sulphide, and 
 after filtration the precipitation of magnesium ammonium 
 phosphate in presence of ammonium sulpho-stannate. 
 
 2. Complete oxidation of the phosphorus cannot be accom- 
 plished by means of nitric acid alone, although many of the 
 published methods are based on the quantitative separation of 
 phosphorus along with the tin in the residue obtained after 
 treatment with nitric acid. This process leads to erroneous 
 results, as some of the phosphorus always remains unoxidized. 
 Treatment with aqua regia seems necessary for complete 
 oxidation. 
 
 3. Ammonia is added previous to the addition of ammonium 
 sulphide so as to eliminate any unnecessary separation of 
 sulphur. It forms with copper the complex ion Cu(NH 3 ) 4 ++ . 
 Lead hydroxide remains undissolved in the ammonia, but is 
 converted into sulphide by ammonium sulphide. 
 
 4. If the first precipitate of magnesium ammonium phos-
 
 DETERMINATION OF CARBON IN STEEL 59 
 
 phate be allowed to stand too long, small amounts of metallic 
 sulphides will precipitate with it. Shaking and standing in 
 ice water promotes the separation of magnesium ammonium 
 phosphate and lessens the danger of carrying down sulphides. 
 The second precipitation is made to separate from traces of 
 sulphides and so that the conditions of precipitation may be 
 improved. See note 12, page 27. The ignition of this 
 precipitate must be made with caution. 
 
 DETERMINATION OF CARBON IN STEEL 
 APPARATUS 
 
 The apparatus necessary for the direct combustion of carbon 
 consists of an electrically heated furnace with suitable rheo- 
 stat, an ammeter, a quartz combustion tube, a purifying train, 
 an absorbing train, and a cylinder of compressed oxygen. 
 
 The furnace is preferably of the resistance type made by 
 winding an alundum or clay tube with a resistance wire such 
 as nichrome and surrounding the tube with insulating material 
 to prevent loss of heat. The length of the furnace should be 
 approximately 8 to 12 inches and it should have an opening 
 capable of taking a quartz tube about i| inches in diameter. 
 Suitable furnaces can be made at low cost and several types 
 may be obtained on the market at reasonable cost. A tem- 
 perature of 1000 C. must be obtained without danger of 
 burning out, and must be maintained for a long period of time. 
 A suitable furnace running on a no volt-circuit at 3! to 4 
 amperes should give this temperature. Before placing the 
 furnace in commission, and at intervals of two or three weeks 
 it should be calibrated with a standard thermocouple, so that 
 the operator may know the approximate temperature by the 
 ammeter readings. By regulation of the rheostat, tempera- 
 tures may be obtained and maintained within 5 to 10 degrees 
 of that desired. 
 
 The quartz combustion tube should be about 24 inches in 
 length and i inch in diameter. It should be glazed on the 
 inner surface. Within the tube should be placed at a position 
 corresponding to the front end of the furnace a two-inch roll
 
 60 QUANTITATIVE ANALYSIS 
 
 of ignited asbestos wrapped in nichrome gauze. This should 
 be left permanently in place. A similar roll should be in- 
 serted in the rear of the tube after the introduction of the boat 
 containing the sample. These rolls serve the purpose of 
 protecting the rubber stoppers from the radiated heat. 
 
 In the center of the tube should be kept a cylinder of nickel 
 foil slightly longer than the combustion boat to be used. This 
 serves the purpose of protecting the walls of the quartz tube 
 from oxide of iron formed during combustion. 
 
 The purifying train consists of a 1 2-inch porcelain tube 
 f inches in diameter containing granulated copper oxide, and 
 a furnace capable of heating the tube to a dull red for 4 to 6 
 inches of its length. Attached to this tube is a tower con- 
 taining fresh soda-lime and this in turn is connected with the 
 rear end of the combustion tube by means of a bubble tube 
 containing a small amount of concentrated sulphuric acid. 
 
 At the front end of the combustion tube there is connected 
 a bubble-tube containing granulated zinc and attached to this 
 is a U-tube containing granulated calcium chloride, and be- 
 yond this is a U-tube or one of the many improved absorption 
 tubes filled with granulated soda-lime, or a special mixture 
 made by granulating fresh sodium hydroxide with about four 
 times its volume of asbestos. 2 
 
 The apparatus is shown in Fig. i . It is almost unnecessary 
 to say that the apparatus should be tested to see that it is 
 perfectly tight, and that before using it a blank should be run 
 exactly as in the determination itself. 
 
 METHOD 
 
 Weigh 2-3 grams of finely divided sample, and distribute it 
 evenly over the surface of a clay or alundum boat about half 
 rilled with RR. Alundum. Attach an absorption tube which 
 has previously been weighed with oxygen in it from the blank 
 determination. Place the boat containing the sample within 
 the nickel cylinder in the center of the heated tube, replace 
 the asbestos plug, connect with the oxygen supply and adjust 
 the rheostat so that the temperature will rise rapidly to 950- 
 1000. Regulate the flow of oxygen so that about 2-3 bubbles 
 
 1 The oxygen supplied commercially is so pure and free from hydrocarbon 
 vapors that this part of the train may be omitted. 
 2 J. Ind. 'Eng. Chem., 13, 1052.
 
 DETERMINATION OF CARBON IN STEEL 
 
 61 
 
 per second pass through the sulphuric acid bulb. When com- 
 bustion begins absorption of carbon dioxide will take place. 
 Increase the flow of oxygen slightly during the next one or two 
 minutes. After combustion is complete the flow should be 
 regulated to 2-3 bubbles per second. Continue the passage 
 of oxygen for 5 minutes. Disconnect the absorption tube, 
 stop the supply of oxygen, and place the absorption tube in 
 the balance. The increase in weight represents the carbon 
 dioxide. 
 
 FIG. i 
 
 DETERMINATION OF GRAPHITE IN CAST IRON 
 
 Dissolve i gram of sample in 35 cc. of nitric acid (sp. gr. 
 1.13), boil the solution from 5-10 minutes. Filter the residue 
 on ignited asbestos in a small carbon funnel, wash several 
 times with hot dilute potassium hydroxide solution (sp. gr. 
 i.io), followed by dilute hydrochloric acid, and finally with 
 hot water. Dry at 100 C., transfer asbestos and graphite to 
 a boat and burn as in the determination of total carbon. 
 
 Notes. i. Carbon exists in irons (> 1.70 per cent C.) and 
 steels ( < 1.70 per cent C.) in four forms; Iron Carbide, known 
 as Cementite; Hardening Carbon; Graphite; Graphitic Tem- 
 per Carbon.
 
 62 QUANTITATIVE ANALYSIS 
 
 Iron and carbon unite to form a carbide, Fe 3 C, which is 
 capable of being held in solid solution (austenite) the limiting 
 value of which is 1.70 per cent carbon. It occurs in all slowly 
 cooled (annealed) carbon steels, and in irons which have been 
 chilled in passing from the liquid to the solid state. In hypo- 
 eutectoid (< 0.85 per cent C.) steels, its formation takes place 
 on passing slowly through the critical temperature, Ari, at 
 680. At this temperature austenite decomposes into a iron 
 and cementite, after having previously thrown out some of its 
 iron (ferrite) at the critical temperature, Ar 3 , which is pro- 
 gressively lowered with increase of carbon. 
 
 In hyper-eutectoid (> 0.85 per cent C.) steels the initial 
 temperature of transformation of austenite into cementite 
 rises with increasing carbon and reaches a maximum at 1.70 
 per cent carbon when cementite separates at 1130. The end- 
 ing of the transformation takes place at the same temperature 
 for all compositions (0.85-1.70 per cent C.) at the Ari point. 
 
 In eutectoid steels (0.85 per cent C.) complete transforma- 
 tion of austenite takes place at Ari into an intimate mixture 
 of its two constituents, iron (ferrite) and cementite, this mix- 
 ture being known as pearlite. Thus hypo-eutectoid steels 
 contain excess ferrite with pearlite, and hyper-eutectoid steels 
 excess cementite and pearlite. 
 
 In irons of hypo-eutectic composition (1.70-4.30 per cent 
 C.) cementite separates at 1130 as part of the eutectic of 
 austenite and cementite, the austenite subsequently trans- 
 forms at Ari into pearlite. In hyper-eutectic (> 4.3 per cent 
 C.) irons the initial separation of cementite rises progressively 
 with increase of carbon and the final separation takes place 
 at 1130. 
 
 These facts are shown in the iron carbon equilibrium dia- 
 gram, Fig. 2. 
 
 On slow cooling of irons, however, the above conditions are 
 not realized, as the separated iron carbide almost immediately 
 decomposes into graphite, 
 
 Fe 3 C -> C (graphite) + 3Fe 
 
 and the more slowly cooled the metal, the more complete is 
 this reaction. 
 
 Dissolved in dilute hydrochloric or sulphuric acids, steels 
 containing iron carbide evolve hydrocarbons. In dilute cold
 
 DETERMINATION OF CARBON IN STEEL 
 
 nitric acid iron carbide separates as a brown flocculent residue 
 which goes into solution above 80 and gives a color propor- 
 tional to the amount which is present (Colorimetric Method). 
 The elements manganese, sulphur and chromium favor 
 carbide formation, and silicon, phosphorus, and aluminum 
 oppose it and promote graphite formation. Double carbides 
 are formed with manganese, chromium, tungsten, molybdenum 
 and vanadium. 
 
 1520 
 
 900' 
 
 Ar, Final 
 FerriW 
 
 JSfO. 
 
 FIG. 2 
 
 The effect of carbide carbon upon the physical properties of 
 steel is shown diagrammatically in Fig. 3- 1 
 
 Hardening carbon is formed when a steel or iron is quenched 
 from any temperature at which austenite exists, i.e., from 
 above Aci. It is unstable and tends to pass back slowly on 
 tempering (heating below Aci), and goes over completely to 
 carbide carbon by heating above Aci and cooling slowly. 
 
 It is decomposed by hot dilute acids. In dilute, cold nitric 
 acid it at first separates as an insoluble residue and then goes 
 into solution on shaking. The solution is colored in propor- 
 tion to the amount of hardening carbon present. 
 1 J. H. Nead, Am. Inst. Min. Eng. (1916), 2341.
 
 6 4 
 
 QUANTITATIVE ANALYSIS 
 
 Graphitic temper carbon is produced by heating irons rich 
 in cementite which decomposes at about 1000 into iron and 
 
 EFFECT OF CARBON ON PHYSICAL PROPERTIES 
 Heat Treatment -Annealed just above Ac 8 
 
 A Tensile Strength 
 
 B Elastic Limit Yield Point 
 
 C Elongation 
 
 D Contraction of Area 
 
 .80 1.00 
 
 Per cent Carbon 
 
 FIG. 3 
 
 1.20 
 
 graphitic temper carbon (Malleableizing Process), and it is 
 sometimes formed in annealing high carbon steels. 
 
 Both graphite and graphitic temper carbon are unaffected 
 by boiling dilute acids. Boiling concentrated nitric acid 
 oxidizes them only very slowly. In oxygen, the amorphous
 
 DETERMINATION OF MANGANESE IN STEEL 65 
 
 graphitic temper carbon is burned more readily than the 
 crystalline graphite. 
 
 2. The carbon in plain carbon steels is completely burned 
 at a temperature of 950-1000 C. In some alloy steels 
 chrome-nickel, chrome- tungsten, chrome- vanadium com- 
 plete combustion does not take place below 1050-1100 C., and 
 on this account a flux, Pb 3 04, is sometimes used, but it is pref- 
 erable to burn direct if the temperature can be obtained. 
 Another method employed to burn the alloy steels is to mix 
 with them a known weight of a low carbon iron, such as Armco 
 Ingot Iron. 
 
 3. Instead of using potassium hydroxide as absorbent, 
 some workers prefer soda-lime, or barium hydroxide. If the 
 latter is used the determination may be finished volumetrically 
 by (i) filtering the precipitated barium carbonate and titrating 
 the excess barium hydroxide, or (2) by dissolving the filtered 
 barium carbonate l in excess standard acid and titrating for 
 the excess. 
 
 4. The whole apparatus should be tested to see that there 
 is no leakage of gas before proceeding with any determination. 
 After having assembled the apparatus, and having been 
 assured of its being tight, a blank determination should be 
 made exactly as in the regular determination. The increase 
 in weight of the absorption tube should not amount to more 
 than 3 milligrams. 
 
 5. There is always difficulty in weighing accurately any 
 piece of glass having a large surface area on account of the 
 variations in the amount of condensed moisture. On this 
 account it is always necessary to verify the weight of the 
 potash bulb before beginning a day's work, and it is never 
 allowable to assume the last weight of the day before. The 
 bulb should always be wiped with a clean, dry cloth after 
 handling and before being placed in the balance case. 
 
 Some of the soda-lime bulbs have small surface area and 
 have a very considerable advantage in this respect over the 
 potash bulbs. 
 
 DETERMINATION OF MANGANESE IN STEEL 
 
 VOLHARD METHOD 
 
 Standardization of Potassium Permanganate Solution. 
 Standardize the potassium permanganate solution in the follow- 
 
 1 Cain, J. Ind. Eng. Chem., 6, 465.
 
 66 QUANTITATIVE ANALYSIS 
 
 ing manner. From a burette measure into a 500 cc. Florence 
 flask several portions of about 25 cc. each of a manganous 
 chloride solution, 1 the value of which has been very accurately 
 established according to the method given on page 17. Add 
 20 cc. zinc nitrate solution (300 grams Zn(NO 3 ) 2 ,6H 2 O per 
 liter), dilute with hot water to about 300 cc., add two or three 
 drops of dilute nitric acid, and heat to near the boiling point. 
 Now add rapidly nearly as much permanganate solution as it 
 is thought will precipitate all of the manganese. Rotate the 
 solution thoroughly; and then add, cautiously, enough per- 
 manganate solution to give a permanent color to the super- 
 natant liquid. 
 
 METHOD 
 
 For steels containing about one per cent of manganese 
 weigh a three-gram sample; for higher percentages of manga- 
 nese use a correspondingly lower weight. Place the borings 
 in a porcelain casserole and treat with 30 cc. of nitric acid 
 (sp. gr. i. 20). When solution of the metal is complete raise 
 the watch glass by supporting it on a glass triangle, and 
 evaporate to dryness. Then remove the watch glass and 
 carefully bake the contents of the dish until no more oxides 
 of nitrogen are evolved. After the casserole has cooled, add 
 20 cc. concentrated hydrochloric acid and heat until all ferric 
 oxide is converted into chloride. (If cast iron is being analyzed, 
 the solution should be evaporated to dryness, taken up in 
 hydrochloric acid, and the graphite and dehydrated silica 
 filtered off.) Concentrate the solution to remove the excess 
 acid and transfer it to a 500 cc. calibrated flask, and dilute to 
 about 200 cc. Now add, cautiously, strong sodium carbonate 
 solution a little at a time until the precipitate just redissolves 
 and the solution becomes colored deep red. 2 Then add 
 slowly in small amounts zinc oxide suspended in water, 
 shaking the flask after each addition, and continue this 
 process until a point is reached where the ferric hydroxide 
 
 1 Six grams of the crystallized salt dissolved in one liter of water. 
 1 Compare Spathic Iron Ore Procedure.
 
 DETERMINATION OF MANGANESE IN STEEL 67 
 
 suddenly coagulates. Allow the precipitate to settle, and fill 
 the flask with water exactly to the graduation mark. Mix 
 the contents of the flask thoroughly by pouring back and forth 
 into a dry beaker several times. Allow the mixture to settle, 
 and then filter through a four-inch dry filter paper. Collect 
 250 cc. of the filtrate, after the rejection of the first five cubic 
 centimeters, in a graduated flask. Transfer this solution to 
 a 500 cc. Florence flask, add two drops of dilute nitric acid, 
 heat to near the boiling point, and titrate with potassium 
 permanganate to pink color, exactly as in the standardization. 
 
 Notes. i. Manganese exists in ordinary steels in the form 
 of a solid solution with iron. Small amounts are also present 
 in the forms of manganese sulphide, MnS, 1 and manganese 
 silicate, M^SiaOs, 2 which form part of the slag found in steels. 
 In manganese steels, ferro-manganese and spiegeleisen there 
 is, in addition to the solid solution, some manganese as carbide, 
 Mn 3 C2, or as a double manganese-iron carbide. Manganese 
 is primarily put in steel for the purpose of deoxidizing and 
 desulphurizing the melt, so as to prevent "red shortness" or 
 brittleness during the forging heats. Some excess always 
 appears in the finished product, and in commercial steels is 
 found in amounts varying from 0.2 to about i.oo per cent. In 
 such amounts it increases the strength and hardness and 
 decreases somewhat the ductility. These effects are more 
 pronounced in high carbon steels than in low carbon material. 
 
 In metals containing manganese in excess of i per cent the 
 hardness and brittleness increase rapidly and reach a maxi- 
 mum between 5 and 6 per cent. In such products the hard- 
 ness and brittleness are so great that it is impossible to machine 
 them with any tool. Beyond 7 per cent of manganese there 
 is a progressive increase in the ductility, and steels containing 
 about 13 per cent manganese are characterized by great 
 strength, toughness and high ductility. Such steels are known 
 as manganese steels and have found extensive application in 
 the industries. 
 
 Manganese lowers the critical temperature, An. With in- 
 
 1 Osmond and Werth, Annales des Mines, 1885. 
 8 Stead, Iron and Steel Magazine, 9, 105.
 
 68 QUANTITATIVE ANALYSIS 
 
 creasing percentages of manganese the critical temperature 
 finally falls below the ordinary room temperature, and self- 
 hardening steels result. See notes on Carbon, page 61. 
 
 2. The principle of the Volhard method is based upon the 
 reaction 
 
 2 KMnO 4 + 3 MnSO 4 + 2H 2 O = 5 MnO 2 + 2 KHS0 4 + H 2 S0 4 . 
 That is, when potassium permanganate is added to a neutral 
 manganous salt, all of the manganese in the latter is oxidized 
 to the dioxide and the permanganate is reduced to the dioxide. 
 The reaction never takes place exactly as represented in the 
 equation, except in the presence of a few drops of nitric acid 
 and the salt of some heavy metal. In the absence of these 
 substances the manganese is not oxidized completely to Mn0 2 . 
 This is provided for in the standardization by the addition of 
 zinc nitrate solution, and in the determination by the reaction 
 of the ferric sulphate with the zinc oxide. 
 
 Fe 2 (S0 4 ) 3 + 3 ZnO + 3 H 2 O - 2 Fe(OH) 3 + 3 ZnSO 4 . 
 
 3. Since two molecules of potassium permanganate will 
 oxidize three atoms of manganese, and since, in acid solutions, 
 the same amount of permanganate will oxidize ten atoms of 
 iron, it follows that icFe is equivalent to 3 Mn. Having 
 established the value of the permanganate solution in terms of 
 iron or sodium oxalate, it would be a simple matter to calculate 
 its value in terms of manganese. This calculation should be 
 made so as to check the value obtained by titrating against 
 manganous sulphate, but the latter value should be used on 
 account of its having been obtained under exactly the same 
 conditions as in the actual determination. The difference 
 between the two values is usually extremely small. 
 
 4. In the standardization, and also in the actual titration, 
 it is impossible to add an excess of potassium permanganate 
 and then to titrate back with manganous sulphate solution, 
 on account of the fact that manganese dioxide reacts catalyti- 
 cally with permanganic acid, 1 which liberates oxygen accord- 
 ing to the following equation: 
 
 2 HMn0 4 = H 2 + 2 MnO 2 + 3 O, 
 
 thus losing three-tenths of the oxygen which should have been 
 used in oxidizing manganese sulphate. 
 
 1 Morse, Hopkins and Walker, Am. Chem. /., 18, 401.
 
 DETERMINATION OF MANGANESE IN STEEL 
 
 5. The Volhard method is particularly well adapted for 
 the determination of manganese in those products where the 
 manganese content is high, i.e., above i per cent. With lower 
 content than 0.50 per cent there is difficulty in getting a satis- 
 factory separation of manganese dioxide without unduly 
 increasing the weight of sample. 
 
 The method is not applicable to the analysis of chrome steels 
 as the zinc oxide may precipitate chromium only incompletely, 
 in which case it might be oxidized by the potassium perman- 
 ganate. Cobalt and vanadium interfere in a similar manner. 
 
 In steels rich in nickel it is difficult to obtain a satisfactory 
 end point. 
 
 WILLIAMS METHOD 1 
 
 Weigh two three-gram samples into 750 cc. Erlenmeyer 
 flasks and cover with 100 cc. 
 nitric acid (sp. gr. 1.42). 
 Place the flask covered with 
 a small watch glass on the 
 steam table and allow it to 
 remain until solution is com- 
 plete. As the steel becomes 
 passive on treatment with 
 strong nitric acid, it is of 
 little advantage to try to 
 hurry the solution by boil- 
 ing. When the steel is com- 
 pletely dissolved, add three 
 grams of potassium chlorate, 
 and after five minutes add 
 a second portion of three 
 grams, heating on the steam 
 table during this time. Af- 
 ter ten minutes remove from 
 the steam table and heat for FlG 
 
 five minutes in a hood with 
 good draught until the excess chlorine dioxide is driven off. 
 
 1 Trans. Am. Inst. Min. Eng., 10, 100.
 
 70 QUANTITATIVE ANALYSIS 
 
 Cool, and filter without undue delay on asbestos which has 
 been previously treated with dilute hydrochloric acid and 
 washed with water until the asbestos does not reduce dilute 
 potassium permanganate solution. The asbestos should be 
 placed on a platinum spiral, a few pieces of glass tubing, or 
 some glass beads in the bottom of a manganese funnel. When 
 the excess of nitric acid has drained off, place a clean beaker 
 under the funnel, begin to wash with water, and continue 
 until litmus paper shows no test for nitric acid. Then wash 
 the manganese dioxide and asbestos back into the flask in 
 which the precipitation was made. 
 
 Add 50 cc. ferrous sulphate solution (50 cc. H^SCX sp. gr. 
 1.84; 5 grams FeSO^H^O; 450 cc. H2O) the value of which, 
 in terms of potassium permanganate, has recently been es- 
 tablished. Dilute to 200 cc. with water. Shake until the 
 manganese dioxide has completely dissolved, and titrate to 
 pink color with potassium permanganate. 
 
 Notes. i. The principle of this method is based upon the 
 two reactions: 
 
 Mn(N0 3 ) 2 + 2 HC10 3 = MnO 2 + 2 C1O 2 + 2 HNO 3 . 
 
 2 FeS0 4 + 2H 2 SO 4 = MnSOi + Fe 2 (SO 4 ) 3 + 2H 2 0. 
 
 The oxidation of manganous nitrate to dioxide will take 
 place only in concentrated nitric acid solution. Dilute nitric 
 acid may be used to dissolve the steel, but, if used, the solution 
 must be concentrated before adding the potassium chlorate, 
 otherwise incomplete oxidation will result. 
 
 Since one molecule of manganese dioxide will oxidize two 
 molecules of ferrous sulphate it is evident that one manganese 
 is equivalent to two atoms of iron. This ratio is used in the 
 calculations, and the iron value of the potassium permanganate 
 is taken as the standard. The manganese value of the potas- 
 sium permanganate in the Williams method should be five- 
 thirds of the strength found in the Volhard method. 
 
 2. This method is well adapted to the determination of 
 manganese in steels containing not more than 0.75 per cent of 
 that element. With higher amounts the filtration of the 
 dioxide becomes more and more difficult unless the size of the
 
 DETERMINATION OF MANGANESE IN STEEL 71 
 
 sample is correspondingly decreased, and, further, there is an 
 increasing error with an increase in the amount of manganese 
 dioxide precipitated. As the precipitate is not absolutely 
 pure MnO 2 , but contains slightly less oxygen than represented 
 by the formula, it will be seen that the error increases with the 
 amount of dioxide thrown out of solution. 
 
 As the precipitate loses oxygen slowly on standing, it is 
 important to filter as soon as possible after precipitation. 
 
 The method is not especially well adapted to cast irons 
 on account of the difficulties of filtration in the presence of 
 graphite and gelatinous silica. 
 
 3. Care should be exercised in the addition of potassium 
 chlorate. If added to a boiling solution dangerous explosions 
 may result. 
 
 4. Some asbestos, either because of its containing ferrous 
 iron or organic matter, reduces potassium permanganate. It 
 is, therefore, necessary to test each new lot to see whether or 
 not it will react with very dilute permanganate solution. 
 This difficulty might be obviated by passing the reducing 
 solution, ferrous sulphate, or oxalic acid, through the funnel 
 containing the manganese dioxide, but it is much simpler to 
 avoid the washing this would involve by putting the dioxide 
 and asbestos into the reducing solution. 
 
 5. Almost any standard reducing solution may be used to 
 react with manganese dioxide. Ferrous sulphate and oxalic 
 acid are the most commonly used. Ferrous sulphate in acid 
 solution is fairly stable and has the very distinct advantage of 
 reacting in the cold. Oxalic acid reacts only when warmed and 
 has the disadvantage of changing concentration, unless steril- 
 ized, because of organic growth in the solution. 
 
 Whatever the reducing solution used, a blank should be 
 made before each day's work against the standard potassium 
 permanganate. 
 
 6. It is impossible to ignite the precipitated dioxide and 
 weigh as Mn 3 O4, because of the occluded iron and the indefinite- 
 ness of composition of the ignited oxide. Provision may be 
 made for the separation of the iron, and the determination may 
 be finished as given on page 17. 
 
 7. The method is applicable in the presence of all the con- 
 stituents of the more common steels.
 
 72 QUANTITATIVE ANALYSIS 
 
 THE BISMUTHATE METHOD FOR MANGANESE 
 
 Weigh into a 250 cc. Erlenmeyer flask about one gram of 
 borings and treat with 50 cc. of nitric acid 1 (sp. gr. 1.13). 
 Heat until solution is complete, allow to cool, and then add 
 about 0.5 gram of sodium bismuthate. Rotate the contents 
 of the flask, and heat until the pink color has disappeared. 
 If the solution shows precipitated manganese dioxide, add 
 crystals of ferrous sulphate, free from manganese, until it 
 becomes clear. Heat for two minutes to remove oxides of 
 nitrogen, and cool to about 15. Now add 2 to 3 grams of 
 sodium bismuthate, and agitate the contents of the flask for 
 several minutes. Dilute with 50 cc. of three per cent nitric 
 acid, 2 and filter through asbestos, as prescribed in the previous 
 method, into a 300 cc. Erlenmeyer flask. Wash the asbestos 
 with 50 to 100 cc. of cold three per cent nitric acid. Run into 
 this solution 50 cc. of the ferrous sulphate solution which was 
 used in the Williams method, and titrate back to pink color 
 with standard potassium permanganate. 
 
 The value of the ferrous sulphate solution in terms of 
 potassium permanganate must be determined in the following 
 manner. Measure into a 250 cc. Erlenmeyer flask 50 cc. of 
 cold nitric acid (sp. gr. 1.13), add about 0.5 gram of sodium 
 bismuthate, agitate, and filter through asbestos. Dilute with 
 50 cc. of cold three per cent nitric acid, run in 50 cc. of ferrous 
 sulphate solution, and titrate with permanganate solution to 
 pink color. 
 
 Having determined the value of the permanganate solution 
 in terms of ferrous sulphate, the manganese in the sample is 
 represented by the difference between the amounts of per- 
 manganate solution actually used in the determination and 
 in the titration of a volume of ferrous sulphate equivalent to 
 that used in the determination. 
 
 Notes. i. The method originated with Schneider, 3 who 
 used bismuth tetroxide as the oxidizing agent. This oxide is 
 
 1 One volume HNOs sp. gr. 1.42; three volumes H 2 O. 
 * 30 cc. of HNOs (sp. gr. 1.42) to one liter of water. 
 3 Dingier' s Polytech. J., 269, 224.
 
 DETERMINATION OF MANGANESE IN STEEL 73 
 
 very difficult to prepare free from chlorides, and, as traces of 
 hydrochloric acid interfere with the sharpness of the end point, 
 it was abandoned by Reddrop and Ramage, 1 who proposed the 
 use of sodium bismuthate. More recently the method has 
 been studied and recommended by Blair 2 and by Blum. 3 
 
 Sodium bismuthate is supposed to be the sodium salt of the 
 hypothetical bismuthic acid, HBiOa. The product sold under 
 this name is probably of indefinite and variable composition. 
 
 2. The method is based upon the fact that sodium bis- 
 muthate oxidizes a manganous salt in the cold to permanganic 
 acid, which can be reduced by ferrous sulphate the value of 
 which is known in terms of standard potassium permanganate. 
 Ferrous sulphate is not readily acted upon by cold nitric acid 
 and can consequently be used to reduce the permanganic acid 
 without introducing an error. 
 
 Lead peroxide oxidizes manganese to permanganic acid in 
 hot solution, but the results are not reliable when more than 
 two per cent of manganese is present. In hot solution per- 
 manganic acid oxidizes some of the manganous nitrate to 
 dioxide. 
 
 3. Since two molecules of permanganic acid will give up 
 five atoms of oxygen, it follows that it will oxidize ten atoms 
 of iron. 
 
 Hence 
 
 2HMnO 4 = loFe, 
 iMn = 5Fe. 
 
 From these equations it can be seen that the value of the 
 standard potassium permanganate may readily be obtained in 
 terms of manganese from the iron standard. 
 
 The value may be determined directly in terms of man- 
 ganese by using a definite volume of standard manganous 
 sulphate solution. This solution may be oxidized and deter- 
 mined by the bismuthate process, exactly as described in the 
 procedure. 
 
 4. The addition of a small amount of sodium bismuthate to 
 the hot solution is to oxidize the hydrocarbons formed by the 
 decomposition of iron carbide. The excess bismuthate is 
 
 1 J. Chem. Soc., 67, 268. 
 
 2 /. Am. Chem. Soc., 26, 33. 
 
 8 J. Am. Chem. Soc., 34, 1379.
 
 74 QUANTITATIVE ANALYSIS 
 
 rapidly decomposed by the hot nitric acid, and then the latter 
 reacts with the permanganic acid, precipitating manganese 
 dioxide. This in turn is reduced by the addition of a small 
 amount of ferrous sulphate. , 
 
 5. It is stated that chromium in small amounts does not 
 interfere with this method, but this is contrary to the author's 
 experience. For accurate work it is necessary to remove iron 
 and chromium with zinc oxide, and to oxidize an aliquot part of 
 the filtrate with sodium bismuthate. . 
 
 Tungsten steels, which decompose readily in nitric acid, may 
 ibe heated to precipitate the tungstic acid, and, after filtration, 
 may be carried through as ordinary steels. If solution of the 
 ;steel is troublesome, it will be necessary to use aqua regia and 
 .to remove subsequently every trace of hydrochloric acid by 
 repeated evaporation with nitric acid. 
 
 6. There must be no loss of time between the filtration and 
 the addition of ferrous sulphate, as permanganic acid gradu- 
 ally decomposes on standing, and the warmer the solution is 
 the more rapidly does decomposition take place. 
 
 7. The bismuthate method combines extreme accuracy with 
 great rapidity and simplicity of manipulation. It is capable 
 of being used, without loss of accuracy, for any percentage of 
 manganese. In this respect it has a distinct advantage over 
 the Volhard and Williams methods. 
 
 8. A colorimetric method, in which oxidation to perman- 
 ganic acid is brought about by sodium bismuthate, ammonium 
 persulphate, or lead peroxide, is extensively used for rapid 
 work in the steel mills. The color of the permanganic acid 
 formed is compared with that produced in steels of known 
 manganese content. 
 
 9. A method, very similar in its details and principle, is in 
 extensive use. Instead of using sodium bismuthate as oxidiz- 
 ing agent, ammonium persulphate 1 is used. While, in general, 
 not so accurate, it possesses the distinct advantage that the 
 excess reagent is destroyed by heat, thus avoiding one filtration. 
 
 10. All methods involving oxidation of manganese to per- 
 manganic acid are not applicable to cast irons, unless the 
 separated carbon is filtered off before oxidation takes place. 
 
 1 Walters, Proc. Eng. Soc., Western, Penna., 17 (1901), 257; Chem. News, 
 4, 239.
 
 DETERMINATION OF PHOSPHORUS IN IRON AND STEEL 75 
 
 DETERMINATION OF PHOSPHORUS IN IRON AND 
 STEEL 
 
 STANDARDIZATION OF POTASSIUM PERMANGANATE AND THE 
 
 USE OF THE JONES REDUCTOR 
 
 Dissolve 5 grams of potassium permanganate in two liters 
 of water, and filter this solution through some washed asbes- 
 tos into a clean bottle. It is advisable to keep this solution 
 in a cool place and protected from the light. 
 
 Weigh several portions of pure iron wire of about 0.20 
 gram each into 125 cc. Erlenmeyer flasks, and pour into each 
 flask 30 cc. of water and 10 cc. of concentrated sulphuric 
 acid. Cover with a watch glass and heat until solution is 
 complete. Add enough strong permanganate solution, drop 
 by drop, until the color of the permanganate disappears only 
 slowly. If an excess be added, it is necessary to reduce by a 
 drop of sulphurous acid and to boil off this excess completely. 
 
 Having previously gotten ready a Jones reductor, by 
 washing with hot dilute sulphuric acid followed by hot water, 
 attach the reductor to the filter pump, and, in making blanks 
 or in any determination, proceed exactly as here described. 
 
 Pour 100 cc. of hot dilute sulphuric acid (made by adding 
 25 cc. of concentrated sulphuric acid to one liter of water) 
 into the funnel, and open the stopcock. When only a little 
 remains in the funnel add the solution to be reduced. This 
 solution should also be hot. Have ready 250 cc. of hot, 
 dilute sulphuric acid, and when the solution to be reduced 
 has nearly passed out of the funnel wash out, with this acid, 
 the beaker containing the iron or molybdenum solution, as 
 the case may be, and pass the whole of the 250 cc. of acid 
 through the reductor. Follow this by 100 cc. of hot water. 
 Disconnect the flask and titrate with potassium permanganate 
 solution. 
 
 Before any series of determinations, a blank must be 
 made on the reductor. In this case a mixture of 10 cc. con- 
 centrated sulphuric acid, 5 cc. dilute ammonia, and 50 cc. of 
 water is to be passed through the reductor, preceded and
 
 76 QUANTITATIVE ANALYSIS 
 
 followed by the dilute acid as described above. The amount 
 of potassium permanganate used in this operation, which 
 usually is about 0.3 cc., must be subtracted from the total 
 amount taken in each determination. Having made a blank, 
 this value will serve for a series of determinations, but a new 
 blank should be made before each new series. 
 
 METHOD 
 
 For steels weigh 2 grams, and for irons weigh a i gram 
 sample into a 250 cc. Erlenmeyer flask. Add 100 cc. of 
 nitric acid (sp. gr. 1.13) which has been prepared by mixing 
 one part of nitric acid (sp. gr. 1.42) with three parts of water, 
 and which has been tested by means of a Westphal balance. 
 Cover the flask with a small watch glass, and heat the solu- 
 tion until the oxides of nitrogen have boiled off. Then add 
 10 cc. of potassium permanganate solution (15 grams of the 
 crystals to one liter of water), or, in the case of irons or steels 
 rich in iron carbide, enough of the solution to produce a pink 
 color. Boil until the pink color disappears and then remove 
 from the heat. Now add at intervals crystals of ferrous sul- 
 phate free from phosphorus until all of the manganese dioxide 
 is dissolved. Boil two minutes longer and then cool the flask. 
 Add cautiously 40 cc. of dilute ammonia (sp. gr. 0.96) and 
 rotate the flask until the precipitated ferric hydroxide is dis- 
 solved. When the solution has cooled to about 35 C. add 
 40 cc. of a recently filtered solution of ammonium molyb- 
 date, 1 close the flask with a solid rubber stopper, and shake 
 for five minutes. Allow the precipitate to settle for a few 
 minutes, and filter through a 9 cm. paper which fits the sides 
 of the funnel perfectly. Wash the precipitate several times 
 by decantation, and then completely on the filter with a solu- 
 tion of acid ammonium sulphate (made by adding 25 cc. of 
 
 1 MOLYBDATE SOLUTION. To ioo grams of pure molybdic oxide (MoO 3 ) 
 which has been stirred up with 400 cc. of cold, distilled water add 80 cc. of 
 concentrated ammonia. Filter, and pour the filtrate slowly, with constant 
 stirring, into a solution of 400 cc. nitric acid (sp. gr. 1.42) in 600 cc. of water. 
 After the addition of 50 milligrams of microcosmic salt, allow to stand for 
 twenty-four hours, and filter. If pure molybdic oxide is not available, 125 
 grams of ammonium molybdate may be substituted.
 
 DETERMINATION OF PHOSPHORUS IN IRON AND STEEL 77 
 
 strong sulphuric acid to 15 cc. of strong ammonia in one 
 liter of water). 
 
 Test the wash water for molybdenum by allowing about 
 ten drops to run into a small specimen tube containing one 
 drop of yellow ammonium sulphide. Compare the color in 
 this tube with a similar tube which has been treated with 
 a like amount of acid wash water and a drop of the poly- 
 sulphide. Continue the washing until all of the molybdenum 
 is removed. 
 
 Now pour 5 cc. of ammonia (sp. gr. 0.90) and 20 cc. of 
 water into the flask to dissolve any ammonium phospho- 
 molybdate adhering to the walls; pass this solution through 
 the filter and allow the nitrate to run into a No. 2 beaker. 
 Wash out the flask and the filter with about 75 cc. of water. 
 To this solution add 10 cc. of concentrated sulphuric acid 
 and pass through the reductor exactly as previously described, 
 taking the precaution not to allow any air to enter the re- 
 ductor. This solution is preceded and followed by the dilute 
 sulphuric acid and water, just as in testing the reductor. 
 When the water has finally been drawn through with the ex- 
 ception of a small amount in the funnel, the flask is detached 
 and the reduced solution is titrated at once with the standard 
 potassium permanganate to a permanent pink color. 
 
 Subtract from the reading of the burette the blank for the 
 reductor and calculate, assuming a reduction of the molyb- 
 denum to Mo24O 37 by the zinc, and oxidation to Mo0 3 by 
 the permanganate. 
 
 THE FERRIC ALUM METHOD FOR PHOSPHORUS 1 
 
 Proceed exactly as in the preceding method up to the 
 point of passing the sulphuric acid solution through the re- 
 ductor. Instead of passing the reduced solution into the 
 empty flask, place in the flask 50 cc. of ferric alum solu- 
 tion, 2 and allow the reduced solution to run through a tube 
 
 1 This method was worked out in the laboratory of the Penna. R.R. by 
 Dudley and Pease but was never published by them. It has been in successful 
 use in this laboratory for many years. 
 
 2 100 grams of FdCSC^MNHOjSO^^jO dissolved in 1000 cc. H 2 O, to which 
 has been added 25 cc. H 2 SO 4 (sp. gr. 1.84). To this solution add 40 cc. H 3 PO 4 .
 
 78 QUANTITATIVE ANALYSIS 
 
 extended below the surface of the liquid. In all other respects 
 the two methods are identical. The purpose of the ferric 
 alum is to oxidize immediately the molybdenum, which has 
 been reduced in the reductor to a form corresponding to the 
 oxide Mc^Os, to the higher oxide MoO 3 . The iron which has 
 been reduced in this reaction is then titrated with potassium 
 permanganate solution, and is a direct measure of the molyb- 
 denum reduced. The calculation is based on the assumption 
 that the molybdenum is reduced to M^Oa, and the results 
 should agree with the other method in which reduction to 
 M024O37 was assumed. 
 
 A blank determination must be made with the ferric alum 
 solution in the flask, using the same amounts of reagents as 
 in the regular procedure. 
 
 Notes. i. The determination of phosphorus is an indirect 
 method dependent upon its oxidation to phosphoric acid and 
 precipitation in the form of ammonium phospho-molybdate, 
 the reduction of the molybdenum by means of zinc, and the 
 subsequent oxidation of the molybdenum by means of potas- 
 sium permanganate. 
 
 2. The phosphorus exists in most steels as a solid solution 
 of iron phosphide (FesP) in iron. When present, as it may 
 be in some irons, in amounts greater than 1.7 per cent it 
 separates as free FesP, and the iron will contain the solid 
 solution, and also free iron phosphide. In wrought iron, 
 and probably also in some low carbon Bessemer steels, phos- 
 phorus may also exist in the form of a phosphate. 
 
 3. The preparation of potassium permanganate which is 
 to be used for a series of determinations should be done with 
 extreme care. The crystals of the salt are never absolutely 
 pure, but invariably contain a small amount of manganese 
 dioxide. If this is not filtered off, it reacts with the potas- 
 sium permanganate, giving off oxygen and forming more 
 manganese dioxide. The presence of manganese dioxide 
 accelerates the decomposition, and when it is once formed the 
 decomposition takes place rapidly. A solution carefully 
 prepared will keep almost indefinitely. 
 
 4. It is impossible to obtain absolutely pure iron wire. 
 The determination of the actual amount of iron in any sample
 
 DETERMINATION OF PHOSPHORUS IN IRON AND STEEL 79 
 
 may be done in either of two ways: the impurities in the 
 sample or the actual amount of iron may be determined. 
 The former method involves much skill and experience; the 
 latter method is the simpler and is, perhaps, more accurate. 
 When using the latter method, the potassium permanganate 
 is titrated first against pure electrolytic iron, 1 and after this 
 has been carefully standardized this solution is used to deter- 
 mine the actual amount of iron in a sample of pure piano wire. 
 There are many other methods for the standardization of 
 potassium permanganate solutions. The author prefers the use 
 of sodium oxalate to any of the others. This salt contains no 
 water of crystallization, is very stable, may easily be prepared 
 pure, and may also be used as a standard for alkalimetric work. 2 
 
 5. The addition of strong potassium permanganate to the 
 solution of iron is rendered necessary because of the pres- 
 ence of a small amount of carbon. The decomposition of 
 iron carbide by sulphuric acid gives rise to hydrocarbons 
 which are held in solution and which would affect the titration, 
 if not previously oxidized. 
 
 6. A blank on the reductor is a necessity on account of 
 the fact that all zinc contains some iron which is given up 
 to the dilute acid passing through the column of zinc. The 
 necessity, therefore, of always using exactly the same amounts 
 of reagents is apparent. The time factor is also apparent. 
 It is advisable to measure accurately all reagents, and in 
 each case to allow approximately the same amount of time 
 for each reaction. 
 
 7. The preparation of nitric acid of exact sp. gr. 1.13 is 
 important because in acid of this strength the steel is most 
 readily soluble, the phosphorus is oxidized as well as in acids 
 of greater strength, the silica is not separated in the insolu- 
 ble form, and it is very important to bring down the yellow 
 precipitate always in the same strength solution. Especially 
 on account of this last reason, the solutions should never 
 be allowed to boil longer than the time indicated in the direc- 
 tions. 
 
 8. Boiling off the oxides of nitrogen seems to be necessary 
 
 1 For details in carrying out this method, consult Treadwell-Hall's Analyti- 
 cal Chemistry, Vol. II. 
 
 1 See Spathic Iron Ore Procedure, p. 21.
 
 80 QUANTITATIVE ANALYSIS 
 
 as these reduce potassium permanganate, which is to be 
 added subsequently to this operation, and also on account 
 of the solubility of the ammonium phospho-molybdate in 
 solutions containing nitrous acid. 
 
 9. Potassium permanganate is added to complete the 
 oxidation of the phosphorus to the form of ortho-phosphoric 
 acid and to destroy the carbonaceous material. When car- 
 bide of iron is boiled with dilute nitric acid certain organic 
 compounds are formed, which, if allowed to remain, would 
 hold some of the yellow precipitate in solution. 
 
 For most steels 10 cc. of permanganate solution is more 
 than enough to oxidize the phosphorus and destroy the car- 
 bonaceous material. It is advisable, however, to use just 
 this amount in each case except in high carbon steels or in 
 white irons, which may require more than the specified amount 
 to complete the oxidation. 
 
 10. Instead of using pure ferrous sulphate for the removal 
 of the manganese dioxide, many other substances, such as 
 sulphurous acid, oxalic acid, sugar, sodium thiosulphate, etc., 
 have been recommended, but it is believed that for student 
 work ferrous sulphate is the most satisfactory. 
 
 n. The addition of ammonia is simply to form some 
 ammonium nitrate in which the precipitate is more insoluble. 
 If by some oversight too much nitric acid had been boiled off 
 during the operation, the ammonia will produce a permanent 
 precipitate of ferric hydroxide. If this should happen, it is 
 necessary to add more nitric acid of sp. gr. 1.13. 
 
 12. The composition of ammonium phospho-molybdate 
 depends very largely upon the strength of solutions from 
 which it is precipitated and upon the temperature of pre- 
 cipitation. Produced under the conditions described above 
 it may be represented by the formula (NH 4 ) 3 P04, laMoOs. 
 The ratio of phosphorus to molybdenum is always P : i2MoOa. 
 If precipitation takes place at a temperature higher than 
 35-4O C., there is liability of ammonium arseno-molybdate 
 being precipitated in case there is any arsenic in the sample. 
 
 All of the* molybdenum before passing through the re- 
 ductor is in a form corresponding to the oxide Mo0 3 , and 
 is reduced in the reductor to a form corresponding to the 
 oxide Mo2O 3 . It is, however, so sensitive to the action of
 
 DETERMINATION OF PHOSPHORUS IN IRON AND STEEL 8 1 
 
 oxygen that on coming in contact with the air 1 in the flask 
 it takes up oxygen from the air, and is oxidized to a form 
 corresponding to the oxide MoaAtf. 2 The calculation of 
 results must, therefore, be based upon this reduction. 
 
 13. Having standardized the potassium permanganate 
 solution against iron, the conversion of the iron value to the 
 phosphorus value is very simple, as is shown by the following 
 equations: 
 
 2KMnO 4 o 50 =0= loFe. 
 MoaA; + 35O = 24MoO 3 . 
 2P =0= 24MoO 3 =c= yoFe. 
 
 P * 35Fe. 
 
 In the ferric alum modification advantage is taken of the 
 fact that ferric iron will oxidize reduced molybdenum solu- 
 tions. The molybdenum in the form of MoaOa as it leaves 
 the reductor reduces a corresponding amount of iron to the 
 ferrous condition, and is in turn oxidized to MoOs. As the 
 ferrous iron is more stable than reduced molybdenum solu- 
 tions, the titration is on the whole more satisfactory. 
 
 14. As the ammonium phospho-molybdate has a strong 
 tendency to crawl, it must be filtered through a paper which 
 fits the funnel perfectly. Otherwise it creeps over the top 
 of the paper and into the filtrate. It is necessary to wash 
 the edges of the paper thoroughly, so as to remove the last 
 traces of iron and molybdenum. If the iron is not removed 
 it will color the drop of yellow ammonium sulphide black 
 while it is still alkaline. 'The presence of molybdenum is 
 shown by its coloring the precipitated sulphur brown. When 
 free from molybdenum the sulphur will be perfectly white. 
 
 15. The determination of phosphorus in most of the alloy 
 steels involves no rpecial difficulties and the regular procedure 
 is applicable except in the presence of large amounts of tung- 
 sten, vanadium, or chromium when special methods must be 
 used. In steels containing chromium an insoluble residue, 
 probably chromium carbide, is left on treating with nitric acid. 
 This goes into solution slowly on digestion at a low tempera- 
 ture. It is advisable to keep the solution at constant volume 
 by the occasional addition of some water. 
 
 1 W. A. Noyes, /. Am. Chem. Soc., 16, 553. 
 
 2 Blair and Whitfield, /. Am. Chem. Soc., 17, 747. 
 
 3 Johnson, /. Ind. and Eng. Chem., n, 113.
 
 82 QUANTITATIVE ANALYSIS 
 
 1 6. The best method for the gravimetric determination of 
 phosphorus is the one which makes use of the basic acetate 
 separation of the phosphoric acid along with some of the iron 
 and the subsequent separation, after the addition of citric acid 
 to hold the iron hi solution, as magnesium ammonium phos- 
 phate. 
 
 ALKALIMETRIC METHOD FOR PHOSPHORUS 
 
 HANDY METHOD 
 
 Proceed exactly as in the preceding methods to the point 
 of washing the ammonium phospho-molybdate precipitate. 
 Instead of using acid ammonium sulphate solution, wash with 
 i per cent nitric acid l to remove iron salts, and follow with 
 i per cent potassium nitrate until litmus paper shows freedom 
 from acid. Transfer the precipitate with the filter paper to 
 the flask in which precipitation was made, cover with enough 
 tenth-normal sodium hydroxide solution to completely dis- 
 solve the yellow precipitate, using a slight excess, and then 
 titrate with tenth-normal nitric acid using phenolphthalein 
 as indicator. 
 
 Notes. i. The reaction upon which this method is based 
 is the following: 
 
 2(NH 4 ) 3 PO 4 .i2MoO 3 + 46NaOH = 
 
 2(NH 4 ) 2 HPO 4 + (NH4) 2 Mo0 4 + 2 3 Na 2 MoO 4 + 22H 2 O. 
 
 It is thus seen that the ratio of phosphorus to sodium 
 hydroxide is, P =c= 23NaOH, and upon this basis the calcula- 
 tions are made. 
 
 2. This method is extensively used, is more rapid than the 
 methods involving reduction of the molybdic acid, and is 
 accurate for the average steel. 
 
 3. Standardization of the sodium hydroxide solution is 
 made preferably against the yellow precipitate obtained from 
 a steel in which the phosphorus has been accurately deter- 
 mined by some other method. It is needless to say that the 
 sodium hydroxide solution must be protected from carbon 
 dioxide of the air and must be prepared free from carbonates 
 
 1 Add 13 cc. nitric acid sp. gr. 1.42 to i liter of water.
 
 PETERMINATION OF SULPHUR IN STEEL 83 
 
 by the introduction, at the time of solution, of a small volume 
 of barium hydroxide solution. 
 
 DETERMINATION OF SULPHUR IN STEEL 
 
 EVOLUTION METHOD 
 
 Standardization of Iodine and Sodium Thiosulphate Solu- 
 tion. Prepare a solution of sodium thiosulphate by dissolving 
 five grams of the crystallized salt, Na 2 S2O 3 .5H 2 O, in one liter 
 of water. Prepare a solution of iodine by dissolving about 
 2.5 grams of iodine in 100 cc. of water containing 5 grams of 
 pure potassium iodide, and diluting, after solution is complete, 
 to one liter. Standardize these two solutions against each 
 other, using freshly-prepared starch solution as indicator, and 
 then standardize the sodium thiosulphate against potassium 
 permanganate in the following manner. Draw 10 cc. of 
 standardized potassium permanganate solution 1 into a No. 5 
 beaker containing one gram of potassium iodide dissolved in 
 350 cc. of cold distilled water. Add dilute sulphuric acid by 
 drops, until the solution has just cleared from precipitated 
 manganese dioxide. The liberated iodine is then titrated with 
 sodium thiosulphate, freshly-prepared starch paste being 
 added when the yellow color of the iodine solution has nearly 
 disappeared. The sharpness of the end reaction is much 
 increased if the solution be cooled by a few lumps of ice. 
 Knowing the value of the potassium permanganate solution, 
 the strength of the sodium thiosulphate solution can be 
 calculated. 
 
 METHOD 
 
 Construct an evolution apparatus as follows: Use a 250 cc. 
 round bottom flask as a generating flask. Fit into this a 
 doubly perforated rubber stopper, through one hole of which 
 is passed a dropping funnel extending to near the bottom of 
 the flask, and through the other a delivery tube bent twice at 
 right angles. This tube leads down about two inches below the 
 
 1 The potassium permanganate may be standardized as described on p. 21 
 by using sodium oxalate as standard.
 
 84 QUANTITATIVE ANALYSIS 
 
 rubber stopper into a 250 cc. Erlenmeyer flask. From this 
 flask have a short tube bent at right angles and connected by 
 rubber tubing with a second tube bent at right angles and 
 extending to the bottom of a second Erlenmeyer flask. By a 
 similar arrangement connect a third Erlenmeyer flask. 
 
 Place 10 cc. of a cadmium chloride solution and 50 cc. of 
 water in the second and third absorption flasks, and 50 cc. of 
 water in the first flask. (Cadmium chloride solution contains 
 120 grams CdCl 2 in 1500 cc. HsO, and 600 cc. NELiOH, sp. gr. 
 0.90.) 
 
 Weigh into the generating flask 5 grams of steel, and place 
 50 cc. of hydrochloric acid (sp. gr. 1.2)* in the dropping funnel. 
 
 When the apparatus is shown by test to be tight, allow the 
 acid to enter the flask in small portions, maintaining an even 
 and fairly rapid flow of gas through the apparatus. When 
 action ceases or slackens, warm gently, and finally boil the 
 liquid until steam condenses in the first flask. Open the 
 stopcock of the dropping funnel, and at the same time remove 
 the flame. 
 
 Precipitation usually takes place only in the second flask. 
 Filter the liquid containing the cadmium sulphide precipi- 
 tate. Wash out the flask twice with water containing a 
 small amount of ammonia, then transfer the filter and pre- 
 cipitate to a beaker and pour over it 300 cc. of water. By 
 means of vigorous stirring, break up the filter paper into 
 shreds. Rinse out the guard flask and the absorption flask 
 and tubes with 50 cc. of water to which 5 cc. hydrochloric 
 acid (sp. gr. 1.20) has been added. This should remove all 
 sulphide. Add this liquid to the liquid containing the filter 
 paper and precipitate, and at once add an excess of iodine 
 solution, titrating back with the thiosulphate solution after a 
 few moments. Take great care to avoid loss of hydrogen 
 sulphide before the addition of the iodine. 
 
 Notes. i. Sulphur occurs in irons and steels in four forms: 
 (i) iron sulphide; (2) manganese sulphide; (3) titanium sul- 
 phide; and (4) a form, or forms, not yet identified, but char- 
 acterized by certain reactions. Manganese sulphide is the 
 most commonly occurring form, and the sulphur occurs as iron 
 1 Pulsifer, /. Ind. Eng. Chem., 10, 545.
 
 DETERMINATION OF SULPHUR IN STEEL 85 
 
 sulphide only when there is insufficient manganese present to 
 combine with it. The occurrence of titanium sulphide is rare, 
 although it is occasionally found in irons made from ores 
 carrying small amounts of titanium. The unidentified form 
 does not respond to the reactions of the other three forms and 
 there is some evidence, although not by any means conclusive, 
 that there may be a combination with carbon, or silicon, or 
 both. 
 
 2. The effect of sulphur, when existing as iron sulphide, is 
 to make the metal "red-short," i.e., brittle at the forging 
 temperatures. This effect is overcome by the addition of 
 manganese, which converts iron sulphide into manganese 
 sulphide. The red-shortness is destroyed, but in some cases 
 the sulphide of manganese is rolled out into filaments, and 
 these, being brittle and weakly cemented to the steel surround- 
 ing them, become points of weakness. 
 
 The most important effect of sulphur is its influence upon 
 the forms of carbon. 1 In castings where form is more of a 
 factor than strength, high sulphur irons may be used so as to 
 get a very fluid metal. In these castings the carbon is always 
 in the combined form, and the iron is white or gray. 
 
 3. The evolution method for sulphur hi steel is based upon 
 the assumption that all of the sulphur will be evolved as 
 hydrogen sulphide, which may be determined by titration 
 with iodine. The reactions involved are the following: 
 
 1. MnS + 2 HC1 = MnCl 2 + H 2 S. 
 
 2. H 2 S + 2l = 2HI + S. 
 
 The standardization of the iodine and sodium thiosulphate 
 solutions is based upon these reactions: 
 
 1. 2KMnO 4 + toKI + 8H 2 SO 4 = 6K 2 S0 4 + 2MnSO 4 + 
 
 8H 2 + 5 I 2 . 
 
 2. 2l + 2Na 2 S 2 O 3 = 2NaI 
 
 This method of standardization is most accurate and con- 
 venient if one has ready a standard potassium permanganate 
 solution. Other methods of standardization, such as the 
 titration of iodine against pure arsenious acid, the titration of 
 sodium thiosulphate against pure iodine, or the use of a steel of 
 
 1 See note, p. 62.
 
 86 QUANTITATIVE ANALYSIS 
 
 known sulphur content, may be used, but the time involved 
 is usually greater, and no greater accuracy is secured. 
 
 4. Instead of titrating hydrogen sulphide against iodine, 
 the hydrogen sulphide may be passed directly into some 
 oxidizing agent, such as ammoniacal hydrogen peroxide, alka- 
 line potassium permanganate, or hydrochloric acid saturated 
 with bromine. The sulphuric acid formed may then be pre- 
 cipitated and weighed as barium sulphate. Indirect oxidation 
 may also be used by absorbing in an alkaline lead salt, or 
 ammoniacal zinc or cadmium salts and subsequently oxidizing 
 the sulphides formed. 
 
 5. The assumption that all of the sulphur of an iron or steel 
 is evolved as hydrogen sulphide is not strictly correct, in fact 
 some, steels and irons yield none of their sulphur as hydrogen 
 sulphide. Part of it may be evolved as a sulphur ether 
 (CHs^S, or some analogous compound, and part of it may 
 remain as a residue unacted upon by hydrochloric acid. 1 
 
 Phillips 2 was the first to point out that certain irons gave 
 low results, due to loss of sulphur hi combination with carbon 
 and hydrogen. He showed that oily drops containing sulphur 
 were formed and that sulphur in this form was not oxidized 
 by the usual oxidizing, absorbing agents, nor did this form 
 react with lead, zinc or cadmium salts to form sulphides. To 
 overcome this difficulty, he proposed passing the evolved 
 gases through a heated tube before passing into the absorbing 
 agent. In this way the organic sulphide was broken down in 
 the presence of hydrogen to hydrogen sulphide. 
 
 Various workers have recommended annealing the metal 
 before placing in the evolution flask, and have obtained 
 improved results. Elliot 3 recommends annealing in a closed 
 muffle a five-gram sample mixed with 0.25 gram dry potas- 
 sium ferrocyanide and wrapped in one n cm. filter paper. 
 The best temperature for annealing is between 750 and 850 C. 
 and the time twenty minutes. Steels so treated give satis- 
 factory results by the evolution process. 
 
 Because of the difficulties mentioned the evolution process 
 should never be used where an absolute value is wanted. In 
 
 1 Blair, J. Am. Chem. Soc., 19, 114. 
 
 2 Phillips, /. Am. Chem. Soc., 17, 891. 
 
 3 J. Iron and Steel Inst., ign, No. i, 4.12.
 
 DETERMINATION OF SULPHUR IN CAST IRON 87 
 
 a majority of cases it will yield fairly accurate results, and by 
 annealing the sample, improved results are obtained, but 
 where there is any question of controversy the Bamber method 
 should be employed. 
 
 6. Iodine and sodium thiosulphate solutions are subject 
 to change, and their values should be verified frequently. 
 The titration of iodine by sodium thiosulphate in weak acid 
 solution gives accurate results, but the acid must be weak and 
 the solution cool. 
 
 7. Wiborgh 1 has devised a colorimetric process for sulphur 
 in steel, based upon the depth of color attained by a cloth 
 treated with a cadmium salt. The evolved gases are passed 
 through the cloth placed over the mouth of the generating 
 vessel, and this cloth is compared either with a standard, or a 
 steel of known sulphur content. 
 
 DETERMINATION OF SULPHUR IN CAST IRON 
 
 BAMBER METHOD 2 
 
 Weigh five grams of borings into a 250 cc. porcelain casserole, 
 and treat with 40 cc. of concentrated nitric acid. The reaction 
 may be violent, and it is wise to have ready a beaker of cold 
 water into which to dip the casserole in order to slow down the 
 reaction. Evaporate to about one-half the volume and then 
 add 5 grams of sodium carbonate. Transfer to a platinum 
 dish, and evaporate to dryness, stirring the pasty mass from 
 time to time. Ignite over an alcohol flame until the oxides of 
 nitrogen are completely driven off. After cooling, treat with 
 20 cc. of a five per cent solution of sodium carbonate and 
 bring to boiling. Allow to settle, and decant through a filter 
 into a No. 4 beaker. Repeat this process several times. 
 Acidulate the filtrate with hydrochloric acid, evaporate to 
 dryness, re-dissolve in water with a few drops of hydro- 
 chloric acid, filter if necessary, and precipitate in the boiling 
 solution with barium chloride. Allow to stand until the 
 precipitate has settled, filter, wash, and ignite as in the analysis 
 of pyrite. 
 
 1 Stahl und Eisen, 6, 240. 
 
 * J. Iron and Steel Inst., 1894, No. i, 319.
 
 88 QUANTITATIVE ANALYSIS 
 
 Notes. i. Some steels and irons will not yield all of their 
 sulphur in the form of sulphuric acid when treated with con- 
 centrated nitric acid or with aqua regia, nor as hydrogen 
 sulphide on treatment with hydrochloric acid. Furthermore, 
 the precipitation of barium sulphate in the presence of iron 
 salts is always an uncertain and inaccurate process. In order 
 to overcome these difficulties and to obtain all of the sulphur, 
 a fusion process must be used. The direct fusion of the steel 
 is a very slow and uncertain process. Bamber proposed to 
 overcome this by a combination solution and fusion process. 
 In this method partial oxidation is accomplished by nitric 
 acid, and complete oxidation by fusion with sodium nitrate 
 and by the decomposition of ferric nitrate into oxide of iron 
 and the strongly oxidizing oxides of nitrogen. Bamber's 
 method has the distinct advantage over other methods in that 
 all of the sulphur, in whatever form present, is obtained, and, 
 further, that precipitation of the barium sulphate is made in 
 the absence of iron and under well-controlled conditions. 
 
 2. Concentrated nitric acid frequently renders steel or iron 
 passive, and solution takes place only with difficulty. In 
 order to hasten solution, it is advisable to add, from time to 
 time, one or two drops of concentrated hydrochloric acid. 
 The hydrochloric acid must never be present in excess, as loss 
 of sulphur as sulphur dioxide will result. 
 
 3. Evaporation to dryness of the ferric nitrate solution is 
 a tedious operation. To prevent spattering and caking of 
 the surface, the mass must be stirred frequently. 
 
 An alcohol flame is not absolutely necessary for the ignition, 
 but is desirable. By placing the dish in an opening in an 
 asbestos board, gas may be used for the ignition, but the board 
 must be large enough to deflect the products of combustion 
 of the gas. The platinum dish is always attacked to an 
 appreciable extent by the ignition, and the process has, there- 
 fore, the disadvantage of the wear on the platinum. 
 
 4. Sodium carbonate is used to wash out the sulphates, so 
 as to decompose any basic salts which may be present. 
 
 5. Pulsifer 1 has recommended the use of chloric acid as 
 solvent and oxidizing agent for the determination of sulphur 
 in steel, and claims excellent results for the method. He says 
 
 1 /. Ind. Eng. Ghent., 8, 1115.
 
 DETERMINATION OF SILICON IN CAST IRON 89 
 
 that the Bamber method is liable to give slightly high results 
 on account of the tendency to absorb sulphur from outside 
 sources, and the nitric acid oxidation method gives low results 
 because of the solubility of barium sulphate in ferric chloride 
 solution. 
 
 An excellent bibliography of the determination of sulphur 
 in iron and steel is given in this paper. 
 
 DETERMINATION OF SILICON IN CAST IRON 
 
 METHOD OF DROWN 
 
 For the determination of silicon in cast iron, weigh a one- 
 gram sample; for wrought iron and steel, weigh a three-gram 
 sample. Treat the borings in a casserole with a mixture of 
 15 cc. of water and 15 cc. of nitric acid (sp. gr. 1.42). As soon 
 as the reaction has ceased, add 25 cc. of sulphuric acid, made by 
 mixing equal parts of the concentrated acid and water. 
 Evaporate to dryness on the steam table and then heat over 
 a free flame in a well ventilated hood until the sulphuric acid 
 fumes strongly. Allow to cool and then add 150 cc. of water 
 and 5 cc. of dilute hydrochloric acid. Heat until the solution 
 of the ferric sulphate is complete and filter by suction at once. 
 Wash with dilute hydrochloric acid (one part acid sp. gr. 1.12 
 to three parts of water) until the filter paper is free from 
 iron stain, and with hot water to remove the hydrochloric 
 acid. Place the filter and silica in a platinum crucible, char 
 the paper slowly and ignite with the crucible on its side and 
 the lid partially covering the mouth. 
 
 The graphite from cast iron burns off very slowly and con- 
 sequently the crucible should be so placed as to have as much 
 air as possible. When the ignition is complete, cool in a 
 desiccator, weigh, and treat the contents of the crucible with 
 hydrofluoric and sulphuric acids exactly as in silicate analysis. 
 Weigh again. The difference in weight is silica. 
 
 Notes. i. Silicon exists in cast irons and steels for the 
 most part in the form of a solid solution of the compound 
 in iron. In wrought iron and some low carbon steels,
 
 90 QUANTITATIVE ANALYSIS 
 
 part of it may also exist in the oxidized form as silicate of iron 
 or silicate of manganese. A silicate of manganese having the 
 composition 2MnO.3SiC>2 has been isolated from steel by 
 Stead. 1 
 
 In ferro-silicon, the silicon may exist as a solid solution of 
 the compound Fe 2 Si hi iron; as the compounds Fe 2 Si or FeSi, 
 or mixtures of these two; and as mixtures of FeSi with silicon, 
 the relative proportions of each being dependent upon the 
 percentage of silicon present. 
 
 2. The influence of silicon hi small amounts upon steel has 
 not been well established, except hi its effect upon the shrink- 
 age hi cast steels. In such steels, the excessive shrinkage can 
 be counteracted by the introduction of silicon, the amount 
 remaining hi the metal being about 0.30 per cent. It also 
 diminishes blow holes hi castings. 
 
 In cast irons, the silicon influences the form of carbon very 
 markedly. With increasing percentages of silicon, there is a 
 corresponding increase hi the amount of carbon in the form of 
 graphite. While cast irons are graded upon the appearance 
 of graphite hi the fractured surface, it is the silicon which 
 influences the character of this appearance. The appearance 
 of the fracture is also undoubtedly influenced by temperature 
 and time of cooling. 2 
 
 3. Nitric acid is used hi preference to hydrochloric or sul- 
 phuric acids for solution because it will decompose any titanium 
 carbide which might be present hi titaniferous pig irons. In 
 the other acids it is insoluble and would be left in the siliceous 
 residue. 
 
 4. Evaporation to free fuming of the sulphuric acid is 
 necessary, otherwise there is incomplete dehydration of the 
 silica. More complete dehydration takes place hi sulphuric 
 acid than in hydrochloric acid solutions. 
 
 Nitric acid, when being freely removed, will fume hi moist 
 air and this must not be mistaken for sulphuric acid. 
 
 During the evaporation, the ferric sulphate is liable to 
 spatter. It is, therefore, necessary to have a watch glass of 
 sufficient size to cover the casserole and to rotate the casserole 
 during the final heating. 
 
 1 Stead, Iron and Steel Magazine, 9, 105. 
 Adamson, /. I. & S. Inst. 1911, No. 2, 86.
 
 DETERMINATION OF COPPER IN STEEL 91 
 
 5. The filtration of the silica must take place soon after 
 the solution of the ferric sulphate is complete on account of 
 the tendency of the silica to go back into solution. Losses 
 amounting to fully half of the total silica may occur on allow- 
 ing the solution to stand over night. 
 
 6. In cast irons, the burning of the graphitic residue is a 
 tedious operation and cannot be helped much by the use of the 
 blast lamp. It is advisable to start this ignition, and, in the 
 meantime, continue other work until the silica becomes white. 
 
 7. In some ferro-silicons, solution in acids takes place 
 extremely slowly. In high silicon products such as ferro- 
 silicon, "Duriron," etc., it is, therefore, more advantageous 
 to fuse at once with sodium carbonate or with a mixture of 
 sodium carbonate and sodium peroxide and to treat the fusion 
 as in the analysis of feldspar. 
 
 DETERMINATION OF COPPER IN STEEL 
 
 Dissolve 5 grams of steel in 75 cc. of nitric acid (sp. gr. 1.20) 
 by gently heating in a covered 250 cc. porcelain casserole. 
 When solution is complete, add 25 cc. of dilute sulphuric acid 
 (i : i) and evaporate until the acid fumes freely. Dissolve 
 by heating with 50 cc. of water and 10 cc. of dilute sulphuric 
 acid, filter into a 500 cc. Erlenmeyer flask and wash the silica 
 with hot water. Dilute the filtrate to 300 cc., add 25 cc. of 
 concentrated freshly prepared ammonium bisulphite solution, 
 and heat to boiling. When the iron is completely reduced, 
 add 30 cc. of a 20 per cent solution of sodium thiosulphate, 
 and continue the boiling until the sulphide of copper is coagu- 
 lated. Filter, wash with hot water, and ignite in a porcelain 
 crucible. Dissolve in a small amount of concentrated nitric 
 acid, nearly neutralize with ammonia, transfer to a platinum 
 crucible and electrolyze, making the crucible the cathode, 
 Wash the crucible with water after all of the copper is de- 
 posited, then with alcohol, dry and weigh. 
 
 Notes. i. Copper may be found in small amounts in 
 nearly all steels, and is added to sheet steel in amounts, usually 
 under 0.25 per cent, in order to retard corrosion. Steels thus
 
 92 QUANTITATIVE ANALYSIS 
 
 treated resist ordinary corrosive action of the weather much 
 better than ordinary steels. 
 
 The effect of copper upon the physical properties has not 
 been definitely determined. In high carbon steels there is 
 a tendency for the copper not to alloy with the iron. 
 
 In high-speed tool steels copper is detrimental. 
 
 2. The principle involved in this method is dependent upon 
 the precipitation of the copper as sulphide in a solution in 
 which the iron has been reduced, its conversion into nitrate 
 and the deposition of metallic copper by electrolysis. 
 
 3. Instead of depositing the copper as metal, the sulphide 
 may be ignited and weighed as oxide, but it is always con- 
 taminated with iron and hence the results are less accurate. 
 
 The solut'.on obtained in the evolution method for sulphur, 
 or the filtrate from the silicon determination may be used 
 for the precipitation of copper. 
 
 4. The purpose of the ammonium bisulphite is to reduce the 
 iron to the ferrous condition. Sodium thiosulphate could be 
 used, but it would involve the separation of a large amount of 
 free sulphur. 
 
 5. The reaction between copper salts and sodium thiosul- 
 phate is complicated. The principal reaction is the reduction 
 of the cupric ion to cuprous ion and at the same time oxi- 
 dation of the S-A" ion to S 4 O 6 ", SO 3 " and SO 4 " ions. There 
 are also formed complex cupro- thiosulphate ions (CuSaOs)', 
 which decompose in boiling solution with formation of Cu 2 S. 
 
 6. The volumetric method may be used for the determina- 
 tion of copper. In case it is desired to do this, proceed exactly 
 as in the Low method after having dissolved the copper sul- 
 phide in nitric acid. 
 
 DETERMINATION OF NICKEL IN STEEL 
 
 METHOD OF BRUNCK 1 
 
 Weigh 0.5-0.6 gram of steel into a No. 4 beaker, and dis- 
 solve in 20 cc. of hydrochloric acid, sp. gr. 1.12. To the 
 solution add about i cc. concentrated nitric acid and boil to 
 oxidize all of the iron. The solution is filtered if silica has 
 separated, or in the case of a high silicon steel, it should be 
 1 Z. angew. Chem. (1907), 1844.
 
 DETERMINATION OF NICKEL IN STEEL 93 
 
 evaporated to dryness to remove silica. To the solution is 
 added 2 to 3 grams of tartaric acid, and it is then diluted with 
 water to about 300 cc. Add ammonia in slight excess and if 
 no ferric hydroxide separates, make slightly acid with hydro- 
 chloric acid, and heat to near boiling. Add 20 cc. of a i per 
 cent alcoholic solution of dimethylglyoxime, and ammonia, 
 drop by drop, till present in slight excess. Allow to stand 
 for one hour, collect the deep-red precipitate on a Gooch 
 crucible, wash with hot water, dry at 110 for 45 minutes, 
 and weigh as NKC^^CfeV 
 
 Notes. Nickel forms with iron a series of homogeneous 
 solid solutions for all concentrations. Its addition to steel 
 raises the strength and particularly the elastic limit without 
 corresponding loss in ductility. (See Carbon.) By the use 
 of nickel steel it is possible to get a greater strength with smaller 
 cross section than with a carbon steel. For this reason it is 
 used extensively in manufacturing parts where the total weight 
 is an important consideration, such as in bicycle and automo- 
 bile parts. 
 
 The most useful nickel steels are those which contain 2-3.75 
 per cent nickel, and 0.1-0.65 P er cen t carbon. For marine 
 boiler tubes, condensers, etc., low carbon, 25-30 per cent 
 nickel steels are used, which have the advantage of being more 
 non-corrosive than carbon steels, and relatively very much 
 lighter in weight. 
 
 Nickel affects the carbon condition to a very marked extent. 
 Ordinary nickel steels containing 3.5 per cent of nickel, and 
 carbon up to 0.80 per cent are pearlitic, but with increasing 
 percentages of carbon are either martensitic or austenitic. 
 With lower carbon these same changes are produced with 
 increase of nickel. For example, a 0.2 per cent carbon steel 
 is pearlitic up to 8 per cent nickel; martensitic with 12-20 
 per cent nickel; and austenitic above 25 per cent nickel. 
 
 2. The principle involved in the separation of nickel from 
 iron is based upon the insolubility of nickel dimethylglyoxime 1 
 in slightly ammoniacal solution, the ferric oxide being held in 
 solution in the form of a complex ion in combination with tar- 
 
 1 Tschugaeff, Berichte, 38, 2520; Brunck, Stahl und Risen, 28, 331; Brunck, 
 Z. angew. Chem. (1907), 1844; Prettner, Chem. Ztg., 33, 396.
 
 94 QUANTITATIVE ANALYSIS 
 
 taric acid. The method separates nickel from cobalt, iron, 
 chromium, manganese, copper, tungsten, and vanadium. If 
 much manganese is present, the separation should be made in 
 acetic acid solution. 
 
 3. Another principle frequently used for the separation of 
 iron from nickel is based upon the solubility in ether of ferric 
 chloride 1 in hydrochloric acid solution, sp. gr. i.n. By this 
 method most of the ferric chloride can be removed not only 
 from nickel, but from cobalt, manganese, chromium, aluminum 
 and titanium, thus giving a simple means of removing most of 
 the iron from these elements. This principle has also been 
 used to advantage for the removal of iron in sulphur deter- 
 minations, the ferric chloride being soluble, and the sulphate 
 insoluble. The removal of the small amount of iron left in 
 solution after the ether separation is a comparatively simple 
 operation. The separation from nickel may be made by means 
 of ammonia and ammonium chloride, or by the basic acetate 
 process. 
 
 4. In some methods for the determination of nickel, it is 
 necessary to remove copper first. This is so when iron is 
 removed by means of ammonia and ammonium chloride, or by 
 ether. The nickel is subsequently precipitated as sulphide, 
 ignited and weighed as oxide. When nickel is determined 
 volumetrically by means of potassium cyanide, it is also 
 necessary to remove copper first. 
 
 5. Nickel may be determined volumetrically without the 
 removal of copper by making use of the reaction between 
 potassium cyanide and the nickel ammonia complex, 
 Ni(NH 3 ) 6 ++ , as expressed by the equation: 
 
 (NO 3 ) 2 + 4KCN = K 2 [Ni(CN) 4 ] + 6NHa + 2KNO 3 . 
 
 The nickel glyoxime is separated by filtration on an asbestos 
 filter, formed on a perforated porcelain plate in an ordinary 
 funnel. The precipitate is washed with water and then 
 transferred in the funnel to a dean eight-ounce suction flask. 
 Dissolve the precipitate on the filter with 10 to 20 cc. of hot 
 concentrated nitric acid, added drop by drop, and wash four 
 or five tunes with water, the process being conducted with 
 suction. The nickel solution is then poured into the beaker 
 
 1 Rothe, Mittheilungen Konig. Tech. Versuchs. zu Berlin, 1892.
 
 DETERMINATION OF CHROMIUM IN STEEL 95 
 
 in which the precipitation was made, several crystals of 
 ammonium persulphate added and the solution is boiled for 
 3 to 5 minutes. The solution should be perfectly clear at 
 this point, but if any turbidity persists, it should be filtered. 
 Ice may be added to cool the solution, which is then made 
 slightly, but distinctly ammoniacal and 10 cc. each of silver 
 nitrate and potassium iodide solutions are added as indicator, 
 and titration is conducted with a standardized cyanide solution. 
 
 Silver nitrate 0.5 gram AgNO 3 in 1000 cc. water. 
 
 Potassium iodide 20 grams KI in 1000 cc. water. 
 
 Cyanide approximately 4.57 grams of KCN in 1000 cc, 
 water. 
 
 DETERMINATION OF CHROMIUM IN STEEL 
 
 METHOD OF BARBA 1 
 
 Weigh 1.5-2 grams of steel into a 300 cc. Erlenmeyer flask, 
 and dissolve in 20 cc. sulphuric acid (1:5). Heat to boiling, 
 and add drop by drop concentrated nitric acid until all the iron 
 is completely oxidized. If a residue remains it should be 
 filtered off, ignited and fused with a small amount of sodium 
 carbonate and potassium nitrate. The fusion should then be 
 taken up in a small amount of dilute sulphuric acid and added 
 to the main solution. Boil off the oxides of nitrogen, dilute 
 with boiling water to 150 cc. and add 5 cc. saturated potassium 
 permanganate solution or more if necessary to produce a 
 permanent pink color. Boil 15 to 20 minutes, and if after this 
 time the color of the permanganate persists, add 25 cc. strong 
 ammonia. Digest at a low heat with occasional stirring for 
 half an hour or until the excess permanganate is destroyed. 
 Then add cautiously 20 cc. dilute sulphuric acid (i : i), and 
 heat for i to 2 minutes to boiling. Cool the solution to room 
 temperature, transfer to a 250 cc. graduated flask and dilute 
 to the mark. Mix thoroughly by pouring back and forth into 
 a dry beaker. Allow the precipitate to settle, filter through 
 a 4-inch Swedish filter, rejecting the first few cc. ; measure off 
 200 cc. of the clear filtrate, add an excess of standard ferrous 
 1 Iron Age, 52, 153.
 
 96 QUANTITATIVE ANALYSIS 
 
 sulphate, and determine the excess by means of standard 
 potassium permanganate. 
 
 Notes. i. Chromium as a constituent of steel has the 
 effect of increasing the strength and the hardness, and chrome 
 steels are always finer grained than plain carbon steels. They 
 are always used in the heat treated condition and the properties 
 are susceptible of great modifications by heat treatment. 
 Where extreme hardness is desirable, chrome steels are used 
 and for dies, rolls for cold rolling metals, balls, and roller bear- 
 ings give excellent service. The Aci point is slightly raised, 
 and, therefore, chrome steels must be hardened at a higher 
 temperature than plain carbon steels. 
 
 Chrome-nickel steels have many desirable properties in the 
 heat-treated condition and are used extensively, particularly 
 in automobile parts. For such uses the chromium may vary 
 from 0.50 per cent to 1.50 per cent and the nickel from 1.25 
 per cent to 3.50 per cent. 
 
 These steels are also used for armor plate and projectiles. 
 
 Mayari steel is made from an ore mined in Mayari, Cuba, 
 which contains chromium and nickel. This steel is used for 
 rails and automobile parts. 
 
 High Speed tool steels contain chromium and tungsten with 
 about 2.50 to 5.00 per cent of chromium and about 18.00 per 
 cent of tungsten. They may be free from vanadium, but 
 most of those on the market now carry small amounts (0.50- 
 2.00 per cent) of this element. 
 
 2. The principles involved in this method are based on the 
 oxidation of chromium to chromic acid in dilute sulphuric 
 acid solution by means of concentrated potassium perman- 
 ganate, the reduction of the chromic acid by means of ferrous 
 sulphate, and the estimation of the excess ferrous sulphate by 
 means of standard potassium permanganate solution. The 
 reactions involved in the calculation of results are as follows: 
 
 i. H 2 Cr 2 7 + 6FeS0 4 + 6H 2 SO 4 = 3 Fe 2 (SO 4 ) 3 + Cr 2 (SO 4 ) 3 
 
 2. 2KMn0 4 + ioFeSO 4 + 8H 2 S0 4 = 5Fe 2 (SO 4 ) 3 + K 2 S0 4 
 2 MnS0 4
 
 DETERMINATION OF CHROMIUM IN STEEL 97 
 
 3. Chromium and iron may be separated by means of ether 
 in the same manner as nickel and iron. The solution contain- 
 ing the chromium ions may be oxidized to chromate ion, and 
 then determined either by gravimetric or volumetric means. 
 
 4. The iron is oxidized with nitric acid in order to avoid 
 the use of a larger amount of potassium permanganate and the 
 subsequent formation of large quantities of manganese salts 
 and of manganese dioxide. 
 
 5. The value of the ferrous sulphate solution in terms of 
 potassium permanganate should be determined each day before 
 using. 
 
 6. Sodium bismuthate may be used to oxidize chromium 
 from the trivalent to the hexavalent condition if added to the 
 hot solution. The chromic acid formed is not decomposed 
 in the boiling solution, and it may be separated by nitration 
 from excess sodium bismuthate and from precipitated manga- 
 nese dioxide. 
 
 DETERMINATION OF CHROMIUM IN CHROME- 
 VANADIUM STEEL 
 
 METHOD OF CAIN 1 
 
 Weigh 1-2 grams of steel into a 300-00. Erlenmeyer flask, 
 and dissolve in concentrated hydrochloric acid, using 10 cc. of 
 acid per gram. When solution is complete dilute to 150 cc. 
 with hot water, and add saturated sodium carbonate solution 
 until nearly neutralized. Add barium carbonate emulsion in 
 slight excess, boil vigorously 10-15 minutes, with small addi- 
 tions of emulsion every two or three minutes. Not more than 
 1-2 grams excess of barium carbonate should be added, and 
 care should be exercised to keep the flask covered as much as 
 possible while adding it. 
 
 Remove the flask from the heat, allow the precipitate to 
 settle and filter on an 1 1 cm. filter paper, washing twice with 
 hot water. Ignite the precipitate in a platinum crucible, 
 mix with two grams of sodium carbonate and 0.25 gram potas- 
 sium nitrate. Fuse for 20 minutes. 
 
 Extract the melt by placing the crucible on the bottom of a 
 small beaker and covering with hot water. Filter into an 
 1 J.Ind. Eng. Ckem., 4, 17.
 
 98 QUANTITATIVE ANALYSIS 
 
 Erlenmeyer flask, add i or 2 cc. of hydrogen peroxide and boil 
 for five minutes to destroy the excess. Add a slight excess of 
 nitric acid, shake vigorously to remove carbon dioxide. Trans- 
 fer to a 250 cc. beaker, and add 2 cc. dilute (i : i) nitric 
 acid for each 100 cc. of solution. Add 20 cc. of a 2o-per cent 
 solution of lead nitrate with constant stirring. Filter on 
 asbestos, wash several times with cold water. Transfer the 
 asbestos mat to a beaker and decompose the lead chromate 
 with hot dilute (i : 4) hydrochloric acid. Cool the solution, 
 dilute to 200 cc. and titrate with standard ferrous sulphate 
 solution, using potassium ferricyanide as outside indicator. 
 
 Notes. i. The principle of this method is based upon the 
 precipitation in boiling solution of chromium hydroxide, in 
 presence of ferrous iron, and the subsequent oxidation of 
 chromium to chromate, the latter being measured by a stand- 
 ard ferrous sulphate solution. 
 
 2. Noyes and Bray 1 have studied the conditions for the 
 separation of chromium and vanadium, and find that by boil- 
 ing with barium carbonate all of the chromium is precipitated, 
 and the separation from vanadium is complete. To further 
 insure separation Cain separates chromium as lead chromate 
 under conditions which would prevent vanadium being 
 precipitated. 
 
 3. The addition of hydrogen peroxide is for the removal of 
 any nitrite which if present might, on acidifying, reduce 
 chromium or vanadium. The excess hydrogen peroxide must 
 be removed as it would reduce chromic acid. 
 
 4. The removal of carbon dioxide is simply for the purpose 
 of getting a precipitate which will settle rapidly. 
 
 DETERMINATION OF TUNGSTEN IN STEEL 
 
 Weigh 2-5 grams of sample into a porcelain casserole and 
 dissolve in 60 cc. nitric acid (sp. gr. 1.20) with the addition, 
 from time to time, of a few drops of strong hydrochloric acid. 
 If the precipitated tungstic acid is dark colored, digest for 
 some time with occasional addition of a few drops of hydro- 
 chloric acid until the precipitate becomes a light yellow. 
 1 Tech. Quart. (1908), 21, 14.
 
 DETERMINATION OF TUNGSTEN IN STEEL 99 
 
 Evaporate the solution to dryness and heat to decompose 
 nitrates. Moisten with concentrated hydrochloric acid and 
 evaporate to dryness. Take up in concentrated hydro- 
 chloric acid and warm until all ferric oxide is in solution. 
 Evaporate off the excess of hydrochloric acid, and continue 
 until the solution becomes syrupy. Dilute with 50-75 cc. of 
 dilute hydrochloric acid, stir, allow to settle and filter. Wash 
 the precipitate on to the filter with dilute hydrochloric acid 
 (i: 5), adding some paper pulp if there is any tendency to 
 pass through the filter paper; and complete the washing with 
 hydrochloric acid. Remove any tungs tic oxide adhering to 
 the casserole with a small piece of filter paper moistened with 
 ammonia, add to the precipitate, ignite over a Tirrill burner, 
 at first at a low temperature, and finally at full heat. 
 
 Treat with 2-3 drops of sulphuric acid and about 2-3 cc. of 
 hydrofluoric acid, evaporate to remove silica, and ignite. 
 
 Fuse with six times its weight of sodium carbonate, allow 
 to cool, digest with hot water, and filter, washing the residue 
 with hot water. Nearly neutralize the filtrate with nitric 
 acid, boil off carbonic acid, allow to cool, and make slightly 
 acid with nitric acid. Add mercurous nitrate 1 solution to 
 the gently boiling solution, allow to settle and test with more 
 mercurous nitrate. Then add an emulsion of mercuric oxide 
 and water until excess nitric acid is neutralized. Allow the 
 precipitate to settle, filter, wash several times by decantation, 
 adding a few drops of mercurous nitrate solution to the hot 
 water, and finally wash with water. 
 
 Dry and ignite the precipitate in a porcelain crucible and 
 weigh as WOs. 
 
 Notes. i. Steels containing about 6 per cent of tung- 
 sten and about 0.60 per cent of carbon are extensively used in 
 the hardened condition for permanent magnets. When used 
 for electric meters they are seasoned, after hardening, by heat- 
 ing for a long time at 100. Tools for making finishing cuts at 
 
 1 ioo grams of salt dissolved in one liter of water, to which some free 
 mercury has been added.
 
 100 QUANTITATIVE ANALYSIS 
 
 relatively high speed are made of steels .containing i per cent 
 of carbon and 3-4 per cent of tungsten. According to Arnold 
 and Read 1 carbon is combined with iron in low tungsten steels, 
 but with increasing tungsten more of the carbon enters into 
 combination with it and hi steels containing 11.5 per cent or 
 more all of the carbon is in combination with tungsten. 
 
 High speed steels contain 13-18 per cent of tungsten, 
 2.5-5.00 per cent of chromium, 0.60-0.70 per cent of carbon, 
 and usually under i.oo per cent of vanadium. They are 
 characterized by their ability to be worked at a speed 3-5 
 times greater than with plain carbon steels, and to maintain 
 their hardness at a red heat ("red-hardness"). These steels 
 are hardened at a temperature, white heat, which would ruin 
 a plain carbon steel. It is believed that at the high tempera- 
 ture which is necessary for satisfactory hardening, double 
 carbides 2 of chromium and tungsten are formed, which persist 
 on rapid cooling, and also to some extent on slow cooling. 
 
 2. The principle of this method is dependent upon the pre- 
 cipitation of tungstic acid anhydride, WOg, from an oxidizing 
 acid solution. This precipitate is subsequently converted 
 into mercurous tungstate, ignited and weighed as WO 3 . 3 
 
 3. If the residue obtained after removal of silica is weighed, 
 it will give a fairly accurate result for tungsten. This residue, 
 however, invariably contains traces of iron, chromium, and 
 manganese, and it is the purpose of the fusion with sodium 
 carbonate to remove tungsten as soluble sodium tungstate 
 from the elements. 
 
 4. In ferro-tungsten and tungsten metal, insoluble in acid, 
 decomposition may be accomplished by fusing the powdered 
 metal with a mixture of sodium carbonate and potassium 
 nitrate, or with Eschka 4 mixture. In either case the product 
 is soluble sodium tungstate which may be precipitated as 
 mercurous tungstate and ignited to tungstic acid. 
 
 1 Proc. Inst. Mech. Eng. (1914), 223. 
 
 1 Edwards, /. I. and S. Inst., 1908, 2, 104. 
 
 1 WOs begins to sublime 500 below its m.p. at 1473 C.A., 12, 226. 
 
 See page 48.
 
 DETERMINATION OF VANADIUM IN STEEL IOI 
 
 DETERMINATION OF VANADIUM IN STEEL 
 
 Dissolve a 2-gram sample in 20 cc. of hydrochloric acid 
 (sp. gr. i. 20) adding sufficient nitric acid to completely oxidize 
 the iron. Evaporate to small volume, and in case tungstic 
 acid separates, dilute, filter, wash with dilute hydrochloric 
 acid, and evaporate the filtrate to a syrupy consistency. 
 Pour the solution into a separatory funnel, and wash the 
 casserole free from ferric chloride by successive small addi- 
 tions of hydrochloric acid (sp. gr. i.io), taking care to keep 
 the volume small. Add an equal volume of ether, and shake 
 for several minutes, releasing the pressure occasionally by 
 opening the stopcock of the funnel, holding the tube of the 
 funnel pointing upward and allowing time for any ejected 
 solution to return into the bulb. Draw off the hydrochloric 
 acid solution from the ether layer into a second separatory 
 funnel and repeat the shaking with ether. 
 
 Again separate the hydrochloric acid layer from the ether, 
 and evaporate to small volume. Convert the hydrochloric 
 acid solution into a nitric acid solution by several evaporations 
 to small volume after the addition of some concentrated 
 nitric acid between evaporations. Dilute slightly with water, 
 add two or three drops of sulphurous acid, heat to boiling and 
 pour slowly into 100 cc. of boiling sodium hydroxide solution 
 (100 grams per liter). Boil a few minutes, filter, and wash 
 with hot water. Acidify the filtrate slightly with nitric acid, 
 make slightly alkaline with sodium hydroxide and boil. 
 Filter, and to the filtrate add 10 cc. of a solution of lead ace- 
 tate, and then enough acetic acid to give a decided odor. 
 Heat to boiling and filter off the lead vanadate on asbestos, 
 washing with hot water. Dissolve by pouring through the 
 precipitate hot dilute hydrochloric acid. Evaporate nearly 
 to dryness, add 50 cc. concentrated hydrochloric acid and 
 again evaporate. Add 10 cc. concentrated sulphuric acid 
 and evaporate until the acid fumes freely. Cool, dilute to
 
 102 QUANTITATIVE ANALYSIS 
 
 150 cc. and titrate with dilute potassium permanganate, 
 keeping the solution at a temperature of 60-70. 
 
 Notes. i. The so-called vanadium steels nearly always 
 contain chromium and should in reality be called chrome- 
 vanadium steels. Many plain carbon steels have been treated 
 with vanadium, and some of them contain the element in small 
 amounts, the treated metal having desirable properties. The 
 amount of vanadium contained in them is so small that there 
 is no more reason why they should be called vanadium steels 
 than the ordinary steel should be called manganese steel on 
 account of its having been treated with ferro-manganese. 
 These steels probably have greater fatigue resisting qualities 
 than plain carbon steels. 
 
 Chrome-vanadium steels are extensively used in the manu- 
 facture of automobiles and are used in those parts where it is 
 desirable to have high elastic limit and at the same time high 
 ductility. Such steels contain approximately i.oo per cent 
 of chromium and 0.20 per cent of vanadium, the carbon 
 content varying, depending upon the use. 
 
 Vanadium is frequently used in high-speed tool steels in 
 amounts varying from 0.40 to 2.50 per cent, but is usually 
 under i.oo per cent. The steels containing vanadium are 
 supposed to have superior properties. 
 
 2. Advantage is taken of the solubility of ferric chloride 
 in ether to separate most of the iron from vanadium. The 
 remainder of the iron, chromium, manganese, nickel, and 
 copper are separated by means of sodium hydroxide. The 
 vanadium is then separated as lead vanadate, subsequently 
 being reduced by hydrochloric acid to vanadyl salt and then 
 oxidized in sulphuric acid solution with potassium perman- 
 ganate to vanadic acid. 
 
 The essential reactions involved are the following: 
 
 Pb 3 (VO 4 ) 2 + 6HC1 = 3PbCl 2 + 2H 3 VO 4 , 
 2 H3V0 4 + 6HC1 = 2 VOC1 2 + 6H 2 + C1 2 , 
 VOC1 2 + H 2 S0 4 = VOSO 4 + 2HC1 
 ioVOSO 4 + 2KMnO 4 + 8H 2 S0 4 = 
 
 SV 2 02(S0 4 ) 3 + K 2 S0 4 + 2 MnS0 4 + 8H 2 O.
 
 ANALYSIS OF DURALUMIN 103 
 
 3. The Rothe 1 separation of ferric chloride by means of 
 ether is an extremely useful process and may be applied to the 
 separation of iron from chromium, aluminum, titanium, nickel, 
 and manganese. When these elements are present in relatively 
 small amounts, it serves as a convenient process for concen- 
 trating them from the larger amount of iron. 
 
 Molybdenum chloride 2 is also soluble in ether and in a 
 separation will be found with the ferric chloride, from which it 
 may be separated as sulphide by precipitating under pressure 
 with hydrogen sulphide. It may then be converted into sul- 
 phate, reduced as in the method for phosphorus, and titrated 
 with potassium permanganate. 
 
 4.' It is necessary to convert the hydrochloric acid solution 
 into nitric acid solution on account of using lead acetate as a 
 precipitant. The removal of one volatile acid by another is not 
 a question of the strength of the acid removed, but one of vol- 
 atility and the amount of acid which is present in excess. 
 
 Precautions must be used in evaporating nitric acid solu- 
 tions containing large amounts of iron as hydrolyzed products 
 separate which are soluble only in hydrochloric acid. 
 
 5. The few drops of sulphur dioxide are to reduce any 
 chromium which may have been oxidized to chromate ions, 
 and the second treatment with sodium hydroxide removes all 
 of the chromium and nickel. 
 
 6. Some differences of opinion has been expressed as to the 
 reduction of vanadium by means of hydrochloric acid. Cam- 
 pagne 3 has obtained satisfactory results, and others have 
 verified his results. Apparently the conditions must always 
 be the same, and in order to avoid variations some workers 
 prefer reduction with sulphurous acid. 
 
 ANALYSIS OF DURALUMIN 4 
 
 DETERMINATION OF SILICON AND IRON 
 
 Dissolve one gram of well mixed drillings in 45 cc. of acid 
 made by mixing 5 cc. nitric acid (sp. gr. 1.42), 10 cc. hydro- 
 chloric acid (sp. gr. 1.20), 5 cc. sulphuric acid (sp. gr. 1.84), 
 
 1 Rothe, Mittheilungen Konig. Tech. Versuchs. zu Berlin, 1892. 
 
 J Blair, J. Am. Chem. Soc., 30, 1228. 
 
 1 Comptes Rendus (1903), 137, 570. 
 
 * Standard Methods of Analysis (modified), Aluminum Co. of America.
 
 104 QUANTITATIVE ANALYSIS 
 
 with 25 cc. of water, using a casserole provided with a cover 
 glass. When the drillings are completely dissolved, evaporate 
 the solution to dryness and raise the heat till the sulphuric 
 acid fumes freely. Continue until a few drops have fallen 
 from the center of the glass. This insures the freedom of the 
 solution from hydrochloric acid and nitric acid, and the com- 
 plete dehydration of the silica. Take up the residue with 5 cc. 
 of concentrated sulphuric acid and about 100 cc. of water; 
 boil until the sulphate is completely dissolved, and filter. 
 Wash the residue at least seven times with hot water to 
 remove sulphates, and ignite in a weighed platinum crucible. 
 Weigh the ignited crucible. The difference between -the two 
 weights represents the silicon as silica. Calculate the silica to 
 silicon. 
 
 DETERMINATION OF IRON 
 
 Pass the hot nitrate and wash water from the silicon deter- 
 mination through a Jones reductor, following the procedure 
 given on page 75 and titrate with potassium permanganate. 
 The Jones reductor should have been prepared for service by 
 washing with definite amounts of hot, dilute sulphuric acid 
 and water/ and the solution in which the iron is to be deter- 
 mined should be preceded and followed by hot dilute sul- 
 phuric acid as directed under the Determination of Phosphorus 
 in Iron and Steel. 
 
 DETERMINATION OF COPPER AND MAGNESIUM 
 
 Place one gram of the alloy together with 5 grams of solid 
 sodium hydroxide in a 500 cc. porcelain casserole and add just 
 enough water to cover; when violent action has ceased, dilute 
 to 200-250 cc. with hot water and boil until solution is com- 
 plete. Filter while hot and wash the residue until free from 
 alkali. The aluminum and most of the zinc dissolve, while 
 the other metals remain in the metallic state. 
 
 Dissolve the residue in hydrochloric acid (i : i) using a little 
 nitric acid if necessary. Neutralize with ammonia, make
 
 ANALYSIS OF DURALUMIN 105 
 
 weakly acid with hydrochloric acid, heat, and precipitate 
 copper with hydrogen sulphide. Filter and wash. Ignite 
 the precipitate in a porcelain crucible, cool, add 2 to 3 cc. of 
 concentrated nitric acid and heat gently until all the copper 
 is in solution. Transfer the solution to a 250 cc. wide mouth 
 Erlenmeyer flask containing 0.5 cc. of concentrated sulphuric 
 acid, and evaporate to complete dryness. Cool, add 25 cc. of 
 12 per cent acetic acid and heat until solution of sulphates is 
 completed. Add 5 cc. of saturated Na2HPO4 solution. When 
 cold add 3 grams of potassium iodide, and allow to stand for 
 ten minutes. 
 
 Titrate the free iodine with sodium thio-sulphate. A few 
 drops of fresh starch solution are added near the end of the 
 titration. 
 
 To the nitrate from the hydrogen sulphide precipitation 
 add 5 cc. concentrated hydrochloric acid, then a small excess 
 of ammonia and precipitate with hydrogen sulphide. Filter 
 and wash with water containing a little ammonium sulphide. 
 
 Make the nitrate faintly acid with hydrochloric acid and 
 boil off hydrogen sulphide, adding a little ammonium per- 
 sulphate toward the latter part of the boiling to oxidize 
 sulphur. 
 
 Add 30 cc. saturated solution of microcosmic salt and cool. 
 Add ammonia drop by drop, stirring vigorously to make the 
 precipitate crystalline. Then add an excess of strong am- 
 monia. Let stand .for two hours in ice water. Filter, wash 
 with cool water containing 5 per cent of strong ammonia and 
 5 per cent ammonium nitrate. Ignite until completely white. 
 Weigh as Mg 2 P 2 O 7 . 
 
 DETERMINATION OF MANGANESE 
 
 Weigh out accurately about 0.2 gram of the sample in which 
 the manganese is to be determined. Place sample in a 150 cc. 
 beaker and add 50 cc. of water, 10 cc. of nitric acid (sp. gr. 
 1.42), and 10 cc. sulphuric acid (sp. gr. 1.84). Heat until 
 solution is complete, and all nitrous fumes driven off. Re-
 
 106 QUANTITATIVE ANALYSIS 
 
 move from source of heat, cool, add 50 cc. water, 10 cc. silver 
 nitrate solution containing 0.003 gram per cc. and one gram 
 ammonium persulphate and heat gently until color is fully 
 developed. Cool to room temperature and titrate with stand- 
 ard sodium arsenite solution. 
 
 i. Duralumin 1 is an alloy of aluminum having approxi- 
 mately the composition 
 
 Per cent 
 
 Copper 3.75 
 
 Magnesium o. 50 
 
 Manganese o. 60-0. 70 
 
 Aluminum Balance 
 
 and with small amounts of silicon and iron which the alum- 
 inum contained as impuritites. It can readily be forged and 
 is always used in this condition. It also responds to heat 
 treatment, and the properties can be materially improved by 
 this means. If, after forging, the alloy be reheated to 400- 
 500 C. and quenched in water its physical properties, hard- 
 ness, tensile strength, and elongation, will be about the same 
 as when forged, viz.; tensile strength 35,000 pounds per 
 square inch, and 10 per cent elongation in two inches. Almost 
 immediately it begins to harden and strengthen so that after 
 about four days the strength has increased to 55,000 pounds 
 per square inch and the elongation to 15-20 per cent. This 
 change in properties which is known as "aging" is much 
 accelerated by heating to ioo-2oo C. The change at the 
 higher temperature is rapid, so that after a certain critical 
 time corresponding to maximum physical properties a pro- 
 longation of the time produces an annealing effect. 
 
 The material which has been subjected to the aging effect 
 may have its strength further increased by cold work. 
 
 The effect of aging duralumin may be explained by taking 
 into account the lowering of the solubility of the compound, 
 CuAl2, from 500 to room temperature. At the higher 
 temperature CuAl 2 is held in solid solution and this state is 
 maintained at the ordinary temperature by quenching, but 
 it will separate out if slowly cooled. In the quenched alloy 
 
 1 Wilm, Metallurgie 8, 225; Circular, Bureau of Standards, No. 76.
 
 ANALYSIS OF DURALUMIN 107 
 
 separation of CuAl 2 takes place slowly at ordinary temperature 
 more rapidly on heating at ioo-2oo, and maximum hardness 
 corresponds closely to a certain average size of particle sepa- 
 rated. 
 
 Duralumin in the hardened condition is remarkably resis- 
 tant to corrosion which is very important in view of its ex- 
 tended use in air-plane construction. 
 
 2. The method of analysis is applicable not alone to dur- 
 alumin but to other aluminum alloys as well. 
 
 The solubility of aluminum and some of its alloys hi simple 
 acids is so slow that special mixtures of acids are used to accel- 
 erate the action. 
 
 3. Aluminum may contain some or all of its silicon in the 
 graphitic form which is not attacked by acids nor oxidized 
 by heat. If present, the siliceous residue must be treated 
 with hydrofluoric and sulphuric acids, heated, and the brown 
 insoluble residue weighed as graphitic silicon. 
 
 4. The determination of manganese in aluminum alloys is 
 based upon its oxidation in the presence of silver nitrate to 
 permanganic acid by means of ammonium persulphate, the 
 destruction of the excess of the latter by heat, and the titration 
 of the permanganic acid by means of sodium arsenite. The 
 calculations are based upon the following reaction: 
 
 2HMnO 4 +5Na 3 AsO 3 +4HN03 =
 
 io8 
 
 QUANTITATIVE ANALYSIS 
 LOGARITHMS OF NUMBERS 
 
 J 
 
 -, 
 
 
 
 
 
 
 
 
 
 
 
 PROPORTIONAL PARTS 
 
 If 
 
 
 
 i 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 
 
 
 
 
 
 
 
 
 i 1 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 oooc 
 
 0043 
 
 0086 
 
 0128 
 
 0170 
 
 0212 
 
 0253 
 
 0294 
 
 0334 
 
 :37- 
 
 4 
 
 8 
 
 12 
 
 17 
 
 21 
 
 25 
 
 9 
 
 33 
 
 37 
 
 II 
 
 0414 
 
 0453 
 
 0492 
 
 0531 
 
 0569 
 
 0607 
 
 0645 
 
 0682 
 
 9719 
 
 0755 
 
 4 
 
 8 
 
 II 
 
 15 
 
 H 
 
 23 
 
 4 
 
 30 
 
 34 
 
 a 
 
 0792 
 
 0828 
 
 0864 
 
 0899 
 
 0934 
 
 0969 
 
 1004 
 
 1038 
 
 1072 
 
 I ICt 
 
 3 
 
 7 
 
 10 
 
 14 
 
 J " 
 
 21 
 
 24 
 
 28 
 
 31 
 
 13 
 
 14 
 
 1139 
 
 1461 
 
 1173 
 1492 
 
 1206 
 1523 
 
 1239 
 1553 
 
 1271 
 
 1584 
 
 I3O3 
 l6l4 
 
 1335 
 1644 
 
 1367 
 
 1673 
 
 I.V.I' 
 I7C3 
 
 U3C 
 1732 
 
 3 
 3 
 
 6 
 6 
 
 9 
 
 12 
 
 S 
 
 18 
 
 21 
 
 Jf> 
 24 
 
 ?9 
 27 
 
 15 
 
 1761 
 
 1790 
 
 1818 
 
 1847 
 
 1875 
 
 1903 
 
 193 
 
 1959 
 
 11,8: 
 
 JCI4 
 
 3 
 
 t 
 
 8 
 
 II 
 
 14 
 
 17 
 
 20 
 
 22 
 
 25 
 
 1 6 
 
 2041 
 
 2068 
 
 2095 
 
 2122 
 
 2148 
 
 2175 
 
 220 
 
 2227 
 
 053 
 
 2270 
 
 3 
 
 5 
 
 8 
 
 II 
 
 13 
 
 if, 
 
 18 
 
 21 
 
 24 
 
 17 
 
 2304 
 
 2330 
 
 2355 
 
 2380 
 
 2405 
 
 2430 
 
 245 
 
 2480 
 
 25:4 
 
 252y 
 
 a 
 
 5 
 
 7 
 
 1C 
 
 13 
 
 IS 
 
 17 
 
 2: 
 
 22 
 
 It 
 
 2553 
 
 2577 
 
 2601 
 
 2625 
 
 2648 
 
 2672 
 
 269 
 
 2718 
 
 2742 
 
 27'->5 
 
 2 
 
 5 
 
 7 
 
 9 
 
 12 
 
 14 
 
 16 
 
 19 
 
 21 
 
 19 
 
 2788 
 
 2810 
 
 2833 
 
 2856 
 
 2878 
 
 2900 
 
 2923 
 
 2945 
 
 29<'7 
 
 1999 
 
 i 
 
 4 
 
 7. 
 
 9 
 
 11 
 
 13 
 
 16 
 
 18 
 
 20 
 
 to 
 
 30ic 
 
 3032 
 
 3054 
 jtf,t 
 
 3075 
 
 00 Q. 
 
 3096 
 
 3118 
 
 3139 
 
 3160 
 
 pft 
 
 3901 
 
 2 
 
 4 
 
 a 
 
 I 
 
 11 
 
 13 
 
 IS 
 
 17 
 
 i a 
 
 21 
 
 3222 
 
 3243 
 
 3203 
 
 32O4 
 
 3304 
 
 3324 
 
 3345 
 
 3365 
 
 S S S S 
 
 ^4"4 
 
 
 
 
 
 
 
 
 
 iS 
 
 93 
 
 3424 
 3617 
 
 3444 
 3636 
 
 ?K?n 
 
 3464 
 3655 
 
 iQlQ 
 
 3483 
 3674 
 2ft rfi 
 
 3502 
 3692 
 
 -28-7 
 
 3522 
 37U 
 
 ?8n? 
 
 354 1 
 
 3729 
 
 3560 
 3747 
 
 357i J 
 
 >7' r ' 
 
 i5i>S 
 
 n4 
 
 k . f " -> 
 
 2 
 
 4 
 
 6 
 
 7 
 
 9 
 
 
 
 13 
 
 IS 
 
 17 
 17 
 
 25 
 
 3979 
 
 3o2O 
 
 3997 
 
 33o 
 4014 
 
 35 
 
 4031 
 
 3o/4 
 
 4048 
 
 39 2 
 4065 
 
 399 
 
 4082 
 
 3927 
 
 4099 
 
 30. :> 
 
 nfl 
 
 ~>.' - 
 
 1133 
 
 i 
 
 4 
 3 
 
 5 
 5 
 
 7 
 
 9 
 
 it 
 
 12 
 
 14 
 
 IS 
 
 36 
 
 4150 
 
 4166 
 
 4183 
 
 4200 
 
 4216 
 
 4232 
 
 4249 
 
 4265 
 
 pfe 
 
 P98 
 
 i 
 
 3 
 
 5 
 
 7 
 
 8 
 
 10 
 
 II 
 
 13 
 
 IS 
 
 37 
 
 4314 
 
 433 
 
 4346 
 
 4362 
 
 4378 
 
 4393 
 
 4409 
 
 4425 
 
 444: 
 
 MSt 
 
 i 
 
 3 
 
 5 
 
 6 
 
 a 
 
 I 
 
 II 
 
 13 
 
 14 
 
 rt 
 
 4472 
 
 4487 
 
 4502 
 
 4518 
 
 4533 
 
 4548 
 
 4564 
 
 4579 
 
 4SSM 
 
 1009 
 
 i 
 
 3 
 
 5 
 
 6 
 
 8 
 
 I 
 
 II 
 
 12 
 
 14 
 
 19 
 
 4624 
 
 4639 
 
 4654 
 
 4669 
 
 4683 
 
 4698 
 
 4713 
 
 4728 
 
 474; 
 
 4757 
 
 i 
 
 3 
 
 4 
 
 a 
 
 7 
 
 '. 
 
 10 
 
 12 
 
 IJ 
 
 9> 
 
 4771 
 
 4786 
 
 4800 
 
 4814 
 
 4829 
 
 4843 
 
 4857 
 
 4871 
 
 pu 
 
 4900 
 
 i 
 
 3 
 
 4 
 
 i 
 
 - 
 
 I 
 
 10 
 
 H 
 
 13 
 
 a 
 
 49M 
 
 4928 
 
 4942 
 
 4955 
 
 4969 
 
 4983 
 
 4997 
 
 5011 
 
 9004 
 
 ntf 
 
 i 
 
 3 
 
 4 
 
 a 
 
 7 
 
 S 
 
 10 
 
 II 
 
 12- 
 
 3-> 
 
 3051 
 
 5065 
 
 5079 
 
 5092 
 
 5105 
 
 5H5 
 
 5132 
 
 5145 
 
 5159 
 
 ^172 
 
 i 
 
 .3 
 
 4 
 
 s 
 
 7 
 
 - 
 
 I 
 
 II 
 
 I 
 
 33 
 
 5185 
 
 5198 
 
 -fyft 
 
 5211 
 
 5224 
 
 5237 
 --jrfVl 
 
 5250 
 
 5263 
 
 5276 
 
 g*g 
 
 5303 
 
 - i< 
 
 i 
 
 .3 
 
 4 
 
 5 
 
 ' 
 
 a 
 
 9 
 
 1C 
 
 ia 
 
 34 
 3S 
 
 5315 
 5441 
 
 o32o 
 
 5453 
 
 534 
 5465 
 
 5353 
 5478 
 
 ;>3w 
 5490 
 
 537S 
 5502 
 
 539 ! 
 5514 
 
 54O3 
 5527 
 
 ^41'' 
 
 5539 
 
 ^ --' s 
 5551 
 
 i 
 
 2 
 
 4 
 
 S 
 
 I" 
 
 7 
 
 9 
 9 
 
 I ; 
 
 IX 
 
 |B 
 
 5563 
 
 5575 
 
 5587 
 
 5599 
 
 56II 
 
 5623 
 
 5635 
 
 5647 
 
 5658 
 
 5670 
 
 i 
 
 2 
 
 4 
 
 g 
 
 h 
 
 7 
 
 e 
 
 K 
 
 II 
 
 37 
 
 5682 
 
 5694 
 
 5705 
 
 5717 
 
 5729 
 
 5740 
 
 5752 
 
 5763 
 
 srrs 
 
 n 
 
 i 
 
 2 
 
 3 
 
 5 
 
 
 
 7 
 
 i 
 
 9 
 
 ; : 
 
 38 
 
 5798 
 
 5809 
 
 5821 
 
 5832 
 
 5843 
 
 5855 
 
 5866 
 
 5877 
 
 SHI 
 
 5899 
 
 i 
 
 2 
 
 .i 
 
 s 
 
 f. 
 
 7 
 
 B 
 
 9 
 
 ; 3 
 
 39 
 
 59" 
 
 6O2I 
 
 5922 
 
 firi-t-r 
 
 5933 
 
 ri40 
 
 5944 
 A- i 
 
 5955 
 
 6O64 
 
 5966 
 
 /vvr 
 
 5977 
 6085 
 
 5988 
 6006 
 
 099 
 
 iOK 
 
 ! 
 
 2 
 
 .i 
 
 4 
 
 5 
 
 7 
 
 8 
 
 9 
 
 ID 
 
 4 - 
 41 
 42 
 
 6128 
 6232 
 
 OO3I 
 6138 
 6243 
 
 OO42 
 6149 
 6253 
 
 OO;>3 
 6lfc 
 6263 
 
 6170 
 6274 
 
 t>0/D 
 6l8o 
 6284 
 
 6191 
 6294 
 
 6201 
 6304 
 
 '212 
 OM 
 
 'jJJ 
 
 bas 
 
 
 I 
 
 2 
 
 a 
 
 3 
 
 3 
 
 4 
 
 ' 
 - 
 
 '- 
 '' 
 
 7 
 7 
 
 8 
 
 a 
 
 
 
 
 
 43 
 
 6335 
 
 6345 
 
 6355 
 
 6365 
 
 6375 
 
 6385 
 
 6395 
 
 6405 
 
 1415 
 
 Mas 
 
 i 
 
 2 
 
 3 
 
 4 
 
 5 
 
 t 
 
 7 
 
 8 
 
 9 
 
 44 
 
 6435 
 
 (\e-yy 
 
 ft-A') 
 
 6454 
 
 6--T 
 
 6464 
 
 n-fir 
 
 6474 
 
 6484 
 
 6^8O 
 
 6493 
 fi-nn 
 
 6503 
 
 ficnn 
 
 Sft3 
 
 "r- 
 
 tsa 
 
 >"i I S 
 
 ! 
 
 ; 
 
 3 
 
 4 
 
 > 
 
 ' 
 
 7 
 
 8 
 
 a 
 
 
 
 43 
 
 47 
 48* 
 
 0532 
 
 6628 
 6721 
 
 68l2 
 
 "042 
 
 6637 
 6730 
 
 6821 
 
 "551 
 
 6646 
 6739 
 68w 
 
 0501 
 
 6656 
 6749 
 6810 
 
 0571 
 
 6665 
 6758 
 6848 
 
 6675 
 6767 
 6857 
 
 059 
 
 6684 
 
 6776 
 6866 
 
 "599 
 
 6693 
 6785 
 
 fAfC 
 
 ''-:' 
 
 9joa 
 
 c.'-i 
 
 - v < 1 
 
 712 
 -:; 
 S&a 
 
 I 
 I 
 
 I 
 2 
 
 3 
 
 a 
 
 4 
 4 
 
 S 
 
 5 
 
 < 
 
 s 
 
 7 
 E 
 
 7 
 
 7 
 
 C> 
 8 
 
 S 
 
 V} 
 
 49 
 
 6902 
 
 6911 
 
 0030 
 6920 
 
 3y 
 6928 
 
 6937 
 
 6946 
 
 6955 
 
 OO75 
 
 6964 
 
 tn 
 
 iw 
 
 fHa 
 
 I 
 
 2 
 
 3 
 
 4 
 
 
 s 
 
 ( 
 
 7 
 
 I 
 
 SD 
 
 6990 
 
 6998 
 
 7007 
 
 7016 
 
 7024 
 
 7033 
 
 7042 
 
 7050 
 
 rag 
 
 yofi) 
 
 T 
 
 9 
 
 3 
 
 3 
 
 
 5 
 
 6 
 
 7 
 
 s 
 
 SI 
 
 7076 
 
 7084 
 
 7093 
 
 7101 
 
 7110 
 
 7118 
 
 7126 
 
 7135 
 
 7143 
 
 Tisa 
 
 I 
 
 2 
 
 3 
 
 a 
 
 
 5 
 
 6 
 
 7 
 
 ft 
 
 P 
 
 7160 
 
 7168 
 
 7177 
 
 7185 
 
 7193 
 
 7202 
 
 7210 
 
 7218 
 
 jot 
 
 7235 
 
 I 
 
 2 
 
 2 
 
 3 
 
 
 5 
 
 6 
 
 7 
 
 r 
 
 a 
 
 7243 
 
 7251 
 
 7259 
 
 7267 
 
 ~ 7i |Q 
 
 7275 
 
 Ticfi 
 
 728 4 
 
 ~?f\1 
 
 7292 
 
 7300 
 
 ITiRrt 
 
 13* 
 
 - -ss 
 
 7316 
 
 I 
 
 2 
 
 2 
 
 3 
 
 
 5 
 
 6 
 
 6 
 
 T 
 
 54 
 
 7324 
 
 7332 
 
 734 
 
 734** 
 
 735 
 
 734 
 
 7372 
 
 73"O 
 
 - ."-^ 
 
 7 ;./" 
 
 
 
 
 
 
 
 

 
 LOGARITHMS OF NUMBERS IOQ 
 
 LOGARITHMS OF NUMBERS (Continued} 
 
 i 
 
 ]i 
 
 ce c 
 
 
 55 
 56 
 57 
 58 
 59 
 
 60 
 61 
 62 
 63 
 64 
 
 6s 
 66 
 67 
 68 
 69 
 
 70 
 7i 
 
 72 
 
 73 
 
 74 
 
 75 
 76 
 77 
 78 
 79 
 
 80 
 81 
 82 
 83 
 
 84 
 
 85 
 86 
 87 
 88 
 89 
 
 90 
 91 
 92 
 93 
 94 
 
 95 
 96 
 97 
 98 
 99 
 
 
 
 I 
 
 2 
 
 3 
 
 4 
 
 5 
 
 7443 
 7520 
 7597 
 7672 
 7745 
 
 7818 
 7889 
 7959 
 8028 
 8096 
 
 8162 
 8228 
 8293 
 8357 
 8420 
 
 8482 
 8543 
 8603 
 8663 
 8722 
 
 8779 
 8837 
 8893 
 8949 
 9004 
 
 9058 
 
 6 
 
 7 
 
 8 
 
 9 
 
 PROPORTIONAL PARTS. 
 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 7404 
 7482 
 -559 
 7634 
 7709 
 
 7782 
 7853 
 7924 
 7993 
 8062 
 
 8129 
 8i95 
 8261 
 8325 
 8388 
 
 8451 
 8513 
 
 7412 
 7490 
 7566 
 7642 
 7716 
 
 7789 
 7860 
 7931 
 8000 
 8069 
 
 8136 
 8202 
 8267 
 8331 
 8395 
 
 8457 
 8519 
 8579 
 8639 
 8698 
 
 8756 
 8814 
 8871 
 8927 
 8982 
 
 9036 
 9090 
 9143 
 9196 
 9248 
 
 9299 
 9350 
 9400 
 9450 
 9499 
 
 954 
 959 
 964 
 968 
 973 
 
 978 
 982 
 987 
 991 
 996 
 
 7419 
 7497 
 7574 
 7649 
 
 7723 
 
 7796 
 7868 
 7938 
 8007 
 8075 
 
 8142 
 8209 
 8274 
 8338 
 8401 
 
 8463 
 8525 
 8585 
 8645 
 8704 
 
 8762 
 8820 
 8876 
 8932 
 8987 
 
 9042 
 9096 
 9149 
 9201 
 9253 
 
 9304 
 9355 
 9405 
 9455 
 9504 
 
 9552 
 9600 
 964 
 969 
 974 
 
 978 
 983 
 987 
 992 
 996 
 
 7427 
 7505 
 7582 
 7657 
 7731 
 
 7875 
 7945 
 8014 
 8082 
 
 8149 
 8215 
 8280 
 8344 
 8407 
 
 8470 
 8531 
 8591 
 8651 
 8710 
 
 8768 
 8825 
 8882 
 8938 
 8993 
 
 9047 
 
 7435 
 7513 
 7589 
 7664 
 7738 
 
 7810 
 7882 
 7952 
 8021 
 8089 
 
 8156 
 8222 
 8287 
 8351 
 8414 
 
 8476 
 8537 
 8597 
 8657 
 8716 
 
 8774 
 8831 
 8887 
 8943 
 8998 
 
 9053 
 
 7451 
 7528 
 7604 
 7679 
 7752 
 
 7825 
 7896 
 7966 
 8035 
 8102 
 
 8169 
 8235 
 8299 
 8363 
 8426 
 
 8488 
 8549 
 8609 
 8669 
 8727 
 
 8785 
 8842 
 8899 
 8954 
 9009 
 
 9063 
 
 7459 
 7536 
 7612 
 7686 
 7760 
 
 7832 
 793 
 7973 
 8041 
 8109 
 
 8176 
 
 466 
 
 (43 
 5iS 
 
 XM 
 7<>7 
 
 R38 
 
 yu 
 98t 
 
 34* 
 II 6 
 
 [8a 
 
 474 
 
 551 
 'ae> 
 701 
 
 774 
 
 M 
 
 !"7 
 987 
 
 C5.S 
 
 i 
 z 
 
 -> 
 
 i 
 i 
 i 
 t 
 
 2 
 
 <: 
 i 
 <, 
 I 
 
 \ 
 
 i 
 
 i 
 
 ^ 
 
 i 
 
 1 
 
 i 
 
 2 
 2 
 
 i 
 J 
 
 i 
 i 
 
 i 
 i 
 
 1 
 
 
 IS.. 
 
 z 
 
 , 
 
 
 
 
 
 
 
 8241 
 8306 
 8370 
 8432 
 
 8494 
 8555 
 8615 
 8675 
 8733 
 
 8791 
 8848 
 
 8904 
 8960 
 9015 
 
 9069 
 
 Jia 
 
 37* 
 
 439 
 
 S 
 
 561 
 6xi 
 
 r.si 
 238 
 
 797 
 
 8 
 
 <!K 
 
 96s 
 
 ox 
 
 of), 
 
 31'.i 
 3S2 
 
 445 
 
 5Cfi 
 5<>7 
 67 
 
 6K6 
 745 
 
 Boa 
 859 
 
 Sin 5 
 8971 
 
 ic 2('i 
 
 1079 
 
 i 
 i 
 i 
 
 i 
 i 
 
 ; 
 
 a 
 
 3 
 
 1 
 
 i 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 8573 
 8633 
 8692 
 
 8751 
 8808 
 8865 
 8921 
 8976 
 
 9031 
 9085 
 9138 
 9191 
 9243 
 
 9294 
 9345 
 9395 
 9445 
 9494 
 
 9542 
 9590 
 963 
 968 
 973 
 
 977 
 982 
 9868 
 991 
 995 
 
 i 
 i 
 
 i 
 i 
 
 2 
 t 
 
 2 
 
 t 
 t 
 
 - 
 
 1 
 
 ; 
 
 I 
 
 4 
 
 ; 
 i 
 
 i 
 i 
 
 t 
 
 5 
 
 s 
 
 i 
 
 
 
 
 
 
 1 
 
 i 
 
 
 
 
 
 
 1 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 9154 
 9206 
 9258 
 
 9309 
 936o 
 9410 
 9460 
 9509 
 
 955 
 960 
 965 
 969 
 974 
 
 979 
 983 
 988 
 992 
 996 
 
 9159 
 9212 
 9263 
 
 9315 
 9365 
 9415 
 946 
 951 
 
 956 
 9609 
 965 
 970 
 975 
 
 979 
 984 
 988 
 993 
 997 
 
 9165 
 
 9170 
 9222 
 9274 
 
 9325 
 9375 
 9425 
 947 
 952 
 
 957 
 961 
 9666 
 971 
 975 
 
 980 
 985 
 989 
 993 
 998 
 
 9175 
 9227 
 9279 
 
 9330 
 938o 
 9430 
 9479 
 9528 
 
 9576 
 9624 
 967 
 971 
 976 
 
 9809 
 985 
 989 
 994 
 998 
 
 ,is. 
 p& 
 
 B3 
 9385 
 943 
 
 MS 
 
 153 
 
 ,58 
 fe 
 
 <c 
 ITS 
 
 .7" 
 gl 
 
 5 
 0C 
 
 I'M 
 ggi 
 
 Il8 
 O3 
 
 US 
 
 134 
 09 
 94 ; 
 948 
 
 S3 
 
 158 
 163 
 
 
 
 173 
 
 .77 
 ft 
 
 06 
 19 
 195 
 
 UU< 
 
 i 
 i 
 
 
 
 
 r 
 
 e 
 
 r 
 i 
 
 t 
 
 i 
 i 
 i 
 
 i 
 i 
 i 
 
 i 
 
 
 
 
 
 
 
 9217 
 9269 
 
 9320 
 
 9370 
 
 9420 
 9469 
 9518 
 
 9566 
 961 
 966 
 9708 
 975 
 
 9800 
 984 
 989C 
 993 
 997 
 
 2 
 
 2 
 
 3 
 
 3 
 
 1 
 
 4 
 
 
 
 
 
 
 
 1 
 
 1 
 I 
 
 I 
 
 a 
 
 2 
 
 a 
 
 a 
 a 
 a 
 
 - 
 a 
 
 2 
 2 
 
 a 
 a 
 
 3 
 3 
 
 3 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 
 4 
 4 
 
 ! 
 1 
 1 
 
 2 
 
 2 
 2 
 
 i 
 ! 
 
 3 
 
 1 
 1 
 
 c 
 
 r 
 
 i 
 t 
 
 I 
 I 
 
 - 
 
 - 
 
 ! 
 
 3 
 
 ; 
 4
 
 no 
 
 QUANTITATIVE ANALYSIS 
 ANTILOGARITHMS 
 
 $ 
 
 .CD 
 
 .01 
 .09 
 .03 
 04 
 
 .05 
 .> 
 .07 
 .08 
 09 
 
 .10 
 
 .11 
 
 .12 
 
 13 
 .14 
 
 .I* 
 .17 
 .IS 
 IS 
 
 .30 
 .21 
 
 .22 
 23 
 .24 
 
 .25 
 .20 
 31 
 .98 
 
 .20 
 
 30 
 
 -31 
 -32 
 
 .33 
 .34 
 
 -35 
 
 .38 
 
 -37 
 .3 
 
 .39 
 
 4 
 43 
 41 
 
 43 
 
 44 
 
 .45 
 
 .4''' 
 .47 
 .48 
 
 49 
 
 
 IOOO 
 
 1023 
 1047 
 1072 
 
 1096 
 
 1 122 
 1148 
 H75 
 1202 
 1230 
 
 1259 
 1288 
 1318 
 1349 
 1380 
 
 1413 
 1445 
 1479 
 ISM 
 1549 
 
 1585 
 1622 
 1660 
 1698 
 1738 
 
 1778 
 1820 
 1862 
 1905 
 1950 
 
 I99S 
 2042 
 2089 
 2138 
 2188 
 
 2239 
 2291 
 2344 
 2399 
 2455 
 
 2512 
 
 2570 
 2630 
 2692 
 27S4 
 2818 
 2884 
 2951 
 3020 
 309 
 
 I 
 
 IOO2 
 1026 
 1050 
 1074 
 1099 
 1125 
 
 II5I 
 1178 
 1205 
 1233 
 1262 
 1291 
 1321 
 1352 
 1384 
 '1416 
 1449 
 1483 
 ISI7 
 1552 
 
 1589 
 1626 
 1663 
 1702 
 1742 
 
 1782 
 1824 
 1866 
 1910 
 1954 
 
 2OOO 
 2046 
 2094 
 
 2143 
 2193 
 2244 
 2296 
 2350 
 2404 
 2460 
 
 2518 
 2576 
 2636 
 2698 
 2761 
 
 2825 
 2891 
 2958 
 3027 
 3097 
 
 2 
 
 IOO5 
 1028 
 1052 
 1076 
 1102 
 
 1127 
 "53 
 1180 
 1208 
 1236 
 
 1265 
 1294 
 1324 
 1355 
 1387 
 
 1419 
 1452 
 1486 
 1521 
 1556 
 
 1592 
 1629 
 1667 
 1706 
 1746 
 
 1786 
 1828 
 1871 
 I9M 
 1959 
 2004 
 2051 
 2099 
 2148 
 2198 
 
 2249 
 
 2301 
 
 2355 
 2410 
 2466 
 
 2523 
 2582 
 2642 
 2704 
 2767 
 
 2831 
 2897 
 2965 
 3034 
 3105 
 
 3 
 
 1007 
 1030 
 1054 
 1079 
 1104 
 
 1130 
 1156 
 1183 
 
 I2II 
 
 1239 
 1268 
 1297 
 1327 
 1358 
 1390 
 1422 
 1455 
 1489 
 1524 
 1560 
 
 1596 
 1633 
 1671 
 1710 
 1750 
 
 1791 
 1832 
 1875 
 1919 
 1963 
 2009 
 2056 
 2104 
 2153 
 2203 
 
 2254 
 2307 
 2360 
 2415 
 2472 
 
 2529 
 2588 
 2649 
 2710 
 2773 
 2838 
 2904 
 2972 
 3041 
 3"2 
 
 4 
 
 s 
 
 6 
 
 1014 
 1038 
 1062 
 1086 
 
 7 
 
 8 
 
 1019 
 1042 
 1067 
 1091 
 
 9 
 
 1021 
 
 
 PROPORTIONAL PARTS. 
 
 I 
 O 
 
 2 
 
 3 
 
 i 
 
 4 
 
 i 
 
 5 
 
 6 
 
 i 
 
 7 
 
 2 
 
 8 
 
 2 
 
 2 
 
 2 
 2 
 
 1009 
 1033 
 1057 
 1081 
 
 1012 
 1035 
 
 1059 
 1084 
 
 1016 
 
 1040 
 1064 
 1089 
 
 
 
 i 
 
 IC<59 
 IOQ4 
 
 O 
 O 
 
 
 
 
 
 i 
 i 
 i 
 
 i 
 
 i 
 i 
 
 i 
 i 
 
 I 
 
 2 
 2 
 
 1132 
 "59 
 1186 
 1213 
 1242 
 
 1271 
 1300 
 1330 
 1361 
 1393 
 1426 
 1459 
 M93 
 1528 
 1563 
 1600 
 1637 
 1675 
 1714 
 1754 
 
 1795 
 1837 
 1879 
 1923 
 1968 
 
 2014 
 2061 
 2109 
 2158 
 2208 
 
 2259 
 2312 
 2366 
 2421 
 2477 
 
 2535 
 2594 
 2655 
 2716 
 2780 
 
 2844 
 2911 
 2979 
 3048 
 3"9 
 
 H35 
 1161 
 1189 
 1216 
 1245 
 
 1274 
 1303 
 1334 
 1365 
 1396 
 
 1429 
 1462 
 1496 
 IS3I 
 1567 
 
 1603 
 1641 
 1679 
 1718 
 1758 
 
 1799 
 1841 
 1884 
 1928 
 1972 
 
 2018 
 2065 
 2113 
 2163 
 2213 
 
 2265 
 2317 
 2371 
 2427 
 2483 
 
 2541 
 2600 
 2661 
 2723 
 2786 
 
 2851 
 2917 
 2985 
 3055 
 3126 
 
 1138 
 1164 
 1191 
 1219 
 1247 
 1276 
 1306 
 1337 
 1368 
 1400 
 
 1432 
 1466 
 1500 
 1535 
 1570 
 
 1607 
 1644 
 1683 
 1722 
 1762 
 
 1803 
 1845 
 1888 
 1932 
 1977 
 
 2023 
 2070 
 2118 
 2168 
 
 2218 
 2270 
 2323 
 2377 
 2432 
 2489 
 
 2547 
 2606 
 2667 
 2729 
 2793 
 2858 
 2924 
 2992 
 3062 
 3133 
 
 1140 
 1167 
 1194 
 
 1222 
 1250 
 
 1279 
 1309 
 1340 
 1371 
 1403 
 
 1435 
 1469 
 1503 
 1538 
 1574 
 1611 
 1648 
 1687 
 1726 
 1766 
 
 1807 
 1849 
 1892 
 1936 
 1982 
 
 2028 
 2075 
 2123 
 2173 
 
 2223 
 
 2275 
 
 2328 
 2382 
 2438 
 
 2495 
 
 2553 
 2612 
 2673 
 2735 
 2799 
 2864 
 2931 
 2999 
 3069 
 3i4i 
 
 1143 
 
 1169 
 1107 
 1225 
 1253 
 I2S2 
 1312 
 1343 
 1374 
 140(3 
 
 M39 
 1472 
 
 1507 
 1542 
 1578 
 
 1614 
 1652 
 
 rfgo 
 
 1730 
 1770 
 
 iSii 
 
 1854 
 1897 
 1941 
 1986 
 
 2C32 
 2C.SC 
 2128 
 2178 
 2228 
 
 22.SO 
 2333 
 23<X 
 2443 
 25CC 
 
 2550 
 
 2ti IS 
 
 1079 
 
 2742 
 2815 
 
 2871 
 1938 
 1006 
 
 VJ6 
 
 5148 
 
 II 4 6 
 1172 
 
 095 
 
 1227 
 
 I2.SO 
 
 1285 
 1315 
 1340 
 1.377 
 
 140<i 
 
 1442 
 I47t> 
 ISM 
 
 1.545 
 1581 
 
 1618 
 rtW 
 
 i94 
 
 1734 
 1774 
 
 iSifi 
 
 1858 
 1901 
 1945 
 1991 
 
 2037 
 2084 
 3133 
 3183 
 3334 
 
 2JSC, 
 
 2339 
 2393 
 2449 
 
 art 
 
 3364 
 
 3034 
 25 
 
 3748 
 
 2812 
 
 3877 
 
 2944 
 10 13 
 K3 
 
 it 55 
 
 
 O 
 
 
 
 
 
 z 
 
 i 
 
 z 
 
 2 
 
 2 
 
 2 
 2 
 
 2 
 2 
 
 O 
 
 
 
 
 I 
 I 
 
 I 
 
 I 
 
 z 
 z 
 
 i 
 
 i 
 
 a 
 a 
 
 2 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 O 
 
 
 I 
 I 
 
 I 
 I 
 
 ! 
 
 a 
 a 
 
 a 
 a 
 
 2 
 
 2 
 
 3 
 
 3 
 
 O 
 O 
 
 I 
 I 
 
 I 
 I 
 
 i 
 
 2 
 2 
 
 2 
 
 2 
 
 2 
 2 
 
 3 
 3 
 
 O 
 
 O 
 
 O 
 
 z 
 
 I 
 I 
 
 I 
 I 
 
 z 
 a 
 
 
 
 
 
 a 
 a 
 
 a 
 a 
 
 3 
 3 
 
 3 
 3 
 
 O 
 O 
 
 I 
 I 
 
 ! 
 
 a 
 a' 
 
 
 
 
 
 a 
 
 a 
 
 3 
 
 3 
 3 
 
 3 
 
 3 
 
 3 
 
 O 
 
 I 
 
 I 
 
 a 
 
 a 
 
 3 
 
 O 
 
 
 
 I 
 
 I 
 
 ; 
 
 a 
 a 
 
 a 
 a 
 
 3 
 3 
 
 3 
 3 
 
 4 
 4 
 
 
 
 
 
 
 
 
 
 O 
 
 
 z 
 z 
 
 I 
 
 Z 
 
 I 
 z 
 
 ; 
 
 a 
 a 
 
 a 
 a 
 
 3 
 3 
 
 3 
 3 
 
 4 
 4 
 
 
 
 a 
 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 4 
 
 4 
 4 
 
 z 
 z 
 z 
 
 I 
 I 
 I 
 
 a 
 
 a 
 
 3 
 
 
 
 
 
 
 5 
 5 
 
 
 
 
 
 4 
 
 
 
 
 
 
 
 z 
 z 
 
 z 
 I 
 I 
 
 I 
 
 2 
 2 
 2 
 
 2 
 
 a 
 3 
 
 3 
 3 
 
 3 
 
 3 
 3 
 
 3 
 
 4 
 4 
 4 
 
 4 
 
 4 
 4 
 4 
 
 5 
 
 5 
 5 
 5 
 
 5 
 5 
 5 
 6 
 
 
 z 
 z 
 I 
 
 I 
 I 
 I 
 
 2 
 2 
 2 
 
 3 
 3 
 3 
 
 3 
 4 
 4 
 
 4 
 4 
 4 
 
 5 
 5 
 \
 
 ANTILOGARITHMS 
 
 III 
 
 ANTILOGARITHMS (Continued) 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 PROPORTIONAL PARTS. 
 
 fs 
 
 
 
 I 
 
 2 
 
 3 
 
 4 
 
 s 
 
 6 
 
 7 
 
 8 
 
 9 
 
 
 
 
 
 
 
 
 
 " 
 
 .50 
 
 3162 
 
 3170 
 
 3177 
 
 3184 
 
 3192 
 
 3199 
 
 3206 
 
 3214 
 
 1,221 
 
 3398 
 
 X 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 .51 
 
 52 
 .53 
 
 3236 
 3311 
 3388 
 
 3243 
 3319 
 3396 
 
 3251 
 3327 
 3404 
 
 3258 
 
 3334 
 3412 
 
 3266 
 3342 
 3420 
 
 3273 
 
 3350 
 3428 
 
 3281 
 3357 
 3436 
 
 3289 
 3365 
 
 3443 
 
 V9<~> 
 U73 
 5451 
 
 3304 
 3381 
 
 5459 
 
 l 
 
 X 
 
 2 
 2 
 
 a 
 
 2 
 
 3 
 3 
 
 
 s 
 
 5 
 
 s 
 (, 
 
 
 -51 
 
 3467 
 
 3475 
 
 3483 
 
 3491 
 
 3499 
 
 3508 
 
 35i6 
 
 3524 
 
 is.y 
 
 (540 
 
 X 
 
 a 
 
 2 
 
 3 
 
 
 S 
 
 fl 
 
 
 .55 
 
 3S48 
 
 3556 
 
 3565 
 
 3573 
 
 358i 
 
 3589 
 
 3597 
 
 3606 
 
 <6i.) 
 
 ata 
 
 x 
 
 2 
 
 2 
 
 .i 
 
 
 5 
 
 B 
 
 
 .56 
 
 3631 
 
 3639 
 
 3648 
 
 3656 
 
 3664 
 
 3673 
 
 3681 
 
 3690 
 
 <6<>S 
 
 (707 
 
 X 
 
 2 
 
 .5 
 
 .i 
 
 
 5 
 
 (i 
 
 
 -57 
 
 3715 
 
 3724 
 
 3733 
 
 3741 
 
 3750 
 
 3758 
 
 3767 
 
 3776 
 
 J7S4 
 
 1793 
 
 I 
 
 2 
 
 3 
 
 .i 
 
 
 5 
 
 B 
 
 
 -5.H 
 
 3802 
 
 3811 
 
 3819 
 
 3828 
 
 3837 
 
 3846 
 
 3855 
 
 3864 
 
 073 
 
 (B83 
 
 X 
 
 2 
 
 3 
 
 4 
 
 4 
 
 5 
 
 B 
 
 
 59 
 
 3890 
 
 3899 
 
 3908 
 
 3917 
 
 3926 
 
 /lrT 
 
 3936 
 
 3945 
 
 .-.., 
 
 3954 
 
 Ar,Af, 
 
 |p6a 
 
 W72 
 
 1 ^i\ 1 
 
 1 
 
 2 
 
 .i 
 
 4 
 
 5 
 
 5 
 
 B 
 
 
 .61 
 
 398i 
 
 4074 
 
 399 
 
 4083 
 
 3999 
 4093 
 
 .,00 
 
 4009 
 4102 
 
 4010 
 
 4III 
 
 4O27 
 
 4121 
 
 4O3O 
 4130 
 
 4O4O 
 4140 
 
 1O.S5 
 1150 
 
 |, I),; 
 4159 
 
 I 
 
 a 
 
 3 
 
 4 
 
 s 
 
 fl 
 
 7 
 
 B 
 
 .62 
 .63 
 
 4169 
 4266 
 
 4178 
 4276 
 
 4Ioo 
 
 4285 
 
 4198 
 4295 
 
 4207 
 
 4305 
 
 4217 
 
 4315 
 
 4227 
 4325 
 
 4236 
 
 4335 
 
 4246 
 J345 
 
 I-,S (l 
 4355 
 
 I 
 
 2 
 
 3 
 
 4 
 
 5 
 
 t 
 
 7 
 
 <s 
 B 
 
 .64 
 
 .65 
 
 .66 
 
 4365 
 
 4467 
 4571 
 
 4375 
 
 4477 
 458i 
 
 4385 
 
 4487 
 4592 
 
 4395 
 
 4498 
 4603 
 
 4406 
 
 4508 
 4613 
 
 4416 
 
 4519 
 
 4624 
 
 4426 
 
 4529 
 4634 
 
 4436 
 
 4539 
 4645 
 
 144*' 
 
 455C 
 1656 
 
 4457 
 
 45< 
 466- 
 
 I 
 I 
 
 
 
 
 
 
 
 fl 
 
 .07 
 .tt 
 9 
 
 4677 
 4786 
 4898 
 
 4688 
 4797 
 4909 
 
 4699 
 4808 
 4920 
 
 4710 
 4819 
 4932 
 
 4721 
 
 4831 
 
 4943 
 
 4732 
 
 4842 
 
 4955 
 
 4742 
 4853 
 4966 
 
 4753 
 4864 
 4977 
 
 47<M 
 4*75 
 4989 
 
 4775 
 (887 
 
 SOCO 
 
 X 
 
 2 
 
 3 
 
 5 
 
 (i 
 
 7 
 
 8 
 
 9 
 
 .70 
 
 5012 
 
 5023 
 
 5035 
 
 5047 
 
 5058 
 
 5070 
 
 5082 
 
 5093 
 
 5105 
 
 5117 
 
 I 
 
 a 
 
 
 5 
 
 (i 
 
 7 
 
 8 
 
 9 
 
 .71 
 
 5129 
 
 5140 
 
 5152 
 
 5164 
 
 5176 
 
 5188 
 
 5200 
 
 5212 
 
 S-224 
 
 5236 
 
 l" 
 
 a 
 
 
 5 
 
 B 
 
 7 
 
 8 
 
 1C 
 
 .7* 
 
 5248 
 
 5260 
 
 5272 
 
 5284 
 
 5297 
 
 5309 
 
 5321 
 
 5333 
 
 5346 
 
 S.ivS 
 
 I 
 
 a 
 
 
 5 
 
 c, 
 
 7 
 
 9 
 
 1C 
 
 .73 
 
 5370 
 
 5383 
 
 5395 
 
 5408 
 
 5420 
 
 5433 
 
 5445 
 
 5458 
 
 5470 
 
 $483 
 
 I 
 
 3 
 
 
 5 
 
 6 
 
 S 
 
 9 
 
 1C 
 
 74 
 
 5495 
 
 5508 
 
 5521 
 
 5534 
 
 5546 
 
 5559 
 
 5572 
 
 5585 
 
 559* 
 
 5610 
 
 I 
 
 3 
 
 
 5 
 
 (i 
 
 B 
 
 9 
 
 1C 
 
 75 
 
 5623 
 
 5636 
 
 5649 
 
 5662 
 
 5675 
 
 5689 
 
 5702 
 
 5715 
 
 5736 
 
 5741 
 
 I 
 
 3 
 
 
 5 
 
 
 8 
 
 g 
 
 1C 
 
 -"6 
 
 5754 
 
 5768 
 
 578i 
 
 5794 
 
 5808 
 
 5821 
 
 5834 
 
 5848 
 
 SS6i 
 
 5875 
 
 I 
 
 3 
 
 
 5 
 
 
 8 
 
 9 
 
 II 
 
 -77 
 
 5888 
 
 5902 
 
 59i6 
 
 5929 
 
 5943 
 
 5957 
 
 5970 
 
 5984 
 
 5998 
 
 '012 
 
 I 
 
 3 
 
 
 5 
 
 
 8 
 
 1C 
 
 II 
 
 .7* 
 
 6026 
 
 6039 
 
 6053 
 
 6067 
 
 6081 
 
 6095 
 
 6109 
 
 6124 
 
 5138 
 
 6152 
 
 X 
 
 3 
 
 
 6 
 
 
 8 
 
 1C 
 
 11 
 
 79 
 
 6166 
 
 6180 
 
 6194 
 
 6209 
 
 6223 
 
 6237 
 
 6252 
 
 6266 
 
 L'Kl 
 
 6-95 
 
 I 
 
 3 
 
 
 6 
 
 
 9 
 
 10 
 
 11 
 
 .80 
 
 6310 
 
 6324 
 
 6339 
 
 6353 
 
 6368 
 
 6383 
 
 6397 
 
 6412 
 
 6427 
 
 M4-> 
 
 1 
 
 3 
 
 
 6 
 
 7 
 
 g 
 
 1C 
 
 12 
 
 .81 
 
 6457 
 
 6471 
 
 6486 
 
 6501 
 
 6516 
 
 6531 
 
 6546 
 
 6561 
 
 '577 
 
 6592 
 
 a 
 
 3 
 
 
 (i 
 
 8 
 
 9 
 
 11 
 
 12 
 
 .81 
 
 6607 
 
 6622 
 
 6637 
 
 6653 
 
 6668 
 
 6683 
 
 6699 
 
 6714 
 
 67JO 
 
 '''745 
 
 a 
 
 3 
 
 
 6 
 
 -s 
 
 9 
 
 11 
 
 12 
 
 .83 
 
 6761 
 
 6776 
 
 6792 
 
 6808 
 
 6823 
 
 6S39 
 
 6855 
 
 6871 
 
 >SS7 
 
 61102 
 
 a 
 
 3 
 
 
 6 
 
 -s 
 
 9 
 
 1 1 
 
 13 
 
 .84 
 
 6918 
 
 6934 
 
 6950 
 
 6966 
 
 6982 
 
 6998 
 
 7015 
 
 7031 
 
 7047 
 
 7063 
 
 2 
 
 3 
 
 
 6 
 
 B 
 
 1C 
 
 1 1 
 
 13 
 
 .85 
 
 7079 
 
 7096 
 
 7112 
 
 7129 
 
 7145 
 
 7161 
 
 7178 
 
 7194 
 
 7*1 ' 
 
 7398 
 
 a 
 
 3 
 
 
 7 
 
 B 
 
 1C 
 
 1-' 
 
 13 
 
 .86 
 
 7244 
 
 7261 
 
 7278 
 
 7295 
 
 73" 
 
 7328 
 
 7345 
 
 7362 
 
 737'.i 
 
 739 
 
 a 
 
 
 
 7 
 
 8 
 
 1C 
 
 12 
 
 13 
 
 .87 
 
 7413 
 
 7430 
 
 7447 
 
 7464 
 
 7482 
 
 7499 
 
 75i6 
 
 7534 
 
 755 1 
 
 7568 
 
 a 
 
 
 
 7 
 
 9 
 
 to 
 
 12 
 
 14 
 
 .88 
 
 7586 
 
 7603 
 
 7621 
 
 7638 
 
 7656 
 
 7674 
 
 7691 
 
 7709 
 
 77-'7 
 
 7745 
 
 a 
 
 
 
 7 
 
 g 
 
 XX 
 
 12 
 
 11 
 
 .'. 
 
 7762 
 
 7780 
 
 7798 
 
 7816 
 
 7834 
 
 7852 
 
 7870 
 
 7889 
 
 7907 
 
 7935 
 
 a 
 
 
 5 
 
 7 
 
 g 
 
 XX 
 
 13 
 
 14 
 
 9 
 
 7943 
 
 7962 
 
 798o 
 
 7998 
 
 8017 
 
 8035 
 8222 
 
 8054 
 8241 
 
 8072 
 8260 
 
 Son 
 
 S WO 
 
 sue 
 s "in 
 
 a 
 
 
 B 
 
 i, 
 
 7 
 
 s 
 
 g 
 
 g 
 
 1 1 
 
 13 
 
 5 
 
 .91 
 
 .<> 
 
 8128 
 8318 
 
 8147 
 8337 
 
 8l66 
 
 8356 
 
 8l85 
 
 8375 
 
 8395 
 
 8414 
 
 8433 
 
 8453 
 
 Kr/V 
 
 8472 
 
 Bajm 
 S492 
 
 2 
 
 
 B 
 
 B 
 
 l< 
 
 l-> 
 
 14 
 
 5 
 
 . 93 
 
 8511 
 
 8S3I 
 
 8551 
 
 8570 
 
 8590 
 
 8610 
 
 8630 
 
 8650 
 
 B670 
 
 B6go 
 
 2 
 
 
 6 
 
 1 
 
 to 
 
 13 
 
 l ! 
 
 a 
 
 .94 
 
 8710 
 
 8730 
 
 8750 
 
 8770 
 
 8790 
 
 8810 
 
 8831 
 
 8851 
 
 8873 
 
 8893 
 
 2 
 
 
 6 
 
 1 
 
 i 
 
 i- 
 
 i i 
 
 ii 
 
 9! 
 
 8913 
 
 8933 
 
 8954 
 
 8974 
 
 8995 
 
 9016 
 
 9036 
 
 9057 
 
 P78 
 
 1099 
 
 2 
 
 
 6 
 
 8 
 
 10 
 
 u 
 
 15 
 
 17 
 
 96 
 
 9120 
 
 9141 
 
 9162 
 
 9183 
 
 9204 
 
 9226 
 
 9247 
 
 9268 
 
 age 
 
 Ml" 
 
 a 
 
 
 I 
 
 I 
 
 ii 
 
 (3 
 
 
 
 7 
 
 >97 
 
 9333 
 
 9354 
 
 9376 
 
 9397 
 
 9419 
 
 9441 
 
 9462 
 
 9484 
 
 9SOC 
 
 WaS 
 
 a 
 
 
 7 
 
 g 
 
 n 
 
 1 3 
 
 15 
 
 7 
 
 98 
 
 9550 
 
 9572 
 
 9594 
 
 9616 
 
 9638 
 
 9661 
 
 9683 
 
 9705 
 
 ym 
 
 WS< 
 
 2 
 
 
 7 
 
 g 
 
 XI 
 
 13 
 
 If, 
 
 18 
 
 99 
 
 9772 
 
 9795 
 
 9817 
 
 9840 
 
 9863 
 
 9886 
 
 9908 
 
 9931 
 
 9954 
 
 '('177 
 
 2 
 
 
 7 
 
 9 
 
 XX 
 
 14 
 
 16 
 
 tl
 
 112 
 
 QUANTITATIVE ANALYSIS 
 
 INTERNATIONAL ATOMIC WEIGHTS, 1921. 
 
 
 Sym- 
 bol. 
 
 Atomic 
 Weight 
 
 
 Sym- 
 bol. 
 
 Atomic 
 Weight. 
 
 Aluminum 
 
 Al 
 
 27 i 
 
 Neodymium 
 
 Nd 
 
 
 Antimony . 
 
 Sb 
 
 I 2O. 2 
 
 Neon 
 
 Ne 
 
 
 Argon 
 
 A 
 
 39.88 
 
 Nickel 
 
 Ni 
 
 58 68 
 
 Arsenic 
 
 As 
 
 74.96 
 
 Niton (radium emana 
 
 
 
 Barium 
 
 Ba 
 
 X 37 37 
 
 tion) 
 
 Nt 
 
 
 Bismuth 
 
 Bi 
 
 209 o 
 
 Nitrogen 
 
 N 
 
 
 Boron. 
 
 B 
 
 no 
 
 Osmium. 
 
 Os 
 
 
 Bromine 
 Cadmium 
 
 Br 
 Cd 
 
 79-92 
 112.40 
 
 Oxygen 
 Palladium 
 
 
 Pd 
 
 iG.OO 
 
 106 7 
 
 Caesium 
 
 Calcium 
 
 Cs 
 Ca 
 
 132.81 
 4 07 
 
 Phosphorus 
 Platinum 
 
 P 
 Pt 
 
 31-04 
 
 Carbon 
 
 c 
 
 12 OO5 
 
 Potassium 
 
 K 
 
 
 Cerium 
 
 Ce 
 
 140.25 
 
 Praseodymium. 
 
 Pr 
 
 
 Chlorine 
 
 Cl 
 
 35.46 
 
 Radium 
 
 Ra 
 
 226 o 
 
 Chromium 
 
 Cr 
 
 52.0 
 
 Rhodium 
 
 Rh 
 
 102 9 
 
 Cobalt 
 
 Co 
 
 58 97 
 
 Rubidium 
 
 Rb 
 
 gr A? 
 
 Columbium 
 
 Cb 
 
 93 r 
 
 Ruthenium . 
 
 Ru 
 
 
 Copper 
 
 Cu 
 
 63 .57 
 
 Samarium 
 
 Sa 
 
 
 Dysprosium 
 Erbium 
 
 Dy 
 Er 
 
 162.5 
 167 7 
 
 Scandium 
 Selenium 
 
 Sc 
 Se 
 
 45-i 
 
 Europium 
 
 Eu 
 
 52 o 
 
 Silicon 
 
 Si 
 
 28 i 
 
 Fluorine 
 
 F 
 
 19 o 
 
 Silver . 
 
 Ag 
 
 107 88 
 
 Gadolinium 
 Gallium 
 Germanium. : 
 Glucinum 
 
 Gd 
 Ga 
 Ge 
 Gl 
 
 57-3 
 70.1 
 
 72.5 
 9 J 
 
 Sodium 
 Strontium 
 Sulphur 
 
 Na 
 Sr 
 S 
 Ta 
 
 23.00 
 87.63 
 32.06 
 181 5 
 
 Gold 
 
 Helium 
 
 Au 
 He 
 
 97-2 
 4 oo 
 
 Tellurium 
 Terbium. 
 
 Te 
 Tb 
 
 127-5 
 
 ICQ 2 
 
 Holmium 
 
 Ho 
 
 63 5 
 
 Thallium. 
 
 Tl 
 
 
 
 H 
 
 i 008 
 
 
 Th 
 
 2 ?2 l*i 
 
 Indium 
 
 In 
 
 14.8 
 
 Thulium 
 
 Tm 
 
 
 Iodine 
 
 I 
 
 2'6 92 
 
 
 Sn 
 
 Il8 7 
 
 Iridium 
 Iron 
 
 Ir 
 Fe 
 
 93-i 
 55 84 
 
 Titanium 
 Tungsten 
 
 Ti 
 W 
 
 4 8.1 
 
 184 o 
 
 Krypton . 
 
 Kr 
 
 82 92 
 
 Uranium 
 
 u 
 
 2*8 2 
 
 Lanthanum 
 Lead 
 Lithium 
 
 La 
 Pb 
 Li 
 
 39-o 
 
 07 .20 
 
 6 .94 
 
 Vanadium 
 Xenon 
 Ytterbium (Neoytter- 
 
 V 
 Xe 
 
 51-0 
 130.2 
 
 Lutecium 
 
 Lu 
 
 75 - 
 
 bium) 
 
 Yb 
 
 173.5 
 
 Magnesium 
 
 Mg 
 
 24.32 
 
 Yttrium 
 
 Yt 
 
 89.33 
 
 Manganese. . . 
 
 Mn 
 
 54 93 
 
 Zinc. 
 
 Zn 
 
 65 37 
 
 Mercury 
 
 Hg 
 
 oo 6 
 
 Zirconium 
 
 Zr 
 
 90.6 
 
 Molybdenum 
 
 Mo 
 
 96.0 
 
 

 
 INDEX 
 
 PAGE 
 
 Alkalies, determination of in silicate 9 
 
 Allen and Bishop, method for available sulphur in pyrite 31 
 
 Ash, determination in coal 47 
 
 Atomic weights 112 
 
 Bamber, method for sulphur 87 
 
 Barba, method for chromium 94 
 
 Barneby and Isham, method for titanium in iron ore 42 
 
 Basic acetate method 16 
 
 Brunck, method for nickel 92 
 
 Calcium, determination in iron ore '. 18 
 
 Carbon, apparatus, for 57 
 
 determination in steel 59 
 
 forms and effects upon physical properties 61 
 
 Carbon-iron equilibrium diagram 63 
 
 Chromium, determination in chrome-vanadium steel 97 
 
 determination in steel 95 
 
 Coal, analysis of 47 
 
 determination of ash in 47 
 
 determination of moisture in 47 
 
 determination of sulphur in 48 
 
 determination of volatile matter in : 48 
 
 Copper, determination in bronze 54 
 
 determination in duralumin 104 
 
 determination in steel 91 
 
 volumetric determination in ore 43 
 
 Di-methyl glyoxime method for nickel 92 
 
 Drown, method for silicon in iron 89 
 
 Duralumin, analysis of 103 
 
 Feldspar, decomposition by means of sodium carbonate 5 
 
 determination of silica in 5 
 
 determination of alkalies in 9 
 
 Fresenius, method for total sulphur 32 
 
 Gooch, method for titanium 39 
 
 Graphite, determination in cast iron 61 
 
 "3
 
 1 14 INDEX 
 
 PAGE 
 
 Handy, method for phosphorus in steel 82 
 
 Iodine, standardization of solution 83 
 
 Iron, determination in duralumin 104 
 
 determination in spathic iron ore 20 
 
 standard for potassium permanganate 75 
 
 Laumontite, determination of silica in 4 
 
 Lead, determination in bronze 54 
 
 Logarithms 108 
 
 Low, A. H., iodimetric method for copper 43 
 
 Lunge, method for available sulphur in pyrite 29 
 
 Magnesium, determination in iron ore 18 
 
 determination in duralumin 104 
 
 Manganese, determination in duralumin 105 
 
 determination in steel 65, 69, 72 
 
 gravimetric determination in iron ore 17 
 
 Moisture, determination in coal 47 
 
 Nickel, determination in steel 92 
 
 Phosphor bronze, analysis of 50 
 
 Phosphorus, alkalimetric method 82 
 
 determination in bronze 57 
 
 determination in steel 75 
 
 ferric alum method 77 
 
 reductor method 75 
 
 Potassium, determination in feldspar 9 
 
 determination as perchlorate 13 
 
 permanganate solution, standardization of 20, 21 
 
 permanganate, standardization in Volhard method 65 
 
 Sampling i 
 
 Silica, determination in decomposable silicate 4 
 
 determination in refractory silicate 5 
 
 determination in spathic iron ore 15 
 
 dehydration of in silicates 4 
 
 ignition of i 5 
 
 testing the purity of 5 
 
 Silicate, determination of alkalies in 9 
 
 Silicon, determination in cast iron 89 
 
 determination in duralumin 103 
 
 Smith, J. L., determination of alkalies in silicate 9 
 
 Sodium, determination in feldspar 9 
 
 oxalate, standard for potassium permanganate 21
 
 INDEX 115 
 
 PAGE 
 
 Sodium, peroxide method for sulphur in pyrite 33 
 
 thiosulphate, standardization of 43 
 
 Spathic iron ore, analysis of 15 
 
 determination of calcium in 18 
 
 determination of iron in 20 
 
 determination of magnesium in 18 
 
 determination of manganese in 17 
 
 determination of oxides of iron and alumina in 19 
 
 determination of silica in 15 
 
 Sulphur, determination in cast iron 87 
 
 determination in steel by evolution method 83 
 
 determination in pyrite 29, 31, 32 
 
 evolution method 83 
 
 gravimetric method in cast iron 87 
 
 Tin, determination in bronze 50 
 
 Titanium, determination in iron ore 39 
 
 Tungsten, determination in steel 98 
 
 Vanadium, determination in steel 101 
 
 Volatile matter, determination in coal 48 
 
 Volhard method for manganese 65 
 
 Williams method for manganese 69 
 
 Zimmermann-Reinhardt method for iron 20
 
 UNIVERSITY OF CALIFORNIA AT LOS ANGELES 
 
 THE UNIVERSITY LIBRARY 
 This book is DUE on the last date stamped below 
 
 2S 19! i* 
 
 feft ' 
 
 12 193F 
 
 21 t99S 
 
 1 1935 
 
 idaet 
 
 DEC 3 1935 
 
 )G i * 
 
 4- J937 
 
 29 
 
 .""i 6 ? W38 
 
 Form L-9-157n-3,'34 
 
 MOV 
 
 APR 5 - 195b 
 SEP 18 195*
 
 QD 
 
 101 
 
 F29 
 
 Pay - 
 
 crmrse in 
 
 quantitative 
 
 analysis.