LIBRARY UNIVERSITY OF CALIFORNIA. Deceive J Accessions Na7~. O,m No. QUANTITATIVE CHEMICAL ANALYSIS BY ELECTROLYSIS. BY DR. ALEXANDER CLASSEN, PROFESSOR OP CHEMISTRY AND DIRECTOR OF THE INORGANIC LABORATORY IN THE ROYAL SCHOOL OF TECHNOLOGY AT AACHEN. ranslation> SECOND ENGLISH. FROM THE THIRD GERMAN EDITION, REVISED AND GREATLY ENLARGED. BY WILLIAM HALE HERRIOK, A.M., FORMERLY PROFESSOR OF CHEMISTRY IN IOWA COLLEGE, AND IN THE PENNSYLVANIA STATE COLLEGE. SECOND THOUSAND. NEW YORK: JOHN WILEY & SONS, 53 EAST TENTH STREET. 1894. COPYRIGHT, 1887, 1894, BY WILLIAM HALE HERRICK. TRANSLATOR'S PREFACE TO FIRST EDITION. THE attention of the translator was drawn to the original work, of which the following is a translation, by finding it to be the only source of knowledge of the subject outside of scattered articles in the journals, and thus convenient and almost indispensable, as a laboratory handbook, to himself and advanced students. It is in the hope of rendering it more available to all who may have occasion to use electrolytic methods in quantitative analysis, and of increasing and stimulating the use of these valuable methods, that he has undertaken the translation. Twelve or fifteen years ago, when the translator was a student, electrolytic methods in analysis were practically unknown in one of the first laboratories of the country ; they still occupy a very subordinate position in many of the leading laboratories. The translator believes that they can, with advantage, be much more widely used in this country, in both scientific and technical laboratories, especially where a current from a dynamo, or power to run a small dynamo, is available. He trusts that he may be found to have contributed to such a result by iii iv TRANSLATOR'S PREFACE. rendering available, in English, this complete and standard work. As the original is the work of a specialist, and recently published, the translator has had little occasion to add to his labors those of . an .editor. A few .additions have been made, touching matters of more recent date that have fallen under his eye. All such additions are enclosed in brackets. The thanks of the translator are due to Prof. O. D. Allen, of the Sheffield Scientific School of Yale College r for revision of portions of the translation relating to the analysis of metallurgical products ; and to his colleagues,. Profs. I. Thornton Osmond and William Frear, to the former for revision of the portion of the proof relating to the use of the dynamo, and to the latter for frequent suggestions throughout the progress of the work. WILLIAM HALE HERRICK. THE PENNSYLVANIA STATE COLLEGE, CHEMICAL LABORATORY, June, 1887. PREFACE TO SECOND EDITION. SINCE the appearance of the first edition of this book, I have devoted myself exclusively to experimental work having for its object the further development of quantitative analysis by electrolysis as an independent branch of the subject. This new method of analysis may now be considered as established in its essential points. The great advantage of quantitative electrolysis, apart from its greater simplicity, lies unquestionably in the fact that the electric current does the work of the analyst, setting him free to carry on other work. Long experience has shown, that, if the methods are correctly followed, even unskilled analysts obtain results that experienced chemists can with difficulty reach by the ordinary methods of gravimetric analysis. Since a large number of the most different analyses can now be carried on at the same time (a result hitherto impossible to attain), I may be permitted to hope that analysis by electrolytic methods may come more and more into use in scientific and technical laboratories. A. CLASSED AACHEN, September, 1885. TRANSLATOR'S PREFACE TO SECOND EDITION. LITTLE needs to be said in introducing this new edition. The author, in his preface, calls attention to the many addi- tions that have been made. It will be seen that the majority of these additions record the work of others than the author, so that while the earlier editions were largely a record of the work of the Aachen school, this edition is rather a compend of the literature of the subject. This fact alone is sufficient evidence of the great advance, within seven years, in the use and development of electrolytic methods. As before, it has not seemed wise to add much new mat- ter. Thanks are hereby rendered to Professor Edgar F. Smith, our leading American worker in this field, for free permission to use his " Electro-chemical Analysis." It has not seemed best, however, to transfer methods from Professor Smith's book, as practical workers would doubtless have access to both books, and would prefer to consult the original. For the same reason, no attempt has been made to summarize periodical literature on this subject, for the two years since the publication of the German edition. Thanks are due to Dr. Leonard Paget, for placing at the disposal of the translator and users of this book his simple and efficient thermopile. WILLIAM HALE HEREICK. NEW YORK, August, 1894. vi PREFACE TO THIRD EDITION. IN the present edition will be found a large number of new methods and improvements. In the preface to the second edition I expressed the hope that electrolytic analysis would find increasing use in both .scientific and technical laboratories. The methods of elec- trolysis are to-day taught as special methods of quantitative analysis, like gas and spectroscopic analysis, in most of the laboratories of the higher institutions of learning in this and other countries (this book has been translated into French by Professor Bias and into English by Professor Herrick), and are employed for scientific researches, as well as for the pro- duction of absolutely pure metals and determinations of atomic weight. It fills me with satisfaction that the introduction of elec- trolytic analysis into the laboratories of the great industries has made such progress; in some such laboratories two or three thousand determinations are made per year. The statements of various associations of metal-workers, large foundry-owners, etc., agree in emphasizing the decided advantages of electrolytic methods over the gravimetric methods hitherto in use, and their indispensability in certain cases. The special advantages of these methods, according to vii Vlll PREFACE TO THIRD EDITION. the above-mentioned testimony, are accuracy and rapidity. These two factors have made it possible to utilize electrolysis for complete and exact control of certain metallurgical opera- tions, and, as more fully stated elsewhere in this book, for the purchase of ores in foreign markets; results often impossible by the use of previous gravimetric methods. A. CLASSEN. AACHEN, July, 1892. CONTENTS. PAKT I. GENERAL PART INTRODUCTION. PAGE GALVANIC BATTERIES 7 THE LECLANCHE CELL 8 THE MEIDINGER CELL 9 THE DANIELL CELL 11 THE GROVE CELL 14 THE BUNSEN CELL 15 THERMO-ELECTRIC PILES 16 CLAMOND THERMOPILE 16 NOE THERMOPILE 18 PAGET THERMOPILE 21 ELECTRICAL MACHINES 24 LABORATORY DYNAMO AND ACCOMPANYING APPARATUS FOR ELECTRO- LYTIC PURPOSES 25 SECONDARY BATTERIES 37 THE USE OF ACCUMULATORS, AND THEIR ADVANTAGES OVER GAL- VANIC BATTERIES 40 APPARATUS FOR REDUCING THE STRENGTH OF THE CURRENT. RESISTANCES 48 MEASUREMENT OF THE STRENGTH OF THE CURRENT 57 PROCESS OF ANALYSIS. ELECTRODES, SUPPORTS, ETC 64 GRAVIMETRIC DETERMINATION OF METALS. DETERMINATION OF IRON 78 " COBALT 81 " NICKEL 82 ix X CONTENTS. PAGE DETERMINATION OF ZINC 83 " "MANGANESE 85 " " ALUMINIUM, CHROMIUM, URANIUM, BERYLLIUM. 8? " "COPPER , 88 " " BISMUTH 91 "CADMIUM 94 "LEAD 96 " THALLIUM 99 " " SILVER 100 " "MERCURY 102 "PLATINUM 104 " " PALLADIUM 106 " GOLD 106 " ANTIMONY 106 " TIN 110 "ARSENIC 113 " " POTASSIUM, AMMONIA (NITROGEN) 113 " " NITRIC ACID IN NITRATES. . .113 SEPARATION OF METALS. IRON AND COBALT 115 " " NICKEL 115 " FROM COBALT AND NICKEL 116 " AND ZINC 116 IRON, COBALT, NICKEL, AND ZINC, FROM ALUMINIUM 117 IRON FROM MANGANESE 118 NICKEL FROM MANGANESE 121 COBALT AND ZINC FROM MANGANESE 1 22 NICKEL, COPPER, CADMIUM, ZINC, AND MERCURY, FROM MANGANESE 122 MANGANESE FROM COPPER, CADMIUM, MERCURY 123 IRON, COBALT, NICKEL, AND ZINC, FROM MANGANESE AND ALU- MINIUM 123 IRON, COBALT, NICKEL, AND ZINC, FROM CHROMIUM 123 " " " " " "* " AND ALUMINIUM. 124 IRON, COBALT, NICKEL, AND ZINC, FROM MANGANESE, CHROMIUM, AND ALUMINIUM 124 IRON, COBALT, NICKEL, AND ZINC, FROM URANIUM 125 " " " " " CHROMIUM AND URANIUM. . 125 IRON, NICKEL, COBALT, AND ZINC, FROM ALUMINIUM, MAGNESIUM, AND URANIUM 126 MANGANESE FROM BARIUM, STRONTIUM, CALCIUM, MAGNESIUM, AND ALKALIES : . . 126 CONTENTS. XI PAGE IRON FROM BERYLLIUM 12$ " " " AND ALUMINIUM 127 " " ZIRCON 127 " " VANADIUM , 127 " " MANGANESE AND PHOSPHORIC ACID 127 *' " MANGANESE, ALUMINIUM, AND PHOSPHORIC ACID 128 " " " AND SULPHURIC ACID 129 COPPER FROM BISMUTH . . 129 " " CADMIUM 130 " LEAD 180 " " SILVER 130 " " ANTIMONY AND ARSENIC 131 " TIN... 138 COPPER FROM IRON, COBALT, NICKEL, ZINC, MANGANESE, CHROMIUM, ALUMINIUM, AND PHOSPHORIC ACID 133 COPPER FROM BARIUM, STRONTIUM, CALCIUM, POTASSIUM, SODIUM, AND LITHIUM 133 BISMUTH FROM IRON, NICKEL, COBALT, ZINC. MANGANESE, CADMIUM, CHROMIUM. ALUMINIUM, AND URANIUM 134 LEAD FROM CADMIUM 134 ' " BISMUTH 134 " " SILVER 134 " MERCURY 135 LEAD FROM IRON, COBALT, NICKEL, ZINC, CHROMIUM, AND ALU- MINIUM 135 CADMIUM FROM ZINC .... 1 35 NICKEL AND COBALT 137 CADMIUM AND BISMUTH FROM MANGANESE, CHROMIUM, AND ALU- MINIUM 1 37 MERCURY FROM SILVER 137 " " COPPER 137 " ARSENIC AND PALLADIUM 138 MERCURY FROM IRON, COBALT, NICKEL, ZINC, MANGANESE, CHRO- MIUM, AND ALUMINIUM 138 ANTIMONY FROM TIN 138 " " ARSENIC 140 ANTIMONY, ARSENIC, AND TIN 141 TIN FROM PHOSPHORIC ACID 144 PLATINUM FROM IKIDIUM 144 SEPARATION OF GOLD FROM OTHER METALS 144 POTASSIUM FROM SODIUM 145 SODIUM AND AMMONIA 145 Xll CONTENTS. PART II. SPECIAL PART. PAGE ALLOY OF COPPER AND ZINC [LEAD, IRON] (BRASS) 14? " SILVER (SILVER COIN) 148 " " TIN AND LEAD (SOLDER) 149 " " LEAD AND BISMUTH . . 149 " " " ZINC .. 149 " " BISMUTH AND COPPER 150 " COPPER AND TIN (BRONZE) 150 " COPPER, TIN, ZINC, AND PHOSPHORUS (Piiospiioit BRONZE) 131 ALLOY OF COPPER, TIN, ZINC, MANGANESE, AND PH SPHORUS (M AN- GANESE-PHOSPHOR BRONZE) 1 51 ALLOY OF NICKEL AND COPPER (NICKEL COIN) 151 " COPPER, ZINC, AND NICKEL (GERMAN SILVER) 152 " TIN, LEAD, BISMUTH, AND CADMIUM (WOOD'S METAL). . . . 153 " ' TIN, LEAD, BISMUTH, AND MERCURY 153 " LEAD AND ANTIMONY (HARD LEAD. TYPE METAL; 154 " " ANTIMONY AND TIN 154 " " " ARSENIC .... 154 " " ANTIMONY, TIN, AND ARSENIC 155 SPATHIC IRON 155 HEMATITE 156 LTMONITE t 158 CLAY IRONSTONE 158 BOG IRON ORE 158 CHROME IRON-ORE 159 PSILOMELANE 160 SPHALERITE 163 CALAMINE AND SMITHSONITE 165 ULTRAMARINE. 165 REFINERY SLAG 166 COPPER AND LEAD SLAGS 1 66 BLAST FURNACE, CUPOLA, AND BESSEMER SLAGS 168 ZIRCON 169 ARSENOPYRITE 169 COPPER PYRITES 170 NICKEL MATTE. COPPER MATTE 171 COPPER SPEISS. LEAD SPEISS 172 PYRARGYRITE 173 TETRAHEDRITE 173 FURNACE Sows 174 CONTENTS. xiii PAGE STJBNITE f ANTIMONY GLANCE). 175 ULLMANITE. . . . 175 BOURNONITE 176 ZlNKENITE 176 LlNNAEITE 177 COBALTITE 177 COBALTIFEROUS ARSENOPYKITE 178 CERUSSITE . 179 GALENA 179 PYKOMORPHITE '.... 180 LEAD MATTE 180 CINNABAR '.'.... 181 BlSMUTHINITE ... 181 URANINITE (PITCHBLENDE) 182 SOFT LEAD .". 184 HARD LEAD 186 ANTIMONY ... 187 SPELTER . ; . . . 187 BLISTER COPPER 189 REFINED COPPER . 191 TIN . . . . . 192 BISMUTH 192 SILVER 193 NICKEL 198 PIG IRON, STEEL, SPIEGEL, FERROMANGANESE J94 PART III. TABLES FOR CALCULATION OF ANALYSES 199 REAGENTS 203 POTASSIUM OXALATE 203 AMMONIUM OXALATE 203 OXALIC ACID 204 AMMONIUM SULPHATE 204 SODIUM SULPHIDE 204 ALCOHOL 205- ANALYTICAL RESULTS . . . 206 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. PAET L-GESTEEAL PART. INTRODUCTION.* WATER acidified with sulphuric acid is decomposed into its elements, hydrogen and oxygen, when a galvanic current is passed through it; a large number of compound sub- stances conduct themselves in a similar manner. This gal- vanic decomposition is called electrolysis, and the substances which are decomposed by the electric current are known as electrolytes. The substances into which electrolytes are separated by the electric current are naturally divided into two groups : Those which separate at the positive electrode, or anode (connected with the -}- pole of the source of the current), and which are therefore the electro-negative con- stituents, are called anions; those which separate at the negative electrode, or kathode (connected with the pole of the source of the current), the electro-positive constitu- ents, are called kathions. The metalloids, or electro-negative acid groups, therefore appear at the positive electrode, while the metals are sepa- rated at the negative electrode. * An elementary knowledge of galvanic action is assumed. 2 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. For instance, if the electric current is passed through the solution of a haloid salt, the halogen is separated at the anode, the metal at the kathode. CuCl 2 = C1 2 + Cu, = C1 2 Oxygen salts act in a similar manner. CuSO 4 = SO 4 + Cu, Cu(N0 3 ) 2 = (N0 3 ) 2 + Cu. Many acids are decomposed in a similar manner.* H 2 S0 4 = SO 4 + H 2 , 2HC1 = C1 2 + H 2 . The substances formed by electrolytic decomposition, however, generally undergo further chemical change, or are acted on by the electrodes ; various secondary reactions take place. In the electrolysis of a solution of copper sulphate between platinum electrodes, the secondary process consists in the re- action with water of the group SO 4 , which cannot exist uncombined. SQ 4 + H 2 O = H 2 S0 4 + O. T ;i accordance with this reaction, oxygen gas is given off at the positive pole. In the electrolysis of hydrochloric acid, the chlorine set * Some acids are not decomposed by the electric current; e.g., silicic, carbonic, and boric acids. INTRODUCTION. 3 free at the anode reacts with water, forming hypochlorous acid, chloric acid, perchloric acid, etc. Similar secondary reactions are observed in the electrolysis of chlorides. If a solution of ammonium chloride, for example, is submitted to electrolysis, the nascent chlorine acts on the undecomposed salt, with the production, among other substances, of nitro- gen, or nitrogen chloride. Haloid salts of the alkaline earths show similar phenomena. Nitric acid is decomposed by electrolysis in accordance with the re-action 8HNO 3 = 8N0 2 + 8O (anode), 8H (kathode). The nascent hydrogen acts secondarily on nitric acid : 8H + HN0 3 = NH 3 + 3H 2 O. In the presence of sulphuric acid, or a sulphate, this decomposition is complete, the final product being ammonium sulphate. This decomposition of nitric acid is of practical impor- tance in chemical analysis. From a nitric-acid solution which contains copper and zinc, the former metal only is reduced ; this fact can be utilized for the separation of the two metals. If, now, the current is allowed to pass for a long time after the reduction of the copper, the nitric acid is gradually converted into ammonia, and the zinc then separates from the solution. In the electrolysis of a solution of lead nitrate, a complete secondary re-action occurs, the ozonized oxygen (or H 2 O 2 ) which is formed at the anode re-acting on the lead salt with formation of lead peroxide. If sufficient free nitric acid is 4 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. present, all the lead separates as peroxide at the positive electrode. Pb(N0 3 ) 2 = 2N0 2 + 2 (anode), Pb, Pb(N0 3 ) 2 + 2 = Pb0 2 . In the electrolysis of salts of manganese, bismuth, etc., similar decompositions occur. If the salts of metals which decompose water at ordinary temperatures (alkalies and alkaline earths) are electrolyzed, secondary re-actions occur at the negative electrode : K 2 SO 4 = SO 4 (anode), K 2 (kathode), (SO 4 + H 2 O = H 2 SO 4 + O), K 2 + 2H 2 = 2KOH + H 2 . The decomposition products, then, are sulphuric acid and oxygen at one electrode, potassium hydroxide and hydrogen at the other. A similar reaction occurs in the electrolysis of chromic acid in the presence of a free acid : 2CrO 3 = O 6 (anode), Cr 2 (kathode), Cr 2 + 3H 2 O = Cr 2 O 3 + 3H 2 . The metals disengaged at the negative electrode may yield secondary products by acting on the solution. So, for instance, in the electrolysis of cupric chloride, the separated copper reacts with the cupric chloride to form cuprous chloride ; copper acetate yields, at the kathode, a mixture of copper and cupric (or cuprous) oxide. INTRODUCTION. 5 In the electrolysis of organic compounds, the groups set free at an electrode may be decomposed in a manner analo- gous to that noted in inorganic compounds, and yield various products. The electrolysis of potassium acetate should yield, as final products, potassium (potassium hydroxide) and acetic acid. CH 3 COOK = K + CH 3 COO. K -u CHsCOO + H.O = KOH + CH 8 COOH. Instead of this, the acetic acid splits either into carbon dioxide and methyl (dimethyl), or ethylene is formed by the action of oxygen on the dimethyl. Potassium valerianate yields, in addition to valerianic acid, carbon dioxide and dibutyl ; the latter is oxidized by continued electrolysis to isobutylene and water. Sodium succinate yields, among other products, ethylene and carbon dioxide ; potassium lactate breaks up into carbon dioxide and aldehyde. If a solution contains several metals, secondary reactions may occur as follows: one of the metals separated at the negative electrode being more strongly electro-positive than the other which is present in the solution acts upon it, and sets free an equivalent weight of it. For instance, in the electrolysis of a mixture of copper and zinc sulphates, copper and zinc separate out at the negative electrode. The latter sets copper free, as follows : Zn + CuSO 4 = Cu + ZnSO 4 . The elements separated by the electric current sometimes recombine, and exert an electro-motive force opposite to that exerted by the primary current. The electro-motive force thus produced is named polarization of the electrodes. \ \ 6 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. Such counter-currents may also be produced by the union of gases disengaged at the anode and kathode. This counter-current may cause the solution of the metal separated at the negative electrode. So, in the electrolysis- of a solution of copper sulphate, it may happen, under certain circumstances, that the free sulphuric acid formed acts again on the copper to form the sulphate. For the purposes of quantitative chemical analysis, only such solutions are adapted, as indicated by the foregoing, as are decomposed completely by the current without the formation of injurious intermediate products. Solutions which contain a free inorganic acid are well adapted to electrolysis, because they are readily decomposed. In the presence of a free acid, however, only a few metals (e.g., copper, mercury, and platinum) separate completely ; so that such solutions are capable of only limited use. Of all compounds of the metals, the double oxalates are the best adapted to quantitative analysis.* Oxalic acid i& decomposed by the electric current : C 2 H 2 O 4 = 2CO 2 (anode), H 2 (kathode). When potassium oxalate is subjected to electrolysis, the principal decomposition-products are : K 2 C 2 O 4 = 2CO 2 (anode), K 2 (kathode), K 2 + 2H 2 O = 2KOH + H 2 , 2KHO + 2C0 2 = 2KHCO 3 . * Classen, Ber. d ch Ges., 14,1622. 2771 ; 17,2467 ; 18,1104, 1687 ; 19 r 823 ; 20, 5U-I ,21, ^ GALVANIC BATTERIES. 7 When ammonium oxalate is used, the decomposition- products are hydrogen and hydrogen ammonium carbonate. The latter is partly redecomposed into ammonia and carbon dioxide. In the electrolysis of double oxalates, e.g., of zinc ammo- nium oxalate, decomposition takes place as follows : Zinc oxalate breaks up into zinc and carbon dioxide, and ammoni- um oxalate into ammonium and carbon dioxide. The carbon dioxide, which separates at the positive pole, combines with the ammonium to form hydrogen ammonium carbonate, as above explained. In the decomposition of oxalates, there are no secondary reactions nor counter-currents unfavorable to the electrolysis. All oxalates are decomposed by the electric current with greater or less ease, and the reduced metals are not attacked by the decomposition-products, even when the current be- comes weaker during the reaction. When the reaction is complete, the solution can be poured off at once, and the weight of the separated metal determined. (See further details later.) GALVANIC BATTERIES. The intensity of the current necessary for the reduction of different metals varies greatly. Copper, cadmium, bis- muth, and platinum, e.g., are precipitated from their solu- tions by very weak currents ; while iron, nickel, cobalt, zinc, and other metals require stronger currents. For the former, Meininger, Daniell, and Leclanche' cells are much used; for stronger currents, those of Bunsen and Grove. The electro-motive force of these cells is very unlike ; for instance, that of a Daniell cell is 1.079 volts, of a 8 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. Leclanche* cell 1.48, of a Bunsen cell 1.80, of a Grove cell 1.81. LECLANCHE CELL. This is a one-fluid cell using a solution of ammonium chloride, which surrounds the negative pole, the zinc. The FIG. 1. cell is much used in the form shown in Fig. 1. In the jar, which is square in section, with a rounded projection at one corner, stands a porous clay cup, from which projects a block of carbon K surrounded by coarsely pulverized man- ganese dioxide, or a mixture of manganese dioxide and retort carbon. In the projecting rounded corner is a stout rod Z of amalgamated zinc. The carbon and zinc are both provided with binding screws, and are immersed in a concentrated solution of ammonium chloride. MEIDINGER CELL. 9 Leclanche* also uses, in place of the powdered man- ganese dioxide, compressed prisms (shown in Fig. 2) con- FIG. 2. sisting of 40 parts manganese dioxide, 55 parts gas carbon, and 5 parts shellac ; a little potassium sulphate is also added to increase the conductivity. The porous cup is thus dispensed with. MEIDINGER CELL. In contrast to the Leclanche' cell, that of Meidinger con- tains two liquids, solutions of magnesium and copper sul- phates. The element is constructed as follows : In the glass vessel G (Fig. 3) stands a smaller glass g, and in this a copper cylinder K to which an insulated copper wire D is fastened. 10 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. A second cylinder Z of zinc, to which the projecting wire j is fastened, is placed in the upper part of the vessel G. The balloon-shaped glass B, filled with crystals of copper sulphate, closes the cell. The cell is filled to about three-fourths of its capacity with a solution of 1 part crystallized mag- nesium sulphate in 7 parts of water ; and the balloon- shaped flask containing copper sulphate is filled up with water, closed with a stopper fitted with the glass tube r, and, as the FIG. 3. cut shows, inverted in the cell. The cells used by the author have the following dimen- sions : Jar .... Flask. . . . Zinc cylinder . Copper cylinder 21.5 cm. high, 15 cm. diameter. 21.0 " 10.0 " " 12 " " 6.5 " " 6 " " The strength of the current from a battery of Meidinger cells was determined as follows : OH Gas per Minute. 2 Meidinger elements gave, in the voltameter, 0.3 cc. 4 Meidinger elements gave, in the voltameter, 0.4 cc. 6 Meidinger elements gave, in the voltameter, 0.7 cc. DANIELL CELL. 11 6 freshly filled Meidinger cells gave, after two days' use, 1.5 cc. ; after eight days, 1.9 ; after fourteen days, 2.6 ; after three weeks, 3.1 cc. oxyhydrogen gas. After this time, the battery remained constant. DANIELL CELL. In a jar of glass (Fig. 4) is a porous clay cup T, and in this a cylinder of cast zinc, the negative pole (Fig. 5). The FIG. 4. FIG. 5. porous cup is surrounded by a cylinder of sheet-copper K, the positive pole. The cylinder of amalgamated zinc * stands in dilute sul- * The zinc is easily amalgamated by plunging it into mercury, on the 12 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. phuric acid (1 : 20), and the copper cylinder in a solution of -copper sulphate ; the sulphuric acid may be replaced by a solution of zinc sulphate. The author has in use Daniell cells of the following dimensions : Jar . . . . . 20.0 cm. high, 10.4 cm. diameter. Porous cup . . 20.4 " " 7.0 " " Copper cylinder, 17.0 " " 8.5 " " Zinc cylinder . 19.0 " " 6.0 " " [The modification of the Daniell cell known as the gravity cell is the form commonly in use for telegraph batteries in this country, and is the cheapest and most convenient cell for constant batteries to yield currents of moderate strength in scientific laboratories. It is very generally thus used. The copper is placed at the bottom of the jar ; an insulated copper wire is riveted to it, long enough to pass up through the solutions and connect with a binding screw on the zinc of an adjacent cell, or with the wire which serves to conduct the current to the solution for electrolysis. The bottom of the jar, about the copper, is filled with copper sulphate ; the zinc, a heavy casting with large surface, is suspended a few inches below the top ; and the jar is filled with water some- times acidulated with sulphuric acid. After standing a few hours, the copper sulphate dissolves ; copper is precipitated, and zinc dissolved ; and the jar, in its normal working state, thus contains two solutions ; the heavier, of copper sulphate, below, and the lighter, of zinc sulphate, above. The porous cup of the Daniell cell is thus dispensed with, and the zinc does not require amalgamation. surface of which a little hydrochloric acid has been poured. The amalga- mated cylinder is then placed in a vessel of water to remove the hydrochloric acid, and allow the excess of mercury to drop off. GRAVITY CELL. The cut (Fig. 6) shows one of the simplest gravity cells, having the zinc in the so-called " crow-foot " shape, hanging directly on the edge of the jar, and furnished with a binding- screw. FIG. 6. The outfit of the chemical laboratory of the Pennsylvania State College, while under the translator's charge, was found convenient, and sufficient for the needs of an ordinary labora- tory for instruction. Some twenty "crow-foot" gravity cells were kept in working condition, and eight Grove cells could be set up if needed for a strong current. Four sets of con- necting-wires were run from the battery-room to the labora- tory desk set apart for electrolytic work, each set being so arranged with binding-screws as to be quickly connected with any desired number of cells. (See under " Secondary Bat- teries," p. 47.) Trans.] 14 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. GROVE CELL. The positive pole is a sheet of platinum foil of the form shown in Fig. 7 ; this is placed in a porous cup filled with nitric acid. The negative pole is a cylinder of amalgamated zinc placed in a glass jar containing dilute sulphuric acid (1 : 20). Fig. 8 shows the arrangement of the cell. FIG. 7. FIG. 8. The cells used by the author have the following dimen- sions : Jar . . . . Porous cup . Zinc cylinder Platinum foil 13.5 cm. high, 10.0 cm. diameter. 13.0 " " 5.2 " 14.0 " " 8.5 " 11.0 " " BUNSEN CELL. 15 BUNSEN CELL. In the Bunsen cell, the platinum is replaced by a prism of retort carbon (Fig. 9) standing in a porous cup filled with nitric acid. The negative electrode, as in the Grove cell, is a cylinder of amal- gamated zinc placed in a glass jar filled with dilute sulphuric acid (1 : 20). The screw-clamp shown in Fig. 10 is often used to fasten a metallic connec- tion to the carbon prism. It has, however, the dis- advantage that the clamp is quickly oxidized by the decomposition products of the FIG. 9. FIG. 10. FIG. 11. FIG. 12. 16 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. nitric acid, and the contact thus broken. It is better, there- fore, to insert in the carbon a metallic socket (Fig. 11), the stem of which is closely covered with platinum foil. Fig. 12 shows the Bunsen cell in its most common form. The author uses cells of the following dimensions : Jar 20.0 cm. high, 11.5 cm. diameter. Porous cup . . 19.5 " " 9.0 " " Zinc cylinder . 20.0 " " 9.0 " " Carbon . . . 21.0 " " 5.0 cm. wide, 2.4 cm. thick. THERMO-ELECTRIC PILES. The thermo-electric piles in use are Clamond's and Noe's. (Diamond's pile (shown in Figs. 13 and 14) is composed of a large number of elements, each consisting of a bar of an antimony and zinc alloy, and a strip of tinned sheet-iron ; the iron strips are fastened to the bars as shown in Fig. 15, thus serving to connect the elements. Both the single elements and the superimposed rings of elements are separated by layers of asbestus. The poles of each ring of elements end in binding-screws. The current is produced by heating with illuminating gas which burns from a cylinder of clay or porcelain perforated with numerous openings, which stands in the middle of the pile (Fig. 16, one-third natural size). This tube-burner is cemented in the cylinder with a mixture of powdered asbestus and water-glass, and can be replaced in case of accidental breakage. To keep the flow of gas constant, and prevent excessive heating of the burner, the gas is first passed through a regulator filled with water (r, Fig. 13), the valve of which partly closes the orifice when the pressure rises, and opens it wider when the pressure falls. The water in THERMO-ELECTRIC PILES. 17 FIG. 14. 18 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. the regulator must be replaced as it evaporates. The current attains its full strength when the gas has been burning about one hour ; it then yields 400-450 cc. oxy hydrogen gas per hour. FIG. 15. FIG. 16. After using the pile, care must be taken not to cool the tube-burner too quickly. To this end, the cylinder opening at C (Fig. 14) is first closed with an iron plate d ; after that the cock is closed. The elements of Noe's thermopile are rods of an alloy con- taining 63 per cent antimony and 37 per cent zinc, about 7 mm. in diameter and 27 mm. long (Fig. 17), to each of which is attached a smaller pointed iron rod (e) to conduct the heat to it. The elements are arranged in a circle, on a ring of ebonite, with the iron point resting on a plate which serves to spread the flame of the gas-burner (Fig. 18). The con- THERMO-ELECTRIC PILES. 19 nection of the elements by German-silver strips, nn, etc., is- shown in Fig. 19. The elements are soldered to copper plates set in a circle, which are bent in spiral form, and FIG. 17. FIG. 18. FIG. 19. 20 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. serve to support the elements, and also to cool their outer ends. Fig. 20 gives a general view of the Noe thermopile. If it is only moderately heated, the air cools it sufficiently ; if more strongly heated, it must be placed in a vessel of water. FIG. 20. According to v. Waltenhofen, a pile of 128 elements in 4 groups of ,32 each is equal in electro-motive force to about 2 Daniell cells. The results of the author's investigations are not favor- able to the use- of thermo-electric piles for analytical pur- poses ; they d e tc., volts ; that is, the whole interval of 6 volts is divided into portions of J volt each. If, now, a current, small in proportion to the current passing through the resistance, is taken out between any two binding-screws for an electrolytic determination, the tension between the screws is not materially changed ; the wires carrying this current can be connected with any binding screws without any change in the main current ; moreover, the introduction of a number of such currents does not materially change the tension, and the tension for any given determination can be varied at will without affecting the others, In the apparatus used by the author, Fig. 27 (one- twentieth natural size), the brass wire gauze resistance is divided into 20 equal parts marked 1, 2, 3, etc. As already stated, the machine,, at IjOOO revolutions, has a current- strength of 60 ampdres and a tension of 10 volts. Of the 60 amperes, 40 are conducted through the resistance, so that 20 remain for electrolytic determinations. The difference of tension between two adjacent binding- screws is J i volt. The tension, that is, at the screw marked 19, is volt, at 18 = 1, at 17 = 1J, at 16 = 2, at = 10 volts. The current from the machine enters by a heavy copper conductor at the screw marked 0, and passes out at that marked 20. On the board BBBB are fastened 6 T-shaped galvanized- iron strips, S x , S 2 , S 3 , S 4 , S 6 , S 6 , six resistances of 0.1 ohm each, W u W 2 , W 3 , W 4 , W 5 , W 6 (to allow the strength of 82 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. current in single experiments to be measured), and the brass< strip M 2 . 8j is connected b}^ a wire with W v S 2 with W 2 , S 3 with W 8 , S 4 with W 4 , S 5 with W 6 , and S 6 with W 6 . The iron strips may be connected with the binding-screws 1, 2, 3, etc.,, by means of wires and the brass screws K v K 2 , etc. If the apparatus is used as shown in the cut, and 1, 2, or 3 is connected with S x , 4, 5, or 6 with S 2 , 7, 8, or 9 with S 8 , 10, 11, or 12 with S 4 , 13, 14, 15, or 16 with S 6 , and one of the others with S 6 , the strongest current is at W v and the weakest at W 6 . Any strip may, of course, be connected with any binding-screw. In performing electrolysis, the solutions to be acted on are placed in connection with a negative pole n v n 2 , or w 3 , etc. (on the resistances W v W 2 , or W 3 , etc.), and a positive pole p v p%, or p%, etc., on the brass strip M 2 , the connections being made according to the strength of current desired. Moreover, as shown by the examples given later, several determinations requiring the same strength of current may be connected with any pair of poles, n^ and p v n% and jt? 2 , etc.. In order to connect more conveniently with the platinum dishes containing the solutions for electrolysis, n^ and p v for instance, may be connected with a brass strip Z (the con- nection with W]_ only is shown in the cut), to which are attached a number of binding-screws, z v 2 , etc. The tension and the strength of the current may be measured at each dish. For example, if the tension at the dish connected with W 2 is to be measured, the plugs from the galvanometer are inserted at > 2 and c 2 ; if they are inserted at a 2 and Z> 2 , the tension in the resistance is meas- ured, which, multiplied by 10, gives, in amperes, the strength of the current acting on the solution connected with W 2 . In order to test the working of the apparatus, the tension ELECTRICAL MACHINES. at the divisions of the wire-gauze resistance was directly measured by a torsion galvanometer, with the following results : Resistance marked Connected with Binding Screw, marked Tension in Volts. W, 1 10.300 w i 2 9.900 W l 3 9.400 W 2 4 8.950 W 2 5 8.300 W 2 6 7.750 W 3 7 7.200 W 3 8 6.650 W 3 9 5.950 W 4 10 5.500 W 4 11 5.050 W 4 12 4.500 W 5 13 4.000 W 5 14 3.450 W 5 15 2.850 W 5 16 2.300 W 6 17 1.700 W 6 18 1.100 W 6 19 0.560 W 6 20 0.007 For the measurement of the strength of the current at the screws 1 to 20, a cell was iised which had a copper elec- 34 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. trode,* and contained 150 cc. of a 15 per cent solution of copper sulphate , this cell was connected to the resistance W 6 (binding-screws, n e and p Q ). The screws 1 to 20 were then successively connected with the bar S 6 , and the deviation of the galvanometer read, the plugs connecting it being placed in a Q and 5 6 . After this reading, the tension in the cell was read, for each screw connection, by placing the plugs in b Q and C Q . In order to control the rate of the machine during the experiment, p l and n l on the resistance W 1 were connected through a rheostat ; and the tension at the binding-screw 1 (connected with Sj) was determined by a second torsion galvanometer, the plugs from which were inserted at b 1 and c r The results of these experiments are given in the following table in the columns included under I. A second series of experiments was conducted to deter- mine the strength of the current by the quantity of copper precipitated. Six platinum dishes, as nearly alike as possible, were filled* with 150 cc. each of a 15 per cent solution of copper sulphate, supplied with copper eiectrodes (see note below), and different quantities of copper precipitated in the same time. These experiments were conducted in three series, as follows : Series 1. I., IV., VIII, XII, XVI, XIX. Series 2. II, V, IX, XIII, XVII, XX. Series 3. Ill, VI, X, XI, XIV, XVIII. * The cell consisted of a platinum dish, and the positive electrode was a round piece of sheet-copper (of the form of the platinum electrode shown in Fig. 37), 6 cm. in diameter and 2 mm. thick. The electrodes were 2.5 cm. apart. ELECTRICAL MACHINES. 35 Of the columns included under II., A gives the strength of the current as determined from the precipitated copper ; B, the results, in a few cases, of the measurement of the strength of current by a torsion galvanometer ; and C, the tension measured at the same time with the torsion galvanometer. Binding Screw. I. II. Amperes. Yolts > Volts, Ex P eri - Machine, ment. A, Amperes. B, Am- peres. c, Volts. I. II. 18.018 15.352 7.900 7.400 10.90 9.90 15.97 14.04 - 9.200 9.000 III. IV. V. 13.231 11.615 10.302 7.100 6.650 6.350 10.10 10.10 10.30 10.86 8.87 8.00 10.800 7.900 7.400 7.100 VI. 9.595 6.010 10.40 6.04 - 5.500 VII. 8.383 5.710 10.50 - - - VIII. 6.565 5.300 10.60 4.97 - 5.000 IX. 5.757 5.100 10.60 4.21 3.800 4.500 X. 4.747 4.700 11.10 4.03 - 3.800 XI. 4.040 4.250 10.90 3.75 3.700 3.100 XII. 3.838 3.800 11.00 3.54 - 2.900 XIII. 3.535 3.400 10.90 3.47 " - 2.500 XIV. 3.030 2.850 10.90 8.09 2.700 .2.800_ XV. 2.520 2.400 11.05 - - - XVI. 2.120 1.900 11.00 1.85 - 1.200 XVII. 1.560 1.500 11.00 1.35 - 1.050 XVIII. 0.759 0.890 10.90 0.76 0.605 0.600 XIX. 0.396 0.290 11.00 0.54 - 0.360 XX. 0.000 0.007 11.10 0.00 - 0.007 36 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. The following sixteen experiments were made simulta- neously under the same conditions as before. The numbers- in column A express the quantities of copper precipitated in 6.5 minutes ; those under B, the tensions measured with the torsion galvanometer. A, Copper. B, Volts. ( 0.7616 gm., Binding screw I. to W 1 . . \ 0.7415 " 7.10 1 0.8286 " J f 0.6021 " 1 Screw IV. to W 2 .... \ 0.5716 " 5.30 [ 0.4788 " J 0.4155 " 1 Screw VIII. to W 8 . . . . 0.3510 " 3.30 0.3535 " J ' C.2648 " 1 Screw XII. to W 4 .... 0.2963 " 1.80 0.2652 " J Screw XVI. to W 6 . . . . j 0.1435 " | I 0.1470 " J 0.90 Screw XIX. to W 6 . . . . i ( 0.0363 " | 1 0.0260 " J 0.23 In order to reach a conclusion as to the value of the apparatus for the purposes of quantitative analysis, twelve determinations were carried on simultaneously, at the author's request, by Dr. Robert Ludwig, formerly assistant in the In- organic Laboratory. The solutions used for these experi- ments were of iron, cobalt, tin, antimony, and copper, metals SECONDARY BATTERIES. 37 which, as will be shown later, require currents of widely different strengths for their separation. The results of one series of these experiments are subjoined. Taken. Found. I. 0.3546 gm. Fe 2 O 3 0.2479 gm. Fe = 0.3541 gm. Fe 2 O 3 II. 0.3836 " Fe 2 3 0.2691 " Fe = 0.3844 " Fe 2 O 3 III. 0.2624 " Co 0.2619 " Co IV. 0.2234 " Co 0.2231 " Co Y. 0.1145 " Sn 0.1142 " Sn VI. 0.2290 " Sa 0.2290 " Sn VII. 0.2025 " Sb 2 S 3 0.1444 " Sb = 0.2019 " Sb 2 S 3 VIII. 0.1890 " Sb 2 S 3 0.1348 " Sb = 0.1885 " Sb 2 S 3 IX. 0.1670 " Sb 2 S 3 0.1189 " Sb = .0.1663 " Sb 2 S 3 X. 0.8374 " CuSO 4 0.2133 " Cu = 25.47 % Cu XI. 0.8768 " CuSO 4 0.2225 " Cu = 25.31 % Cu XII. 0.7905 " CuSO 4 i 0.1991 " Cu = 25.29 % Cu i Calculated 25.39 % Cu [SECONDARY BATTERIES. The secondary battery ("accumulator," or, more com- monly, " storage battery ") was already beginning to attract general attention when the earlier edition of this work was issued. To whatever degree it has succeeded or failed in doing the varied work that has been put upon it, it is unques- tionably the most convenient and desirable proximate source of the electric current for analytical purposes. The name " Secondary Battery " accurately describes it, and is therefore decidedly preferable to either " Accumulator" or " Storage Battery." There is no accumulation and no storage of electricity ; while the chemical reaction as a result 88 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. of which the battery gives out an electric current is a second- ary one; that is, it must be preceded by a reverse chemical action, produced by an electric current from some primary source. In the simplest form of secondary battery, introduced by Plante, two or more lead plates, alternately positive and nega- tive, are immersed in dilute sulphuric acid, and a current from a primary battery is passed through these electrodes, at first alternately in opposite directions, until the cell is " formed " that is, until the surface of the lead has become porous r g-iving a large surface for chemical action. When " formed," the charging current is always passed in one direction. To- expedite and intensify the " forming," very many modifica- tions have been made which may be grouped under three heads : (1) extending the surface of the lead plates, (2) pro- ducing oxidation by chemical means, (3) applying mechani- cally an oxide, or mixture of oxides, of lead. The reaction in the secondary battery is essentially the decomposition of water. When the electric current is passed between platinum electrodes through acidulated water, as in the voltameter (p. 58, Fig. 33), the water, as is well known, is- decomposed into oxygen and hydrogen. If now the enor- mously extended lead electrodes are substituted for the plati- num electrodes, the oxygen converts the lead or lead monoxide of the positive electrode into lead peroxide, while the hydro- gen reduces any lead oxide on the negative electrode to metallic lead, and becomes occluded on the surface of tin's lend. (Accordingly, the abundant disengagement of gas- bubbles marks the completion of the charging of a secondary battery.) The battery, thus charged, is chemically in the con- dition of a primary battery that is, there is little or no chem- ical action until the opposite electrodes are put in metallic SECONDARY BATTERIES. 39 connection, when chemical action begins, with the production of an electric current. In the discharge, substantially the reverse reaction takes place. The lead peroxide of the positive electrode is reduced by the occluded hydrogen and by the hydrogen of newly de- composed water, the oxygen of which brings more or less com- pletely to the state of monoxide the surface of the negative electrode. Secondary reactions, between lead or its oxides and the acid used, occur to a very considerable extent, so that the whole subject of the reactions in the secondary battery is complex ; the outline above gives, however, the fundamental reaction, and is sufficient for our purpose. Many forms have been given the secondary battery, and, indeed, some successful batteries use other metals than lead. Any of the batteries to be found in the market will give good results. Fig. 28 is a cut of a form that has been very generally adopted for use with the phonograph. It shows the gen- eral appearance both of the com- plete cell and of the assemblage of alternate positive and negative electrodes as immersed in the acid. Moreover, a secondary battery that will give entirely satisfactory results for analytical purposes can readily be made in any laboratory, and formed by the current, substantially on the Plante system. In undertaking this, it is well to remember that complete isolation of the electrodes from each other is essential, and FIG. 28. 40 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. that extended surface is favorable to rapid and compete " forming." For a full and recent com pen d of facts in regard to the secondary battery see k< The Voltaic Cell." by Park Benjamin, pp. 415-508. Trans.'] THE USB OF ACCUMULATORS AND THEIR ADVANTAGES OVER DYNAMOS AND GALVANIC BATTERIES. The apparatus described on pp. 25-32 for carrying on simultaneously a large number of analyses by a dynamo cur- rent, which was used for years only in the Aachen laboratory, has the disadvantage that at night, and on days when the stenm engine is not in use, electrolytic work must cease. This fact led the author to experiment with accumulators, using the apparatus constructed by Professors Farbaky and Scheneck,* of Schemuitz, Hungary. These gentlemen had the kindness to place two accumulators at the author's service for testing, and he made his first experiments in 1888, work- ing with R. Schelle, at that time professor in the Schemnitz Royal School of Mines. These accumulators have 6 negative and 5 positive lead-plate electrodes, each 6 mm. thick. The weight of the electrodes is 15.5 kg., the volume of the 33$ sulphuric acid 3.5 1., the total weight of each, cell 35 kg. The active surface of the electrodes is 3133 sq. cm., so that the internal resistance is very low, measuring between 0.0166 and 0.0 17 ohm. The accumulators can be charged at a 20 to 25 ampere rate, and yield in discharge at 23, 30, 40 and 60 ampere rates respectively 150, 148, 140 and 125 ampere-hours, with a fall of riot over 10% in the voltage. If the discharge * Cf. Ueber die elektrische Akkumulatoren voii Farbaky und Sche- neck (Dingier, Polyt. Jour. 257, 357); also Bericht liber die Akkumulaloren vou Farbaky und Scbeneck von A. v. Waltenbofeu, Zeit. f. Elektro- technik, 1886. SECONDARY BATTERIES. 41 is lighter, and the fall in electro-motive force less than for lighting purposes, as in electrolytic analyses, an accumulator may yield over 250 ampere-hours. Two such accumulators were fully charged, until OH gas was obviously disengaged, by a current of 20 to 25 amperes from the dynamo, through the brass-wire-gauze resistance, Fig. 27. The current was measured by a Kohlrausch galva- nometer, made by Hartmann & Braun, Bockenheim, Frank- furt a. M., the scale of which read from to 60 amperes. A second Kohlrausch amperemeter, divided from to 15 am- peres, was used to measure the current taken from the accu- mulators for the analyses. A Siemens torsion galvanometer showed a tension, for each charged accumulator, of 2.05 volts; when the two accumula- tors were connected for tension, they yielded 40-50 cc. oxy- hydrogen gas (see p. 57) per minute. By the use of these two accumulators, four to eight analyses were carried on simultaneously, and the accumulators kept in constant use day and night, except for the short intervals needed to change the solutions for analysis. The results of analyses extending over a period of six days are subjoined. FIRST DAY. Tension 2.55 volts. (Voltameter 48 cc. OH gas.) Determination of Copper from Nitric-acid Solution. Taken CuSO 4 ,5H 2 O. Found Cn. 4.0140 gm. 1.0170 gm. = 25.33$ 4.1376 ^ 1.0480 " = 25.33 2.2340 0.5661 " = 25.34 2.3575 0.5978 " = 25.35 42 QUANTITATIVE ANALYSIS BY ELECTROLYSIS Tin from the Acid Ammonium Double Oxalate.* Taken Si]Cl 4 2NH 4 Cl. l.S450gm. 2.0210 " Found Sn. 0.5964 gra. = 32.33 0.6548 >< = 32.39 Antimony from Solution in Sodium Sulphide. f Taken Sb 2 S 3 . Found Sb. n 0.2404 gm. 0.1720 gm. = 71.50 0.2551 " 0.1827 = n.eo SECOND DAY. Tension 1.95 volts. (Voltameter = 42 cc. OH gas.) 2.0490 gm. NiSO 4 +(NH 4 ) 2 SO 4 ,6H 3 O gave 0.3053 gm. Ni = 14.90# ( ( 3.0180 " 3.3400 " CoSO 4 -[-K 2 SO 4 ,6H 2 O " 0.3000 " 0.3440 " =14.91 ( 3.1200 " " 0.3120 " " =1471 ] 89?0 " FeSO 4 +(NH 4 ) 2 SO 4 ,6H 2 O " 0.2697 " Fe = 14.25| ( 3.1210 " " 0.3027 " " = 14.25 .0 ' CuS0 4 ,5H 2 O " 0.2533 " Cu = 25.33[ .0 " 0.2533 " " =25.33 .0 " 2534 " " =25.34 .0 " 0.2537 " " =2537 .9210 " SnCl 4 + 2NH 4 Cl " 0.6219 " Sn = 32.37**- < 3.1320 " " 0.6900 " " =32.36 * Classen's method : see Tiu, later. f " " " Antimony, later. \ " " " Nickel " " Cobalt I " " " Iron T From the acid double oxalate, Classen's method. ** From the acid ammonium double oxalate. SECONDARY BATTERIES. THIRD DAY. Tension 1.95 volts. (Voltameter = 40 cc. OH gas. (Six simultaneous analyses.) 1.0050 gm. CuSO 4 ,5H 2 O 1.0170 " 1.0006 " 1.0013 " 1.5680 " SnCl 4 +2NH 4 C1 2.4520 " gave 0.2550gra. Cu = 25.37 " 0.2580 " " =25.36 " 0.2539 " " =25.37 " 0.2540 " " =25.37 " 0.5070 " Sn = 32.34 " 0.7946 " " =32.40 FOURTH DAY. Tension 1.95 volts. (Voltameter = 40 cc. OH gas.) gave 0.2532 gm. Cu = 25.32# 1.0 gm. CuSO 4 ,5H 2 O 1.0 < 1.0 " < 1.0 1.0 ii " 1.0 K ii 1.0 " < 1.0 < " 2.20 ii NiS0 4 +(NH 4 ) 2 S0 4 ,6H 2 2.45 " 2.1340 " CoS0 4 4- K 2 S0 4 ,6H 2 O 2.4350 ii < t 0.2535 " " =25.35 0.2532 " " = 25.32 0.2536 " " =25.36 02535 " " =25.35 02538 " " =25.38 0.2539 " " =25.39 0.2537 " " =25.37 0.3277 " Ni = 1489 0.3650 " " =14.89 0.3148 " Co = 14.75 0.3587 11 " =i4.?a FIFTH DAY. Tension 1.95 volts. (Voltameter = 40 cc. OH gas.) 1.0 1.0 2.4120 2.2130 gm. CuSO 4 ,5H 2 O " FeS0 4 -f (NH 4 ) 2 S0 4 ,6H 2 O gave 0.2537 gra. Cu = 25.37& " 0.2537 " " =2587 " 0.3438 " Fe = 14.25 " 0.3156 " " =14.26- a u 0.2534 " u 25 .34 u 0.2536 " u r= 25 .36 a 0.2533 u rrz 25 .33 a u 0.2537 (I a = 25 .37 " u 0.2534 u tt = 25 .34 u 0.2536 u " = 25 .36 U a 0.2535 a u 25 .35 44 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. SIXTH DAY. Tension 1.92 volts. (Voltameter = 39 cc. OH gas.) (Eight simultaneous copper determinations.} 1.0 gm. CuSO 4 ,5H 2 O gave 0.2533 gm. Cu = 25. 1.0 " 1.0 " 1.0 1.0 " 1.0 " 1.0 " 1.0 Fifty determinations, it is seen, were made in six days. At the end of the sixth day the tension had fallen to 1.85 volts and the volume of OH gas to 37 cc.; the accumulators were therefore fully charged by a 10-ampere current, receiving an addition of 54 ampere-hours. Since the total capacity of such an accu- mulator is over 250 ampere-hours, it is safe to assume that 60 to TO determinations may be made from one charge; and the experience of years confirms this assumption. To ascertain whether accumulators in use still retain electric energy, their tension may be measured, or the specific gravity of the sulphuric acid in the accumulator may be determined, as this is higher when the accumulator is charged. Still another advantage in the use of accumulators is found in the beauty of the precipitated metal, resulting from the constancy of the current, which much exceeds that from a galvanic battery or a dynamo. In the Aachen laboratory two pairs of accumulators have been constantly used since 1888, without need of repair. Two accumulators are in use in the analytical laboratory, for which they have proved entirely sufficient, and two in the author's private laboratory. The SECONDARY BATTERIES. 45 current from the dynamo is used only to charge the accumu- lators. F. Riidorff * has recently advocated the use of Mei- dinger cells in place of the " much-extolled dynamo, with or without accumulators," f on the ground that the outfit of dynamo and accumulators is attainable only by laboratories commanding abundant supplies and the use of power, while even then the necessary attention to the source of power and the dynamo is irksome. As to the use of Meidinger cells, those who are familiar with the earlier editions of this work and the contributions of the author to the "Berichte" are aware that he early recommended the use of Meidinger bat- teries for the determination of many metals, made investiga- tions as to the strength of current yielded by them, and used them for years in the laboratory under his charge. These cells are entirely satisfactory when they are not needed daily for a long time (e.g., in small commercial laboratories), and when there are not, as a rule, a large number of determinations to be simultaneously made. This has been established by the experience of many former pupils of the author who have worked or still work with Meidinger cells. The use of Meid- inger cells is, however, not to be recommended in labora- tories for instruction, since complete precipitation of the * Zeit. fiir angewandte Chemie, 1892, S. 3. f It is admitted that the author first suggested the use of the dynamo and accumulator. His introduction of the dynamo into the chemical lab- oratory has been recognized by highest authority as a noteworthy service (Kiliani : Berg- u. Htittenm. Zeit., March, 1886). The author is unaware that he has "extolled " these sources of the galvanic current ; his publica- tions have simply stated their advantages over galvanic batteries, giving results of experiment. Dynamos with or without accumulators have been introduced not only in laboratories for instruction, but also in laboratories of factories and smelting-works. [Here followed a list of laboratories, which it is not necessary to reproduce. It includes only Cornell Univer- sity in this country. Trans.} 46 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. metal requires a long time (12 to 14 hours), arid the pupils can- not watch the reaction. For this reason it is not desirable in a laboratory for instruction to leave solutions for analysis to the action of the current during the night. Moreover, while the separation of certain metals (e.g., antimony from tin, and arsenic, cadmium, bismuth, or mercury from other metals) is possible with Meidinger batteries, the separation of many other metals is attended with great difficulties. The possibility of conducting analyses during the night is a recognized advantage of electrolysis; cases often occur, how- ever, in technical work in which rapidity is equally as impor- tant as accuracy, both for control of manufacturing processes, and in the purchase of foreign ores. Now Meidinger cells are totally unsuited to rapid analytic work. According to the manager of the largest smelting works in the Rhine provinces and Westphalia, it is the custom, especially in foreign markets, to send samples of ores that are to be sold at auction to such smelting-works as use them. A rapid and exact determination of the metal in the ore is of the highest importance, inasmuch as the auction sales take place without delay, and the value of the ores to the smelting- works must be promptly given to the seller. Thus it has been demonstrated to the author that his methods and apparatus first made it possible to secure a rapid and exact determination of metals in such cases, and that the application of electrolytic methods is adapted to greatly facili- tate dealing in antimony, lead, silver, gold, and other ores. In the Aachen laboratory, in which determinations were formerly made with Meidinger cells, there have been made for some years past, by the use of two accumulators, some 2500 electrolytic determinations yearly, on the average. Riidorff s contention that only such laboratories as have large resources and power at command can afford to use tiie dynamo with or without acc'nmulators, is not well founded. \ SECONDARY BATTERIES. 47 A small dynamo and two accumulators can be obtained for a few hundred marks, and as to the question of power, Kudorff was considering only laboratories for instruction. In indus- trial establishments (as well as in the Aachen laboratory) the dynamo is run by a belt from shafting, and the portable accumulators are charged without difficulty. The larger factories usually have their own electric-lighting plants, the generator of which can conveniently be used in the daytime to charge accumulators. This method is followed in a smelt, ing establishment near the Aachen laboratory. In cities Avhich have electric-lighting plants the use of the current can be obtained for charging accumulators, reducing the original outlay to the purchase of two accumulators. Physicians often use accumulators which are charged by the local electric-light- ing company. In any case, accumulators may be charged from Bunsen cells; there is needed only an amperemeter and a resistance so as to insure the maximum current needed for the accumulators in use. [The rapidity with which Bunsen cells " run down " makes it inconvenient to use them, as suggested in the last paragraph, to charge secondary batteries. The durability and uniformity under constant discharge of the copper-sulphate cell suggests the use for this purpose of a battery of special large cells of a simple " gravity " type, such as could be put up in any labora- tory, using, e.g , as containers, the large pails or small tubs, of some kind of papier-mache or compressed wood fibre, which are to be found in any grocery, after coating them heavily on the inside with asphalt paint. A series of careful experi- ments recently conducted with the knowledge of the translator on the charging of a secondary battery with a gravity battery of such type gave, however, very indifferent results. The translator believes the most economical, convenient, and permanent source of current for electrolytic work, in 48 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. an isolated laboratory, whether for research, for instruction,, or for technical purposes, to be some form of the Paget thermopile described on p. 21 ff., with a secondary battery of two cells. Such a plant, as has been indicated, can be, if de- sirable or necessary, entirely home-made. Trans .] APPARATUS FOR REDUCING THE STRENGTH OF THE CURRENT. RESISTANCES. In order to obtain currents of any desired strength from a galvanic battery or an accumulator, it is often necessary to place in the current a resistance. As appears hereafter, currents of very different strengths are required to separate different metals from their solutions in a manner adapted to quantitative? analysis. For instance, the determination of iron requires a: current of 0.5 to 1 ampere ; of tin, one of about ampere; of antimony, one of about 0.15 ampere. Constant currents of about 0.15 ampere may be obtained by the use of five or six. Meidinger cells, while it is difficult with this type of cells to> obtain or to maintain constant currents of 0.5, 1, or 1.5 amperes.. To carry on a number of antimony or other determinations by the use of a Meidinger battery requires a large number of cells, while any desired determination or separation may be carried on by the use of Bunsen cells or accumulators with suitable resistances. The author commonly uses, for the reduction of the strength of a current, plug rheostats. As ordinarily con- structed, these are ill adapted to laboratory use, for the plugs are quickly attacked by acid vapors from the cells, or the vapors of the laboratory, and the resistance introduced is thus changed. The ordinary apparatus has also the fault REDUCING THE STRENGTH OF THE CURRENT. 49 that the plugs are liable to become loose. Both difficulties are met by the use of mercury contacts, instead of plugs, to FIG. 29. FIG. 30. 50 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. connect the metallic plates. Fig. 29 shows the arrangement of such a rheostat.* By inserting the contact bars CC in the mercury cups or removing them, any resistance from 0.5 to 80 ohms may be inserted by intervals of 0.5 ohms. The following results of experiment show the action of this rheostat. A current from three Bunsen cells yielding in the voltameter 28 cc. oxyhydrogen gas per minute was reduced as follows : Ohms inserted. CC. Oxyhydrogen Gas per Minute. Ohms inserted. CC. Oxyhydrogen Gas per Minute. 0.5 16.00 15.0 2.20 1.0 12.50 20.0 1.30 1.5 9.75 30.0 1.10 2.0 7.00 40.0 0.80 3.0 6.00 50.0 0.70 4.0 5.00 60.0 0.60 5.0 4.90 70.0 0.50 7.5 4.00 80.0 0.45 10.0 3.50 A current of 16 cc. oxyhydrogen gas pel* minute (yielded by two Bunsen cells) was reduced by 40 ohms to 0.4, and by 80 ohms to 0.15 cc. oxyhydrogen gas. More recently the author has used the simplified form of rheostat shown in Fig. 30, in which brass plates are dispensed with, and the contact with the German-silver coils is made directly by mercury. The following results of experiment show how constant is the current from Bunsen cells when a rheostat is used. In the separation of antimony from tin, the current from two * This rheostat is made, at the author's suggestion, by Fraas Brothers in Wunsiedel. REDUCING THE STRENGTH OF THE CURRENT. 51 Bunsen cells was reduced to 0.6 and 2 cc. oxyhydrogen gas per minute. Columns A and B give the strength of the two Bunsen elements ; columns C and D, that obtained by use of the rheostat. A and C were measured before the experiments; B and D, after them (lapse of time, 14 hours). A. B. C. D. CC. OH Gas. CC. OH Gas. CC OH Gas. CC. OH Gas. 17 16.0 0.6 0.3 24 19.0 0.6 0.4 18 11.5 0.6 0.3 17 15.5 0.6 0.4 With a view to economy in battery power, and especially in rheostats, the author has constructed a simple apparatus FIG. 31. {Fig. 31) which allows a number of determinations to be carried on with the same battery. 52 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. The current from the battery enters at , circulates through the German-silver resistance !N", and returns to the battery through b. In making electrolytic determinations the plati- num dishes serving as negative electrodes may be attached to any one of the binding-screws 1-20, while the platinum foil serving as positive electrode is attached to the binding-screw marked with the + sign. The apparatus, therefore, is suited to carry on eight different determinations simultaneously. Its value for analytical purposes is shown by the following experiments. To determine directly the current strength at the binding-screws 1-20, 150 cc. of a 15$ copper sulphate solution was placed in each of 6 platinum dishes of equal size, copper electrodes* were used, and the current passed for 7 minutes in each case. The current was produced by a battery of 5 Bun sen cells, which gave 56 cc. OH gas in the first and second experiments,, and 68 in the third. FIEST EXPERIMENT. Gm. Cu. Amperes. Binding-screw 1 0.5064 = 3.75 " 2 .*.. 0.3507 = 2.617 " 3 . - . . . 0.2873 = 2.085 " 4 .* ..... 0.2358 = 1.711 " 5 . > . . . 0.1857 = 1.348 " 6 0.1453 = 1.054 " 7 0.1341 = 0.973 " 8 0.1128 = 0.818 * 6 cm. in diameter, 2 mm. thick. The diameter of the platinum dishes was 9 cm., the distance of the electrodes from each other 2.5 cm. REDUCING THE STRENGTH OF THE CURRENT. 53 SECOND EXPERIMENT. Gin. Cu. Amperes. Binding-screw 7 . . . . . 0.2213 = 1.606 " 8 0.1622 = 1.177 9 ..... 0.1356 = 0.984 " 10 0.1083 = 0.786 " 11 . , . , . 0.0846 = 0.614 12 0.0744 = 0.576 13 0.0506 = 0.367 14 0.0410 = 0.225 THIED EXPERIMENT. Gm. Cu. Amperes. Binding-screw 13 ..... 0.1983 = 1.446 14 0.1304 = 0.946 " 15 0.1276 = 0.926 " 16 0.0855 = 0.620 " 17 ..... 0.0605 = 0.439 18 0.0385 = 0.280 " 19 0.0314 = 0.227 " 20 0.0136 = 0.098 From a number of quantitative determinations, which, were made by Norrenburg with this apparatus, the following are selected : SERIES I. i The apparatus was attached to a battery of 5 Bunsen cells, yielding 62 cc. OH gas, and eight iron determinations made simultaneously. The precipitation was complete in 6 hours. QUANTITATIVE ANALYSIS BY ELECTROLYSIS. Taken FeSO 4 , 2(NH 4 ) 2 SO 4 .6H 2 O. 1.2918 gm. 1.4360 " 1.1926 " 1.1964 1.2945 " 1.3218 1.2931 " 1.3255 " Found Fe. 0.1846 gm. = 14.30$ ] 0.2059 " = 14.33 I 0.1708 =14.32 j 0.1700 =14.30 0.1851 =14.30 0.1892 " = 14.31 0.1854 " = 14.34 0.1895 " =14.30 Binding Cc. Screw. OH Gas, 24.0 25.0 24.0 16.8 16.6 17.2 13.2 At the close of the experiment the battery (without resist- ance) still yielded 50 cc. OH gas. SERIES II. Three nickel and five copper determinations were con- ducted simultaneously. The current from 5 Bunsen cells- yielded 65 cc. OH gas. Taken Nickel Ammon. Sulph. 1.2848 gm. 1.4341 1.2008 " Found Nickel. 0.1963 gm. = 15. 0.2201 " = 15.35 0.1842 = 15.34 Binding Cc. Screw. OH Gas. Copper Sulphate. 1.1531 gm. 0.9787 " 1.0092 0.9938 " 1.0088 " Copper. 0.2910 gm. = 25. 0.2476 " = 25.30 0.2556 " = 25.32 0.2515 " =25.30 0.2550 =25.27 DEDUCING THE STRENGTH OF THE CURRENT. SERIES III. This established the applicability of the process to the simultaneous determination of nickel, antimony, and copper. The number of analyses again was eight. The battery, of 5 Bunsen cells, yielded 65 cc. OH gas per minute. Taken Nickel Ammonium Sulphate. 1.3022 gm. 1.1520 " 1.4391 " Antimony Tersulphide. 0.1609 gm. 0.1691 " 0.1626 Copper Sulphate. 0.2527 gm. 0.2550 The current strength of the battery at the close of the last two series was about half that at the beginning. These experiments show plainly the practical advantage of the new rheostat. To perform eight iron determinations (Series I.) simultaneously without this rheostat would require 8 ordinary rheostats and at least 16 cells. For three nickel and five copper determinations would be needed 16 Bunsen cells and 8 rheostats, or 6 cells, 3 rheostats, and 5 Meidinger batteries of 3 or 4 cells each, the latter for the copper deter- minations. The conditions would be similar with the third series. [Prof. Edgar F. Smith uses a simple apparatus (Fig. 32), the accompanying figure and description of which he kindly permits the translator to copy : "The writer has for some time employed a much simpler Found Binding Cc. Nickel. Screw. OH Gas. 15.30$) ( 21.0 15.27 L 3 -j 22.0 15.32 j ( 22.0 Antimony. 71.44$) ( 1.0 71.47 I 9 \ 1.0 71.49 ) ( 1.0 Copper. 25.30 | 25.30 J 7 ) 3.6 ) 3.6 56 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. current-reducer, which has the advantage of cheapness and ready construction to recommend it. It consists of a light wooden parallelogram, about 6 feet in length. Extending from end to end, on both sides, is a light iron wire, measuring FIG. 32. in all about 500 feet. With the binding-posts at a and 5, and a simple clamp, it is possible to throw in almost any resistance that may be required. It answers all practical purposes," " Electro-chemical Analysis," p. 29. TransJ] MEASUREMENT OF STRENGTH OF THE CURRENT. 57 MEASUREMENT OF THE STRENGTH OF THE CURRENT. The unit of strength is the ampdre, that of electro-motive force is the volt, that of resistance is the ohm. In a current, the resistance of which is 1 ohm, a current of 1 ampdre gives the force of 1 volt. 1 ohm is equal to 1.065 Siemens units, ;S. U. (a column of mercury 1 m. long and 1 sq. mm. in area of .section, at 0) ; 1 volt = 0.95 Daniell (D. = the electro-motive force' of a Daniell cell) ; 1 ampere = * ; 1 ampere = 1 o. LJ. 10.436 cc. oxy hydrogen gas at and 760 mm. pressure = 19.69 mg. copper = 67.1 mg. silver in 1 minute. The strength of current in cc. of oxyhydrogen gas, multiplied by 0.0958, gives the strength in amperes. For the measurement of the strength of the current, either the chemical or the magnetic action of the current is used. In the former case, the strength of the current is determined by the galvanic decomposition of water, and measurement of the volume of oxyhydrogen gas produced. A special appar- atus, the voltameter, is used for this purpose ; its construction is shown in Fig. 33. The cylindrical vessel g is partly filled with pure dilute 33 per cent sulphuric acid. The platinum wires d and d' welded to the platiiium strips p and p^* are fused into the walls of the vessel g. This latter stands in a large cylinder C of water which serves to cool it. The platinum wires end in the screws s and ', which are con- nected with the battery. The oxyhydrogen gas, as it is formed, passes through the tube r, which contains a little water, and then is collected in the measuring-tube R, which is graduated into - cc., and filled with water. To measure an electric current with the voltameter, the water over which * In the apparatus used by the author, the platinum electrodes are 31 X 13 mm., and are distant from each other 20 mm. 58 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. the gas is collected is first saturated with oxy hydrogen gas r and then, by the use of a watch with second-hand, the volume of gas is observed, which the current yields in a FIG. 33 minute, or, if the current is weak, in a longer time. To compare observations, the volume should be reduced to and 760 mm. pressure. v =. observed volume of oxyhydrogen gas. v l = normal volume (at and 760 mm.). t =z observed temperature. h = pressure reckoned in mm. of mercury. v h Vl " 1 + 0.00367* ' 760' MEASUREMENT OF STRENGTH OF THE CURRENT. 50 Let I indicate the height of the column of liquid, s the density of the liquid, and I the barometric height ; then The foregoing long-used form of voltameter is inconvenient in use, and, moreover, admits of no allowance for pressure in reading off the volume of gas. Unless -the measuring-tube is completely filled with water before each experiment, the re- sults from the same current will differ according to the quan- tity of OH gas which the eu- diometer already contains. J. Walter has constructed an ap- paratus for electrolytic analysis which is much more convenient than the preceding, and is also free from the fault above men- tioned. Walter's form of the apparatus cannot, however, be used in connection with this work, as the author's observa- tions are all based on the use of a voltameter with platinum electrodes 31 mm. long, 13 rnm. wide, and 20 mm. apart. G. Neumann, former assistant in the Aachen laboratory, has therefore modified Walter's ap- paratus to the form shown in Fig. 34. This voltameter con- sists of a tube A containing 40 cc. and graduated into -fa cc., the graduation beginning below * 13.6 = sp. gr. of mercury. FIG. 34. 60 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. the cock a. Below the graduation the tube is so widened that the two platinum electrodes, 31 ram. long and 13 rnm. wide, can readily be placed at the required distance from each other. Below the cylindrical vessel c thus formed the tube is again nar- rowed for the attachment of a rubber tube connecting it with the pressure-tube B. The wires leading from the electrodes are fused into the walls of AND ALUMINIUM. The nitric acid solution is electrolyzed as directed on p. 97, and as large a surface as possible furnished to the lead dioxide by making the platinum dish the positive electrode. CADMIUM AND ZINC. The separation depends on the same principle as the separation of copper from other metals (see p. 91). Eliasberg has investigated the subject in the author's laboratory, and finds that the separation is complete and satisfactory when the solution is kept warm throughout, and a current of 0.1- 0.15 cc. oxyhydrogen gas per minute is employed for the separation of the cadmium. 8 to 10 grams potassium oxalate are dissolved in the acid-free solution, it is heated with the addition of 2-3 gms. ammonium oxalate, diluted to about 100 cc., and electrolyzed as directed. The cadmium is com- pletely separated in six to seven hours ; it separates as a partly compact, partly crystalline, deposit on the platinum dish. The zinc is determined in the solution as directed on p. 83. A. Yver * recommends the use of a solution of the ace- * Bull. boc. Chim. de Paris, 34, 18. 136 QUANTITATIVE ANALYSIS BY ELECTROLYSIS rates or sulphates treated with an excess of sodium acetate and a few drops of acetic acid ; the electrolysis to be con- ducted hot, using two Daniell cells. According to Eliasberg's experiments, the action is satisfactory when the current has a strength of 0.5-0.6 cc. oxyhydrogen gas, and the solution is diluted to 80-90 cc. [Smith and Knerr* have also investigated this process, and obtain the best results with a current of not much more than 0.1 cc. oxyhydrogen gas. They also separate cadmium from zinc in a tartrate solution, containing 2 gms. sodium tartrate and excess of tartaric acid, employing a current generating 0.4-0.5 cc. oxyhydrogen gas per minute. The cadmium is deposited rapidly at first, but the last traces are difficult to remove. Duration of process, two and one-half to four and one- half hours. Trans. ~\ In the laboratory of the Technical High School in Munich the following directions are given for Tver's method: To the sulphuric acid solution of the two metals add sodium hydrox- ide solution until a permanent precipitate is obtained, dissolve the precipitate in the smallest possible quantity of dilute sul- phuric acid, dilute the solution to about TO cc. and reduce the cadmium with a current of N. D. 100 0.07 ampere (see p. 63). "When the greater part of the metal is precipitated neutralize the free sulphuric acid with sodium hydroxide, add 3 gm. sodium acetate, heat to about 45, and subject to the action of a current of N. D. 100 = 0.3 ampere. The latter direction as- sumes that the electromotive force is not over 3.6 volts ; if more, it is to be reduced to about 2.4 volts. Edgar F. Smith and Lee K. Frankel recommend the sepa- ration of the two metals by precipitating the cadmium from * Am. Ch. J., 8,210. SEPARATION OF THE METALS. 137 the solution of the double cyanides. The solution is treated with potassium cyanide in such excess that 4.5 gm. are present to 0.4 gin. of the metals, diluted, and a current of 0.3 cc. elec- trolytic gas used for the electrolysis. Precipitation is com- plete in 18 to 24 hours. Zinc is precipitated by this weak current only when all the cyanide is decomposed. It is deter- mined in the cadmium-free solution, as heretofore directed the large amount of alumina which it always contains. If, therefore, commercial sodium sulphide is to be used, it must first be dissolved, and the solution, with exclusion of air, completely saturated with pure hydrogen sulphide gas. It is then filtered from the precipitated impurities, and evaporated in a large platinum or porcelain dish. The further treatment is given in full in the chapter on reagents. It is preferable, however, as the condition of the sodium sulphide solution is of great importance to the success of the process, to prepare the solution as directed in the chapter referred to. The process of separation is as follows : A mixture of the pure sulphides,* or the residue obtained by evaporating a solution of the two metals, is treated in a platinum dish with about 60 cc. of a sodium sulphide solution of sp.gr. 1.22-1.225, and enough concentrated solution of sodium hy- droxide to furnish 1 gm. NaOH. If solution does not take place at once, it is hastened by heating over a low flame, the watch-glass covering the dish is rinsed with 10-15 cc. water, and the solution is allowed to cool thoroughly. It is then submitted to the action of a current yielding 1.5-2 cc. per minute, which may be produced by several Meidinger cells, or by reducing to the required strength the current from Bunsen cells, dynamo, or accumulators. It is best to leave the current at work during the night ; the separation is completed in twelve hours, the antimony appearing as a brilliant coating, adhering closely to the dish. When the action begins, the whole surface of the dish, which is in contact with the solution, becomes quickly covered with a dark coating of antimony which soon takes on a brilliant metallic appearance. * The solution of a mixture of the metallic sulphides and sulphur in sodium sulphide is to be treated like a solution of polysulphides (see p. 108)- 140 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. In the earlier part of the process, the whole solution appears to be filled with small gas bubbles which rise slowly, break at the surface, and cover the watch-glass with minute portions of the solution. In the course of two hours, the disengagement of gas is ended, and the solution is completely clear. To avoid loss, it is best, at this time, repeatedly to wash the under surface of the watch-glass with a drop of water which finally runs down the positive electrode. When the current has been in action for twelve hours, it is stopped, the solution is poured into a second tared platinum dish, and the deposited metal is washed two or three times with about 10 cc. of water. The antimony is treated according to the directions already given, and weighed. As tin cannot be reduced from a sodium sulphide solu- tion, but can be completely precipitated from solution in ammonium sulphide (as stated on p. 110), the sodium sulphide, after the separation of antimony, must be converted into ammonium sulphide according to the directions given on p. 112. If the two metals are to be determined in the yellow solution of polysulphides of the alkalies, the solution is decolorized with ammoniacal hydrogen peroxide (see Anti- mony, p. 112), and evaporated nearly to dryness ; about 60 cc. sodium sulphide solution and the necessary amount of sodium hydroxide are then added, and the process carried on as above directed. ANTIMONY AND ARSENIC. In an alkaline solution, arsenious acid is oxidized to arsenic acid by the galvanic current. If, however, a solution containing antimony and arsenious acid is electrolyzed, a mixture of antimony with arsenic is deposited. The action SEPARATION OF THE METALS. 141 is different if the arsenic is present in the solution as arsenic acid ; in the presence of a free alkali, the antimony alone is deposited from a concentrated sodium sulphide solution. The arsenic, therefore, if present as arsenious acid, must be oxidized to arsenic acid before the metals can be separated. It is heated with concentrated nitric acid or aqua regia, the acid completely removed by evaporation on the water-bath, the residue treated with 50-60 cc. sodium sulphide, sp. gr. 1.22-1. 22& ; a concentrated solution of sodium hydroxide containing about 1 gm. NaOH added, and the solution sub- mitted to the action of a current yielding 1.5-2 cc. oxy- hydrogen gas per minute. The separation is conducted precisely like that of antimony from tin. If antimony and arsenic are to be determined in a solution of polysulphides of the alkalies, the solution is treated as described on p. 108. To determine arsenic, the antimony-free solution is acidified with dilute sulphuric acid, heated in the water-bath to remove hydrogen sulphide, filtered, and the precipitate dissolved in hydrochloric acid with the addition of potassium chlorate. This solution is treated with ammonia in excess, and the arsenic acid precipitated as magnesium ammonium arsenate with magnesia mixture. The precipitate may be dried, at 110, on a weighed filter, and weighed, or converted into magnesium pyro-arsenate by careful ignition in a porcelain crucible. ARSENIC, ANTIMONY, AND TIN. If arsenic is present as arsenic acid, antimony alone is precipitated from a concentrated alkaline solution of the three metals in sodium sulphide ; tin and arsenic remain in solution. The arsenic is converted into arsenic acid, and the antimony precipitated, exactly as heretofore described. 142 QUANTITATIVE ANALYSIS BY ELECTROLYSIS For the separation of tin from arsenic, the solution poured off from the antimony is treated with dilute sulphuric or hydrochloric acid to decompose the sulpho-salts, the mixture of arsenic and tin sulphides and sulphur is filtered off, and oxidized with hydrochloric acid and potassium chlorate, and the arsenic separated as above described. To determine the tin, the solution freed from arsenic is saturated with hydrogen sulphide, filtered, and the tin sulphide dissolved in ammonium sulphide. The tin is determined electrolytically as directed p. 111. In the analysis of a substance which contains arsenic, antimony, and tin, the arsenic may also be first eliminated according to the method of E. Fischer-Hufschmidt simplified by R. Ludwig and the author,* and antimony and tin sepa- rated in the arsenic-free solution. If the sulphides of the metals are to be separated, they are oxidized with concentrated hydrochloric acid and potas- sium chlorate, and the acid evaporated on the water-bath. The residue is washed with fuming hydrochloric acid into a flask of 500-600 cc. capacity,! treated with 20-25 cc. of a saturated solution of ferrous chloride, or better, with about 25 gm. of ammonium ferrous sulphate [FeSO 4 -f- (NH 4 ) 2 SO 4 + 6H 2 O] and fuming hydrochloric acid added till the volume is 150 to 200 cc. A strong current of hydrochloric acid gas is now passed into the solution, and kept up for at least half an hour after the solution seems fully saturated. Then the solution is reduced to about 50 cc. by distilling off the liquid, without a condenser, in a stream of hydrogen chloride gas. A flask of about 1 litre capacity, containing 400-500 cc. * Ber. d. ch. Ges., 18, 1110. t A convenient apparatus is illustrated in the author's " Handbuch de* Quantitative Analyse," 4th edition, p. 78. SEPARATION OF THE METALS. 143 water, is used as a receiver. If the flask is well cooled during the distillation, not a trace of arsenic passes over into a second receiver, even when as much as 0.5 gm., reckoned as As 2 O 3 , is present. The arsenic in the distillate may either be saturated with sodium carbonate, and titrated with iodine solution, or pre- cipitated as As 2 O 3 with hydrogen sulphide, and determined as such on a weighed filter, or the arsenic calculated from the amount of sulphur in the precipitate. The process, in the latter case, is as follows : The distillate is mixed with twice its volume of water, air expelled by a strong current of carbon dioxide, and the arsenic precipitated by passing in pure hydrogen sulphide gas. The excess of hydrogen sul- phide is removed by passing a strong current of carbon dioxide till lead acetate paper is not colored by the escaping gases. The arsenic sulphide is allowed to subside, and the clear solution siphoned off. The remaining strongly acid solution is saturated with ammonia, which dissolves the arsenic sulphide ; the solution is then boiled with an excess of hydrogen peroxide free from sulphuric acid. The solution is then acidified with hydrochloric acid, and the sulphuric acid produced by the action of the hydrogen peroxide deter- mined as barium sulphate in the usual way. To determine the antimony and tin, the strong acid solu- tion in the flask, which contains the iron, is diluted with three times its volume of water. Antimony and tin are precipitated with hydrogen sulphide. After the precipitate has subsided, the clear solution is poured on a filter, the precipitate washed several times by decantation, and after- wards, on the filter, with hot water, till free from hydro- chloric acid. Portions of the sulphides often adhere to the walls of the flask in which the precipitation took place. These are washed out with concentrated sodium sulphide 144 QUANTITATIVE ANALYSIS BY ELECTKOLYSIS. solution, and the solution is poured on the filter containing- the sulphides. The filtrate is collected in a tared platinum dish. The filter, on which some iron sulphide always remains after the solution of the antimony and tin sulphides, is washed with sodium sulphide solution, the necessary amount of sodium hydroxide is added to the filtrate, and the antimony and tin are separated electrolytically as already directed. TIN AND PHOSPHORIC ACID. In the determination of metals, in the presence of phos- phoric acid, the latter is often removed as tin phosphate. The phosphoric acid is then usually determined in a separate portion, as its determination in the tin precipitate is too difficult and slow a process. The precipitate of tin oxide and tin phosphate may, however, be dissolved by digestion with ammonium sulphide, the solution diluted, the tin pre- cipitated by electrolysis, and the phosphoric acid determined as usual. PLATINUM AND IRIDIUM. As stated on p. 105, platinum can be separated from a hydrochloric acid solution, in a compact condition, by a very weak current. This fact may be utilized for its separation from iridium. If the current from two or three Meidinger cells, or a single Bunsen cell, is passed through an acidified solution of platinum and iridium, the platinum is separated without the least trace of iridium. SEPARATION OF GOLD FROM OTHER METALS. Edgar F. Smith has been often quoted as having given much study to the behavior of the metallic cyanides under SEPARATION OF THE METALS. 145 the electric current. He bases thereon a method for the sepa- ration of gold from palladium, copper, nickel, zinc, and plati- num. The conditions are the same as those already repeatedly given. When the solution of about 150 cc. contains about 3 gm. potassium cyanide, a current yielding 0.5-1 cc. electro- lytic gas per minute precipitates gold free from the other metals. This method serves also to separate silver or mercury from platinum. The silver or mercury is thrown down entirely free from platinum. POTASSIUM AND SODIUM. The ordinary method of determining potassium and so- dium in the same solution is to weigh the mixed chlorides, and the potassium as platinchloride ; the sodium is thus deter- mined by difference. The errors of the work, therefore, all fall on the sodium. The potassium may be determined, as already directed (p. 113), by precipitating as potassium platin- chloride, and determining the platinum in the latter by elec- trolysis. To determine the sodium directly, the filtrate from the potassium pLitinchloride is evaporated on the water-bath to remove alcohol, the residue dissolved in water with the ad- dition of a little hydrochloric acid, and the platinum removed by electrolysis. The sodium chloride in the solution poured off from the platinum is determined by evaporating to dry- ness, and weighing the residue. SODIUM AND AMMONIA. The direct determination of both is accomplished as with potassium and sodium ; the ammonia is precipitated as ammo- nium platinchloride, and the process conducted as described above. PAET II. -SPECIAL PART. ALLOY OF COPPER AND ZINC (LEAD, IRON). Brass. FOE the separation of the copper from the other metals, it Is necessary to precipitate it from acid solution. A nitric or sulphuric acid solution may be used. The use of a solution containing free nitric acid has the disadvantage, that if the action of the current is continued after all the copper is precipitated, more or less zinc is carried down with the copper. The presence of nitric acid or a nitrate also hinders the electrolytic separation of the zinc. If this acid is used, therefore, the solution, after removal of the copper, must be repeatedly evaporated to dryness with hydrochloric acid to convert the nitrates into chlorides. For the analysis of the alloy, 0.1-0.2 gm. is dissolved in as little nitric acid as possible, evaporated to dryness in the water-bath, the residue dissolved in 150-200 cc. water, and 15-20 cc. of nitric acid (sp. gr., 1.21) added. From this solution, the copper is precipitated by electrolysis (p. 89). The current is continued as long as a drop of the solution gives a blue color with ammonia. In order to separate copper from the other metals in sulphuric acid solution, the nitric acid solution of the alloy is evaporated on the water-bath, with addition of about 5 cc. 147 148 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. dilute sulphuric acid, till the residue no longer smells of nitric acid, dissolved in 150-200 cc. water, and treated as. before. When the reduction from either nitric or sulphuric acid solution is complete, the precipitate of metallic copper is washed free from acid without interrupting the current (p. 89), the contents of the dish siphoned off, and the copper, after washing with "absolute alcohol, determined as already described (p. 89). To determine the zinc in the solution, it is concentrated to about 100 cc., the free sulphuric acid very nearly neutralized with ammonia, the zinc converted into the double oxalate, and electrolyzed. The process is exactly as- described on p. 83. As a rule, brass contains small quantities of lead and iron. The presence of the former is shown, during the electrolysis of the copper, by the appearance, on the positive electrode, of a slight brown coating of lead peroxide. The positive electrode is then washed as usual, dried, and its increase in weight determined (p. 97). If iron is present, it is reduced with the zinc. It is- determined according to directions on p. 116. ALLOY OP COPPER AND SILVER. Silver Coin. The alloy is analyzed by dissolving 0.1-0.2 gm. in dilute nitric acid, evaporating off the acid in the water-bath, dis- solving the residue in water, and treating the solution accord- ing to directions on p. 130. SOLDEE. 149 ALLOY OF TIN AND LEAD. Solder. A small quantity of the alloy is digested with nitric acid till the tin is entirely converted into oxide, the excess of nitric acid evaporated, the solution diluted with water, and the tin oxide filtered off. The precipitate, after washing with water containing nitric acid, is dissolved in hot concen- trated hydrochloric acid, evaporated on the water-bath, and the aqueous solution of the residue converted into the double acid ammonium oxalate (see Tin, p. 111). On submitting the solution to electrolysis, the small quantity of lead which was precipitated with the tin oxide is deposited as dioxide on the positive electrode. When the reduction is ended, both elec- trodes are therefore weighed. The lead in the nitric acid ;solution is determined as directed on p. 97. ALLOY OP LEAD AND BISMUTH. The alloy is digested with nitric acid till completely dis- solved, the excess of acid evaporated off, and the lead precipi- tated as peroxide (p. 97). The bismuth in the solution, after evaporation of the nitric acid, is converted into bismuth ammonium oxalate, and electrolyzed as directed p. 91. ALLOY OP LEAD AND ZINC. The lead may be separated either as peroxide by elec- trolysis, or as sulphate by evaporation with dilute sulphuric .acid. In the former case, the lead-free solution is evaporated to dryness with addition of hydrochloric acid, and the zinc finally determined as directed p. S3. If the lead has been determined as sulphate, the alcohol is first removed, the 150 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. solution neutralized with ammonia, and the zinc determined as before. ALLOY OP BISMUTH AND COPPER. The alloy is dissolved in nitric acid, diluted with water,, and electrolyzed as directed p. 89 ; the copper, under the action of the prescribed current, is deposited perfectly free from bismuth. To determine the latter, the solution is evaporated to dryness in the water-bath to remove nitric acid, and the residue treated as directed p. 91. ALLOY OF COPPER AND TIN. Bronze. The alloy is oxidized with aqua regia, evaporated to- dryness, and the residue digested with a concentrated solu- tion of sodium sulphide. The copper sulphide which remains undissolved is filtered off, washed thoroughly first with sodium sulphide, then with hydrogen sulphide solution, dis- solved in nitric acid, and electrolyzed as directed p. 89. The tin is determined as directed p. 111. Results close enough for technical analysis are obtained by oxidizing the alloy with nitric acid, filtering off the tin oxide, dissolving it, after washing, in oxalic acid, adding acid am- monium oxalate, and precipitating the tin, as directed p. Ill* The copper is determined, as before, in the filtrate from the tin. oxide. PHOSPHOR-BRONZE. 151 ALLOY OP COPPER, TIN, ZINC, AND PHOSPHORUS. Phosphor-Bronze. When the alloy is digested with concentrated nitric acid to complete oxidation, a precipitate remains, which consists of a mixture of tin oxide and tin phosphate, with small quantities of copper oxide. It is filtered off, washed with water containing nitric acid, and heated with a concentrated solution of sodium sulphide. The residue of copper sul- phide is dissolved in nitric acid, and added to the principal solution. The tin is determined by converting the sodium sulphide into ammonium sulphide, and electrolyzing as directed p. 111. The phosphoric acid is determined in the filtrate in the usual manner. The nitric acid solution contains the copper and zinc. They are separated according to directions for the analysis of brass (p. 147). ALLOY OP COPPER, TIN, ZINC, MANGANESE, AND PHOSPHORUS. Manganese Phosphor-Bronze. The process is as already described ; the manganese remains with the zinc, and is finally separated as directed p. 123 ALLOY OF NICKEL AND COPPER. Nickel Coin. The analysis of this alloy is very simple. It is dissolved in nitric or sulphuric acid, and the copper precipitated 152 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. according to the directions given on p. 89. If the copper is precipitated from nitric acid solution, the acid is removed by evaporation, the nickel brought into solution as chloride, and the process conducted as directed p. 82. If a sulphuric acid solution is used, the acid is neutralized with ammonia, and the nickel determined as before. ALLOY OP COPPER, ZINC, AND NICKEL. German Silver. The alloy is dissolved in nitric acid, and the copper separated as directed p. 89. The copper-free solution is evaporated in the water-bath, the nitrates converted into chlorides by heating and evaporation with hydrochloric acid, the residue dissolved in water, with the addition of hydro- chloric acid, and diluted till 1 gin. of the two oxides corre- sponds to 50 cc. The solution is nearly neutralized by adding sodium carbonate till a slight permanent precipitate is formed, which is redissolved in a few drops of hydrochloric acid. The zinc is precipitated by passing hydrogen sulphide into the cold solution as long as a precipitate is formed, adding a few drops of sodium acetate, and allowing the precipitate to settle. The zinc sulphide is washed with hydrogen sulphide water to which a little ammonium nitrate is added, dissolved in concentrated hydrochloric acid, the acid evaporated off, and the residue treated for determination of zinc as directed p. 83. The filtrate is evaporated to dryness, and the nickel determined in the residue as directed p. 82. WOOD'S METAL. 153 ALLOY OP TIN, LEAD, BISMUTH, AND CADMIUM. Wood's Metal. The alloy is treated with nitric acid, whereby the tin Temains undissolved as oxide, contaminated with small por- tions of the oxides of lead and bismuth. This is filtered off, washed with water containing nitric acid, and dissolved in sodium sulphide. The tin is determined as directed p. 111. The insoluble sulphides of bismuth and lead are dissolved in nitric acid, and the solution added to the principal solution. The lead is determined as peroxide (p. 97), or as sulphate, observing the well-known precautions. The bismuth and cadmium are separated as follows : The solution is evaporated to remove nitric acid, the residue dissolved in the least possible amount of hydrochloric acid, and the bismuth precipitated as oxy chloride by the addition of much water. The precipitate is filtered off, washed, dissolved in a little dilute hydrochloric acid, and the bismuth precipitated as directed p. 91. The filtrate from the bis- muth oxychloride is evaporated to dryness, and the cadmium determined in the residue as directed p. 94. ALLOY OP TIN, LEAD, BISMUTH, AND MERCURY. The tin is separated from the other metals by oxidation with nitric acid, and treated as before. The mercury can now be precipitated from the acid solution (p. 102), and also a portion of the lead as peroxide at the positive electrode. To remove the whole of the lead, the mercury-free solution is again electrolyzed, using the platinum dish as the positive electrode. The lead-free solution is evaporated to dryness, and the bismuth determined as directed p. 91. The lead 154 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. may also be separated from the tin-free solution as sulphate,, the mercury then separated, by electrolysis, from the acid solution, and the bismuth finally determined as before. ALLOY OF LEAD AND ANTIMONY. Hard Lead. Type Metal. The two metals may be separated, either by oxidizing with nitric acid, evaporating to dry ness, and digesting the residue with sodium sulphide, or by heating the finely divided alloy with ten times its weight of anhydrous sodium thio- sulphate in a covered porcelain crucible, over a very low flame, till the mixture is sintered together, and extracting with water. In either case, lead sulphide remains undis- solved, and is filtered off, and washed first with sodium sulphide, and then with hydrogen sulphide, solution. It may be determined directly as sulphide, or as directed p. 97. The antimony is determined, in the filtrate from lead sulphide, exactly as directed p. 107. ALLOY OP ANTIMONY AND TIN. The method of analysis has been already given on p. 138.. The alloy is oxidized with nitric acid, and the residue, after evaporation, dissolved in a concentrated solution of sodium sulphide, sodium hydroxide added, and the process followed throughout as given on p. 139 ALLOY OP ANTIMONY AND ARSENIC. It has already been stated (p. 141) that the two metals can be separated under conditions similar to those in the separation of antimony from tin ; the method requires the arsenic .to be oxidized to arsenic acid. The alloy is digested ALLOY OF ANTIMONY, TIN, AND ARSENIC. 155 with aqua regia, the acid removed by evaporation, the residue dissolved in concentrated sodium sulphide, sodium hydroxide added, and the directions given on p. 139 followed throughout. ALLOY OF ANTIMONY, TIN, AND ARSENIC. When this alloy is oxidized with aqua regia, and a solu- tion in sodium sulphide prepared as above, antimony alone is electrolytically deposited in presence of tin. The method is, described on p. 142. SPATHIC IRON ORE. Constituents : Ferrous Carbonate, with Manganese, Calcium, and Magnesium Carbonates (Gangue). All the constituents of the mineral may be determined in the same solution. About 0.5 gm. of the dry mineral is dissolved in a porcelain dish, in the least possible amount of hydrochloric acid, the acid removed by evaporation, and the residue taken up with water to which a little hydrochloric acid is added. If insoluble gangue is present, this is filtered off, washed with water, and weighed. The metals are con- verted into oxalates by treatment with potassium and ammo- nium oxalate, and the insoluble residue of calcium oxalate filtered off, and washed with hot water. If manganese is pres- ent, the calcium oxalate always carries down some manganese oxalate. * When the precipitate is ignited, a mixture of CaO and Mn 2 O 3 is obtained. It is weighed, and the manganese in it determined volumetrically.f The iron and manganese are separated as directed p. * Classen, Zts. Anal. Ch., 16, 318. f Classen, Quant. Anal., 4th ed., p. 128. 156 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. the manganese finally precipitated as sulphide, and the mag- nesium in the filtrate as magnesium ammonium phosphate. If magnesium is absent, the manganese is determined as rnangano-mangaiiic oxide or sulphate (p. 120). HEMATITE. Constituents : Ferric Oxide, Manganic Oxide (Copper Oxide, Alumina, Lime, Magnesia), Phosphoric Acid, Sulphuric Acid. The iron, manganese, and calcium are determined as above. If copper is present, it is first separated from the other metals by first submitting the double oxalate solution to a very weak current. If, in addition to iron (copper, if present) and manganese, phosphoric and sulphuric acids are to be determined, the metals are converted into double oxalates, and iron and manganese completely removed (see separation of Iron and Manganese, p. 118) ; the two acids may now be determined in the solution entirely free from manganese. If only one acid is to be determined, the whole filtrate can be used ; otherwise it is diluted to a known volume, and aliquot portions taken for analysis. In the determination of either acid, the solution is first acidified with hydrochloric acid,* and then treated either with barium chloride, or with one-third its volume of ammonia, and mag- nesia mixture. About 1 gm. of the mineral is needed for the determination of sulphuric and phosphoric acids. If alumina, as well as phosphoric acid, is present in hematite (its presence is shown by a white turbidity f of * If the acid carbonates produced from the oxalates are not decomposed, small hard crystals of acid carbonates are precipitated together with am- monium magnesium phosphate. These crystals are difficultly soluble in ammonia, and may make the results too high. t A turbidity often appears when the solution is first heated, caused by the driving off of ammonium compounds. DETERMINATION OF IRON, MANGANESE, ETC. 157 aluminium phosphate and hydroxide in the solution under- going electrolysis), the manganese must always be converted into sulphide. The iron-free solution is boiled to decompose hydrogen ammonium carbonate, tartaric acid or a solution of a tartrate added till the precipitate of aluminium hydroxide disappears, and the weakly ammoniacal solution precipitated hot with ammonium sulphide. The green manganous sulphide is determined as hereto- fore directed. The phosphoric acid may be determined with magnesia mixture, in the filtrate from the manganese sul- phide. To determine sulphuric acid in presence of alumina, iron and manganese are removed, by electrolysis, from a separate portion, the solution is -poured off, the ammonium carbonate decomposed by heat, the solution acidified with hydrochloric acid, and the sulphuric acid determined with, barium chloride. Determination of Iron, Manganese, Copper, Calcium, Magnesium, Phosphoric Acid, and Sulphuric Acid. The method of determining iron, manganese, etc., in the same solution has already been given. If it is desired to determine magnesium and phosphoric and sulphuric acids, in the filtrate from manganese peroxide, it is diluted to a known volume, magnesium is determined in an aliquot part with ammonium phosphate, and phosphoric and sulphuric acids in two other portions. 158 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. LIMONITE. Constituents : Ferric Hydroxide, together with Manganese Oxide (Lime, Magnesia), Phosphoric Acid, Sulphuric Acid, Silica, and Gangue. The analysis may be conducted like those of hematite and spathic iron ; but care must be taken, at the outset, to convert the silica into the insoluble modification by evapo- rating the solution, and drying the residue. CLAY IRON-ORE, Constituents: Iron Oxide, Alumina, Manganese, and Water. The mineral is digested with concentrated hydrochloric acid till it is completely decomposed, the insoluble residue is filtered off, the filtrate evaporated to remove free acid, the residue dissolved in water with a few drops of hydrochloric acid, and the iron separated from aluminium and manganese as directed p. 124. BOG IRON-ORE. Mixture of Ferric Hydroxide with Ferrous and Ferric Silicates, Manganese, Alumina, Copper, Calcium, Magnesium, Sulphuric Acid, Phosphoric Acid, Arsenic Acid, Organic Matter, and Gangue. The analysis of the mineral is easily understood from the foregoing. Arsenic and copper are best determined by eliminating the former as chloride, as directed p. 142. and precipitating the copper with hydrogen sulphide in the greatly diluted residue left in the distillation flask. The copper sulphide is dissolved in nitric acid, and determined electrolytically as directed p. 89. CHROME IRON ORE. PSILOMELANE. 159 CHROME IRON ORE. Constituents : Chromium Oxide, Ferrous and Ferric Oxides, Alumina, Manganese, Calcium, Silica. The finely powdered mineral is fused for a long time with sodium carbonate and potassium chlorate, and the fused mass extracted with water. The residue contains oxides of iron, manganese, calcium, magnesium, and aluminium, and traces of chromium and silica ; the solution, chromic acid, silica, and some alumina and lime. The residue is dissolved in hydrochloric acid, the solution evaporated to dryness to separate silica, the residue treated with water and a little hydrochloric acid, and filtered. The metals in the filtrate are converted into double oxalates. If manganese is present, the precipitate of calcium oxalate must be treated as directed p. 155. The filtrate from the calcium oxalate, which contains iron, manganese, aluminium, and chromium, is treated as directed p. 124. The aqueous solution from the fused mass is evaporated to separate silica, the calcium precipitated as oxalate, and the aluminium and chromic acid separated accord- ing to previous directions. Edgar F. Smith recommends the use of the galvanic cur- rent for the decomposition of chrome iron ore. The process, according to his directions, is conducted as follows : Thirty or forty gm. potassium hydroxide are heated in a nickel crucible until the mass is in a condition of quiet fusion. The chrome iron ore for decomposition (about 0.5 gm.) is finely pulverized, weighed on a watch-glass, and gradually added, with the help of a camePs-hair pencil, to the crucible contain- ing the fused alkali. The crucible is then covered with a perforated watch-glass and connected with the anode of the battery or other source of current. The kathode employed is a thick platinum wire, which is plunged through the opening 160 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. in the watch-glass into the fused mass. To regulate the current an amperemeter (p. 61) is inserted, and a switch is also placed in the circuit, so adjusted as readily to produce the reversal of the current, which is necessary toward the close of the process. The current strength must not exceed 1 ampere. After about 30 minutes the current is reversed by the switch, so that the crucible becomes the kathode, and the platinum wire the anode. The object of this reversal is to oxidize completely the last traces of the mineral, minute portions of which may have been protected by metallic iron which had been deposited by the current. After the current has acted in this direction for 10 minutes, the decomposition is complete. The fused mass of course contains the chromium as chromate. The author pursued similar researches some years since,, and can confirm Smith's results. FSILOMELANE. Constituents : Manganous Oxide, Copper Oxide, Ferric Oxide, Nickel Oxide, Cobalt Oxide, Alumina, Lime, Potash, Soda, and Lithia. Determination of Manganese, Copper, Iron, Aluminium, Nickel, Cobalt, and Calcium. A weighed portion of the mineral is dissolved in hydro- chloric acid, evaporated to dryness, dissolved in water with a few drops of hydrochloric acid, converted into double oxalates, calcium oxalate filtered off, and the calcium and manganese in the precipitate determined as directed p. 155. In the filtrate, the copper is first determined electrolyti- cally (p. 89). After the precipitation of the copper is- complete, the solution, which contains the other metals, is decanted from the copper precipitate, and is then again submitted to electrolysis for the precipitation of iron, co- PSILOMELANE. 161 bait, nickel, and manganese, the latter as dioxide at the positive electrode. After the electrolysis is completed, the solution is decanted from the precipitated metals, and the remaining manganese completely precipitated, according to directions given on p. 120. If only the weight of nickel and cobalt together is desired, the precipitate containing the three metals is dissolved in hydrochloric acid, and the iron determined by titration with potassium permanganate as directed p. 115. Otherwise the cobalt and nickel must first be separated from the iron. The precipitate of the metals is dissolved in hydrochloric acid, the acid removed by evapora- tion, the residue oxidized with hydrogen peroxide or bromine water, dissolved in water with a few drops of hydrochloric acid, and the metals converted into double oxalates by addi- tion of potassium oxalate in slight excess. From the boiling solution, which should have a volume of 80-100 cc., the cobalt and nickel are precipitated as oxalates by concen- trated acetic acid. A great excess of acetic acid must be used, and the solution, after the filtrate has subsided, must be tested with the reagent for a further precipitate. The filtrate from the cobalt and nickel oxalates contains all the iron as potassium iron oxalate.* The precipitate of nickel and cobalt oxalates is washed with a mixture of equal parts of alcohol, acetic acid, and water, and, after drying to remove acetic acid and alcohol, is dissolved on the filter with hot water containing potassium and ammonium oxalates. The solution is electrolyzed as directed p. 81. The sum of nickel and cobalt is determined, the metals dissolved in hydrochloric acid, evaporated to dry- ness, the residue dissolved in a few drops of water, potassium hydroxide added in slight excess, and the resulting precipi- * Classen, Zts. Anal. Ch., 18, 189. 162 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. tate dissolved in concentrated acetic acid. The cobalt is precipitated with a saturated solution of potassium nitrite acidified with acetic acid. The precipitate, after standing twenty-four hours, is filtered off, washed with potassium nitrite, and dissolved in hydrochloric acid, the solution is evaporated to dryness, and the residue converted into the double oxalate, and electrolyzed. The nickel is determined by difference. The nickel may also be determined, instead of, or in addition to, the cobalt, by precipitating nickel with potassium hydroxide, in the filtrate from the cobalt potas- sium nitrite, filtering, dissolving in hydrochloric acid, and separating nickel electrolytically as directed p. 82. To determine the iron in the filtrate from cobalt and nickel oxalates, the alcohol and acetic acid are completely removed by evaporation, the residue dissolved in water, and the iron electrolytically deposited from the solution of the double oxalate (p. 78). Determination of Potassium, Sodium, Lithium, Calcium, and Magnesium. The mineral is dissolved in hydrochloric acid, evaporated to remove acid, and treated with an excess of ammonium oxalate. The filtrate from calcium oxalate is electrolyzed, iron, nickel, cobalt, and copper separating as metals, man- ganese as dioxide, and aluminium as hydroxide. The filtrate from the manganese dioxide and aluminium hydroxide con- tains only alkalies, magnesium, and a little manganese. It is boiled to 'remove the hydrogen ammonium carbonate formed by the electrolytic decomposition of ammonium oxalate, con- centrated to about 50 cc., heated to boiling, and at least an equal volume of concentrated acetic acid added. The pre- cipitate consists of manganese and magnesium oxalates. It SPHALERITE (ZINC BLENDE). 163 is filtered off, washed with a mixture of equal volumes of alcohol, acetic acid, and water, and ignited. The residue is MgO + Mn 2 O 3 . It is weighed, dissolved in hydrochloric acid, and the manganese determined by electrolysis as dioxide (p. 84). The alkalies are determined in the filtrate from the man- ganese and magnesium oxalates. It is evaporated to dryness, the ammonium salts removed by gentle ignition, the residue dissolved in water, the solution filtered, and evaporated to dryness after addition of a little hydrochloric acid. The residue is washed into a small stoppered flask with absolute alcohol, an equal volume of water-free ether added, and allowed to stand twenty-four hours. The solution is then filtered from the residue, the alcohol and ether evaporated, and the lithium chloride converted into sulphate and weighed. The residue of potassium and sodium chlorides is dissolved in water, and both metals directly determined as directed p. 145. SPHALERITE (ZINC BLENDE). Constituents : Zinc Sulphide, also Determinable Quantities of Iron, Manganese, Copper, Arsenic, Antimony, and Gangue. In most cases, it is only necessary to determine the zinc. The process is then as follows : About 0.5 gm. of the finely powdered mineral is digested with concentrated nitric acid till fully decomposed, the acid evaporated off, and the nitrates converted into chlorides by evaporation with hydrochloric acid. The residue is dissolved in about 25 cc. water and 10 cc. hydrochloric acid, and hydrogen sulphide passed through the solution. The precipitate of sulphides of lead, copper, etc., is filtered off, washed with water containing hydrogen sulphide and hydrochloric acid, and the filtrate evaporated to dryness. The residue contains chlorides of 164 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. zinc, iron, manganese, calcium, and magnesium. It is dis- solved in water with a little hydrochloric acid, converted into double oxalates (p. 78), the calcium oxalate filtered off t and the filtrate electrolyzed. Zinc and iron separate at the- negative electrode, and manganese, as dioxide, at the positive. The two metals are weighed, dissolved in hydrochloric acid, and the iron determined by titration with potassium per- manganate (p. 115). It is stated on p. 116 that the precipitation of iron and zinc from the same solution is complete only when there is less than one-third as much zinc as iron, and that it can be successfully performed, in other cases, by adding a weighed quantity of an iron salt before the electrolysis. Determination of Lead, Copper, Arsenic, Antimony, Zinc, Iron, Manganese, and Gaiigue. , As when zinc alone is to be determined, the mineral is oxidized with nitric acid, the gangue filtered off, and the acid solution of chlorides treated with hydrogen sulphide. The precipitated sulphides are washed first with hydrogen sulphide water containing hydrochloric acid, and afterward with pure hydrogen sulphide water. The antimony and arsenic are separated from lead and copper by digestion with a concentrated solution of sodium sulphide ; the residue is washed with the same solution, and afterward with hydrogen sulphide solution. The sodium sulphide washings are added to the solution for determina- tion of arsenic and antimony, and the hydrogen sulphide washings separately collected. The necessary amount of sodium hydroxide is added ta the sodium sulphide solution, and the antimony and arsenic separated and determined as directed p. 140. CALAMINE AND SMITHSONITE. ULTKAMAKINE. 165 The sulphides of lead and copper are dissolved in nitric acid, and the metals determined as directed p. 147. Iron, zinc, and manganese are determined according to previous directions. CALAMINE AND SMITHSONITB. Constituents: Zinc (Cadmium), Copper, Lead, Arsenic, Antimony, Iron, Manganese, Calcium, Magnesium, Silica, Carbonic Acid, Water. Zinc and the other constituents are determined as already directed. If the mineral contains cadmium, copper and lead are first precipitated from the nitric acid solution, the decanted solution evaporated to dryness, the cadmium nitrate converted into chloride, and cadmium determined as directed p. 94. ULTRAMARINE. Constituents : Alumina, Potassium, Sodium, Iron, Calcium, Sulphur, Silica, Sulphuric Acid, Chlorine. A weighed portion of the substance is dissolved in hydro- chloric acid, evaporated to dryness to separate silica, the residue dissolved in water with a few drops of hydrochloric acid, filtered from the silica, the free acid neutralized with ammonia, and a great excess of ammonium oxalate added. The calcium oxalate is filtered off, iron and aluminium deter- mined electrolytically, the solution filtered from the alu- minium hydroxide, evaporated to dryness, the ammonium .salts removed by gentle ignition, the residue dissolved in water, and the alkalies converted into chlorides by evapora- tion with hydrochloric acid. Potassium and sodium are determined as directed p. 145. 166 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. REFINERY SLAG. Constituents : Ferrous and Ferric Oxides, Metallic Iron, Copper, Aluminium, Calcium, Magnesium, Silica, Sulphuric and Phos- phoric Acids. A portion of the substance (0.5-1 gm.) is dissolved in hydrochloric acid, evaporated to remove silica, the residue dissolved in hydrochloric acid, evaporated to remove free acid, and the metals converted, as usual, into double oxalates. Calcium oxalate is filtered off, and the manganese in the precipitate determined as directed p. 155. The copper is separated by the action of a weak galvanic current (p. 89), and the iron, manganese, and aluminium separated in the copper-free solution as directed p. 123. For the determination of magnesium and sulphuric and phosphoric acids, see Hematite, p. 156. To determine the metallic iron, about 5 gm. of the finely powdered slag is placed in a small platinum or porcelain dish, and treated with an aqueous solution of copper sulphate. A quantity of metallic copper equivalent to the iron is precipi- tated (CuSO 4 + Fe = FeSO 4 + Cu). The decomposition is hastened by frequent stirring ; the copper and undecom- posed slag are finally filtered off, washed thoroughly, and digested in the water-bath for a long time with nitric acid. In the solution, after filtration, the copper is electrolytically determined, and the quantity of iron calculated from it. COPPER AND LEAD SLAGS. Constituents : Copper, Lead, Iron, Manganese, Barium, Calcium, Magnesium, Silica, Sulphuric Acid, Sulphur, and ordinarily small quantities of Arsenic, Antimony, Bismuth, Cobalt, Nickel, and Zinc. The slag is decomposed by digestion with nitric acid, evaporated to dryness, the residue taken up with water and REFINERY SLAG. COPPER AND LEAD SLAGS. 167 a little hydrochloric acid, and the solution filtered from the residue of silica and barium sulphate, which are separated as usual. The calcium is separated by adding ammonium oxa- late in great excess ; the calcium and the manganese it may contain are determined as directed p. 155. Copper is then precipitated (p. 89), and afterward iron and manganese (p. 118), and magnesium and sulphuric acid are determined as directed p. 156. In the presence of arsenic, antimony, etc., the hydrochloric acid solution, after separation of silica, is treated, first hot and then cold, with hydrogen sulphide gas, and the precipi- tated sulphides are washed with hydrogen sulphide water, and treated with a concentrated solution of sodium sulphide. The insoluble sulphides of lead, copper, etc., are washed first with sodium sulphide, and then with hydrogen sulphide (see p. 154), and antimony and arsenic are separated in the solution as directed on p. 140. The residue of lead sulphide, etc., is digested with nitric acid till thoroughly decomposed, and lead and copper sepa- rated from the solution as directed p. 130. The nitric acid is evaporated off, and bismuth nitrate is converted into chloride, and determined as directed p. 91. The solution filtered from the hydrogen sulphide precipi- tate, which contains iron, manganese, etc., is evaporated almost to dryness to remove hydrogen sulphide and most of the hydrochloric acid, and the metals finally converted into double oxalates. Calcium oxalate is filtered off, and the precautions described on p. 155 are observed in its deter- mination. By electrolysis of the filtrate, iron, cobalt, nickel, and zinc are obtained as metals, and manganese, in part, as dioxide ; magnesium remains in solution. The two latter are determined as directed p. 157. The iron, cobalt, etc., are dissolved in concentrated hydro- 168 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. chloric acid, the solution evaporated to dryness, the residue dissolved in water with a few drops of acetic acid, potassium oxalate added in sufficient quantity to form the double oxa- lates, the solution diluted to 25-30 cc., and precipitated, at boiling heat, with concentrated acetic acid in great excess. After standing about six hours in a warm place, the oxalates of cobalt, nickel, and zinc are filtered off, washed with a mixture of equal volumes of acetic acid, alcohol, and water, and the oxalates converted, by very gentle heating, into oxides. The mixed oxides are dissolved in hydrochloric acid, and zinc separated from nickel and cobalt as directed on p. 152. Iron is determined, in the nitrate from the oxa- lates, as directed on p. 162. BLAST FURNACE, CUPOLA, AND BESSEMER SLAGS. Constituents : Ferrous and Ferric Oxides, Metallic Iron, Man- ganese, Aluminium, Copper Lead, Zinc, Calcium, Magnesium, Alkalies, Silica, Sulphuric and Phosphoric Acids, Sulphur (as Calcium Sulphide). The method of analysis is so similar to the foregoing that it needs only brief mention. The slag is digested with fuming hydrochloric acid, or aqua regia, till completely decomposed, the solution evaporated on the water-bath to dryness, the residue dissolved in water and a little hydro- chloric acid, and the silica filtered off. After conversion into double oxalates, the calcium oxalate, which may contain manganese, is filtered off (p. 155), copper and lead first precipitated (p. 130), then iron and zinc with aluminium and the rest of the manganese ; iron and zinc are determined as directed p. 116, and manganese, aluminium, and magnesium as directed p. 156. The alkalies and sulphuric and phos- phoric acids are determined as heretofore directed. ZIRCON. AESENOPYRITE. 169 ZIRCON. Constituents : Zirconia, Iron Oxide, Lime, Silica. The mineral is decomposed by long-continued fusion with sodium carbonate, the fused mass dissolved in hydro- chloric acid, the solution evaporated to dryness, the residue taken up with water acidified with hydrochloric acid, the silica filtered off, and the filtrate treated with a great excess of ammonium oxalate. To overcome the injurious effect of sodium chloride, about 10 gm. ammonium oxalate must be dissolved by heating in the solution diluted to about 200 cc. Iron and zircon are separated as directed p. 127. If calcium is present, the calcium oxalate precipitate is, of course, to be filtered off before electrolysis, and determined. ARSENOPYRITE. Iron, Arsenic, Antimony, Sulphur, Gangue. A portion of the finely powdered mineral is oxidized with aqua regia till fully decomposed, the gangue filtered off, and the solution evaporated to dryness. The chlorides are con- verted into sulphates by moistening and heating with sul- phuric acid, water is added, the solution heated to 70-80, and hydrogen sulphide passed till it has cooled completely. After standing some twelve hours at a moderate heat, the sulphides of arsenic and antimony are filtered off, and sepa- rated as directed p. 140. To determine the iron, the hydrogen sulphide is driven off from the solution, which is then treated as directed p. 78. 170 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. CHALCOPYRITE (COPPER PYRITES). Constituents : Copper, Iron, Sulphur, Gaiigue. The mineral is oxidized with nitric acid, the gangue filtered off, and copper precipitated in the filtrate (p. 89). To determine iron, nitric acid is removed by evaporation, concentrated hydrochloric acid added, the solution again evaporated, and finally iron is precipitated, after formation of the double oxalate, according to directions on p. 78. Sulphur may be determined in the same portion by pre- cipitating sulphuric acid with barium chloride, and removing the excess of the latter by careful addition of sulphuric acid. Copper is then separated from iron in sulphuric acid solution,, and the latter determined as usual. As already stated on p. 91, copper cannot be precipitated from either nitric or sulphuric acid solution in the presence of any considerable quantity of arsenic and antimony without being contaminated by them. If only the copper is to be determined, the nitric acid solution of the mineral is evaporated to dry ness, the residue dissolved in water with a little acetic acid, and potassium oxalate added in excess. The solution is filtered hot from the gangue, the residue washed with water containing potas- sium oxalate, and the filtrate brought to a volume of about 50 cc. After cooling, almost all the copper crystallizes out as potassium copper oxalate ; the rest is precipitated by addition of much concentrated acetic acid. The precipitate is washed with a mixture of equal volumes of water, acetic acid, and alcohol, dissolved in ammonium oxalate, and elec- trolyzed. If arsenic and antimony are present in larger proportion, the finely pulverized mineral is mixed with four times its NICKEL MATTE. COPPER MATTE. 171 weight of ammonium chloride, and heated gently in a covered crucible. Arsenic and antimony, and the greater part of the iron are volatilized as chlorides.* The residue is dissolved in nitric acid, and treated as. before. NICKEL MATTE. COPPER MATTE. Nickel, Cobalt, Zinc, Iron, Copper, Lead, Arsenic, Antimony, Sulphur, Gangue. The substance is decomposed with aqua regia, evaporated to dryness, the residue dissolved in hydrochloric acid, and filtered from the gangue. In this solution, the metals pre- cipitable by hydrogen sulphide are precipitated by heating to 70-80, and passing hydrogen sulphide gas till the solution becomes cold. The precipitate is filtered off, washed first with a solution containing hydrogen sulphide and hydro- chloric acid, then with pure hydrogen sulphide solution, and heated with^a concentrated solution of sodium sulphide as- directed p. 139, and the arsenic and antimony separated and determined as directed p. 140. The sulphides of lead and copper left undissolved by sodium sulphide are digested with nitric acid, and deter- mined as directed p. 130. The filtrate from the hydrogen sulphide precipitate is evaporated to dryness to remove hydrogen sulphide and hydrochloric acid, the residue dis- solved in water with a little acetic acid, potassium oxalate added in excess, and the solution of 50-100 cc. precipitated boiling hot with a great excess of concentrated acetic acid (at least an equal volume). The precipitate of nickel, cobalt, and zinc oxalates is filtered off, washed with a mixture of equal volumes of alcohol, acetic acid, and water (p. 161), * Classen, Zts. Anal. Ch., 18, 388. 172 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. dried, and converted, by gentle ignition, into oxides. The residue is dissolved in hydrochloric acid, and zinc, cobalt, and nickel separated, and determined as directed pp. 152 and 161. Iron is determined, in the filtrate from the mixed oxalates, as direqted p. 162. ^ COPPER SPEISS, LEAD SPEISS. Antimony or Arsenic Compounds of Iron, Cobalt, and Nickel, together with Sulphur Compounds of Copper, Lead, Silver, Bismuth, Iron, and Zinc. It is best to decompose the finely powdered substance in a suitable apparatus * with chlorine gas, volatilizing arsenic, antimony, iron, and zinc, as chlorides, and collecting them in a receiver containing equal volumes of hydrochloric and tartaric acids. The free chlorine is expelled, by heat, from the solution in the receiver, and hydrogen sulphide passed into the still hot solution until it cools. The sulphides are filtered, washed, treated with sodium sulphide, and arsenic and antimony determined in the solution, as directed p. 140. The insoluble sulphides of iron and zinc are dissolved in hydrochloric acid, evaporated to dryness, the residue dis- solved in water with a few drops of hydrochloric acid, and iron and zinc determined as directed p. 116. After the decomposition with chlorine, the non-volatile chlorides of copper, lead, silver, bismuth, cobalt, and nickel, and a part of the iron and zinc, remain in the bulb. They are dissolved in dilute hydrochloric acid, and lead, copper, silver, and bismuth precipitated with hydrogen sulphide. The sulphides are digested with nitric acid till completely * Classen, Quantitative Analyse, 4th ed. p. 187. PYRARGYRITE. TETRAHEDRITE. 173 dissolved, and copper and silver precipitated as metals, and lead as peroxide, by electrolysis. Copper and silver are separated as directed p. 130, and bismuth from some residual lead as directed p. 134. The separation of cobalt and nickel from iron and zinc is given on pp. 152 and 161. FYRARGYRITE. Silver, Antimony (Arsenic), Sulphur, Gangue.' The mineral may be decomposed by chlorine gas, or heating with anhydrous sodium thiosulphate. In the former case, the chlorides of antimony and arsenic (and sulphur) go into solution, while silver chloride remains in the bulb tube. In the latter case, when the fused mass is treated with water, silver sulphide remains undissolved, and may be dissolved in nitric acid, and the silver deposited, as metal, from the solu- tion (p. 100). To determine antimony, and separate it from arsenic, the solution of sodium pentasulphide is oxidized with hydrogen peroxide, evaporated, and treated as in the determination of antimony in presence of tin (p. 138). TETRAHEDRITE. Copper, Antimony, Arsenic, Silver, Lead, Iron, Zinc, Sulphur, Gangue. The mineral may be decomposed as heretofore described. When chlorine gas is used, the receiver contains chlorides of antimony, arsenic, iron, and zinc (and sulphur) ; the bulb- tube, copper, lead, silver, and gangue, with a portion of the iron and zinc. The metals are separated as already described. 174 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. FURNACE "SOWS." Alloys of Iron (the principal constituent), Copper, Silver, Lead, Molybdenum, Vanadium, Cobalt, Nickel, and Zinc, with Sulphides and Phosphides of these Metals, and varying amounts of Carbonic Acid and Silica. The substance is best decomposed by chlorine gas. The quantity of iron is so great, however, that two bulb-tubes of the most infusible glass should be used, in the second of which is deposited most of the iron chloride. The substance is heated in a stream of chlorine as long as iron chloride sublimes ; then it is certain that all the molybdenum chloride will have been carried over into the receiver, which also contains vanadium, sulphur, and phosphorus chlorides. Hydrogen sulphide is passed into the solution collected in the receiver until the supernatant liquid is colorless. The precipitate of molybdenum sulphide is filtered off, washed, oxidized with nitric acid, the solution supersaturated with ammonia, and molybdenum oxide precipitated by electrolysis. The filtrate from molybdenum sulphide contains vana- dium and iron. Hydrogen sulphide and hydrochloric acid are evaporated off, double oxalates formed, and the two metals separated electrolytically as directed p. 127. To determine vanadium in the solution decanted from the iron, it is evaporated to dryness, the ammonium salts driven off by careful ignition, and the residue of vanadium oxide converted, by fusion with potassium nitrate, into potassium vanadate. The fused mass is dissolved in water, nitric acid added not to acid reaction, then a concentrated solution of ammonium chloride, and then alcohol in the proportion of one volume to three of the solution. After standing forty- eight hours, the ammonium vanadate is filtered off, and washed with a concentrated solution of ammonium chloride, STIBNITE (ANTIMONY GLANCE). ULLMANITE. 175 and then with alcohol. The salt is heated first in the air, then in a stream of oxygen, and leaves a residue of pure vanadic acid which is weighed. The chlorides remaining in the bulb-tube are heated with hydrochloric acid ; a residue of silver chloride and carbon remains. It is heated with potassium cyanide, the carbon filtered off, and the silver determined by electrolysis. The methods of separation and determination of the metals in the hydrochloric acid solution have already been given. STIBNITE (ANTIMONY GLANCE). Constituents: Antimony and Sulphur, and usually small quantities of Iron, Lead, Copper, and Arsenic. The simplest method of analyzing the mineral is to mix with four or five times its weight of anhydrous sodium thiosulphate, and heat for a long time in a covered crucible (p. 154). The fused mass is exhausted with water ; the solution contains antimony and arsenic, and is treated for decomposition of sodium pentasulphide and determination of the two metals as directed p. 140 ; the undissolved sulphides of lead, copper, and iron are oxidized with nitric acid, and the metals separated according to foregoing directions. ULLMANITE. Antimony, Nickel, and Sulphur. The finely powdered mineral is decomposed in a stream of chlorine (p. 172), all the antimony passing into the receiver as chloride, and nickel chloride remaining in the bulb-tube. The latter is determined by dissolving the con- tents of the bulb in hydrochloric acid, evaporating, convert- ing into the double oxalate, and precipitating by electrolysis. 176 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. Antimony is precipitated, as sulphide, by passing hydro- gen sulphide gas into the solution in hydrochloric and tar- taric acids, dissolved in concentrated sodium sulphide, the solu- tion diluted with water and submitted to electrolysis (p. 106). If the mineral contains iron, it passes over, as chloride, into the receiver ; it may be determined, in the filtrate from antimony sulphide, after supersaturation with ammonia, by precipitation as sulphide with ammonium sulphide. The sulphide thus obtained is dissolved, converted into the double oxalate, and iron determined electrolytically (p. 78). The analysis is made more simply if the mineral is decomposed by heating with sodium thiosulphate ; when the proportion of antimony is large, it is necessary to repeat the process with the residual nickel sulphide. Antimony is determined in the aqueous solution of the fused mass as- directed p. 108. If, on treatment with hydrogen peroxide, or addition of sodium monosulphide, some nickel sulphide separates, it is added to the principal portion. The sulphides of iron and nickel are oxidized with nitric a*cid, the nitrates converted into chlorides, and the two metals separated as directed (p. 115). BOURNONITE. Antimony, Lead, Copper (Iron), and Sulphur. The finely powdered mineral is heated either with chlorine or anhydrous sodium thiosulphate, and the analysis conducted as already described. ZINKENITE. Antimony, Lead (Silver, Copper, Iron), Sulphur. The mineral is most simply decomposed by heating with anhydrous sodium thiosulphate. After exhaustion with LIISTN^EITE. COBALTITE. 177 water, the residue of undissolved sulphides is dried, the filter burnt, and fusion with thiosulphate repeated. Anti- mony is determined according to directions on p. 108. The sulphides of lead, silver, etc., are oxidized with nitric acid ; copper and silver precipitated electrolytically, and separated as directed p. 130. A portion of the lead is separated, as peroxide, by the electrolysis of the nitric acid solution, and is determined as such. The rest is precipitated with hydro- gen sulphide, the filtrate neutralized with ammonia, ammo- nium oxalate added, and iron determined by electrolysis. LINN^BITE. Constituents : Cobalt and Sulphur. The analysis of this mineral is very simple. It is dis- solved in aqua regia, the free acid evaporated off, and chlorides formed by repeated evaporation with hydrochloric acid. The aqueous solution of the residue is treated with an excess of ammonium oxalate, and cobalt precipitated electro- lytically (p. 81). If iron is present, the two metals are separated as directed p. 115. In the solution decanted from the metallic cobalt, ammo- nium carbonate is decomposed by boiling, hydrochloric acid is added, and the sulphur determined by precipitation with barium chloride. COBALTITE. Cobalt, Iron (Copper, Antimony), Arsenic, and Sulphur. The mineral may be decomposed by heating with nitric acid, or with sodium thiosulphate. If nitric acid is used, the free acid is evaporated off, and the nitrates converted 178 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. into chlorides. In the hydrochloric acid solution, arsenic, antimony, and copper are precipitated, as sulphides, by pass- ing hydrogen sulphide into the hot solution till it cools ; the sulphides are digested with sodium sulphide, and the solution treated as directed p. 140. The residue of copper sulphide is dissolved in nitric acid, and the copper separated by elec- trolysis (p. 89). The nitrate from the hydrogen sulphide precipitate is freed from hydrogen sulphide and hydro- chloric acid, and iron and cobalt are separated as directed p. 115. If the mineral is heated with anhydrous sodium thio- sulphate, and exhausted with water, antimony and arsenic go into solution, and are determined as directed p. MO. The sulphides insoluble in water are dissolved in nitric acid, and copper first precipitated (p. 89) ; the nitrates are then converted into chlorides, and cobalt and iron deter- mined (p. 115). Finally, arsenic and antimony may also be determined by removing the arsenic first. The nitric acid solution is heated with sulphuric acid to convert nitrates into sulphates. The arsenic is driven off from this, as chloride, by treatment with ferrous chloride or sulphate, and distillation in a stream of hydrochloric acid (p. 142). To determine antimony, the residue in the flask is saturated with hydrogen sulphide, and filtered ; the precipitate is washed, and treated with sodium sulphide (p. 143). COBALTITEROUS ARSENOPYRITB. Cobalt, Iron, Arsenic, and Sulphur. The mineral is analyzed in the same manner as co- baltite. CERUSSITE. GALENA. 179 CERUSSITR Lead, Iron, Calcium, Carbonic Acid. The pulverized mineral is dissolved by heating with nitric acid, and the lead determined, as peroxide, by connecting the platinum dish with the positive pole of the battery 9) ; the precipi- tated sulphides of arsenic, antimony, and copper are filtered off, thoroughly washed, and digested with sodium sulphide. The solution is treated like one containing polysulphides, and arsenic and antimony separated as directed p. 140. Determination of Phosphorus. About 2 gms. of iron is digested with nitric acid, sp. gr. 1.2, till decomposition is complete. If a carbonaceous residue is left, the nitric acid solution is poured off, and the residue heated with aqua regia. Nitric acid and aqua regia are completely removed by evaporation to dryness, and the nitrates converted into chlorides by repeatedly moistening with concentrated hydrochloric acid, and evaporating to dryness. The residue is treated with water, heated, and the iron brought into solution by the addition of the least possible quantity of hydrochloric acid. To convert the iron, 'etc., into double oxalates, six or eight times the weight of the iron, reckoned as oxide, of a mixture of 1 part potassium oxalate and 5-6 parts ammonium oxalate, is dissolved by 198 QUANTITATIVE ANALYSIS -BY ELECTROLYSIS. heating in the solution, it is diluted to 250-300 cc., and electrolyzed at a temperature of about 80. The heat is maintained during; the reaction; the solution must by no means be heated to boiling, lest the iron scale off. The solution is poured off when the reduction is complete, and phosphoric acid determined as directed p. 127. Two grams of iron are enough for the determination of phosphorus, even when the percentage is small. If a larger quantity is taken, it is best to divide the solution, after conversion into oxalates, and precipitate in several dishes. If not more than 2 gms. iron are present, it is all separated in two hours by a current of 20 cc. oxyhydrogen gas per minute, while the reduction of 4 gms. requires seven hours because of its poor conductivity. As it is not necessary to determine the iron, it may be precipitated just as well in a beaker ; in this case, the negative electrode is a large piece of light platinum foil which is attached by a platinum wire to the conductor from the zinc of the battery. Determination of Sulphur. About 2 grams of iron is oxidized, with aqua regia, to convert sulphur into sulphuric acid, and the insoluble resi- due filtered off. As a portion of the sulphur may be left in the residue, it is fused with a small quantity of a mixture of sodium carbonate and potassium nitrate, the fused mass dissolved in hydrochloric acid, and the solution thus obtained added to the other. The aqua regia is removed, the nitrates converted into chlorides, and the latter into double oxalates, as already directed. After removing the iron by electrolysis, the solution is poured off, boiled to remove ammonia, acidified with hydrochloric acid, and the sulphuric acid precipitated with barium chloride. TABLES FOR CALCULATION OF ANALYSES. 199 TABLES FOR CALCULATION OF ANALYSES. Atomic Weight. Found. Sought. Factor. Aluminium . 27.04 A1 2 3 Al 0.5304 Antimony 120.29 Sb Sb 2 8 1.19902 Sb 2 S 3 1.39879 Arsenic . . . 74.9 As As 2 O 8 1.31962 As 2 5 1.53271 As 2 S 8 1.64192 Barium . . . 136.86 BaSO 4 Ba 0.58819 BaOO 8 Ba 0.69574 BaO 0.77688 Beryllium . . 9.08 BeO Be 0.36262 Bisntuth . 207.5 Bi Bi 2 8 1.11538 Boron .... 10.9 KBF 4 B 0.08639 BA 0.27613 Bromine . 79.7.6 AgBr Br 0.42556 Cadmium . . . 111.7 Cd CdO 1.14288 CdS 1.28630 Caesium . 132.7 Calcium . . . 39.91 CaO Ca 0-.71433 CaCO 3 Ca 0.40006 CaO 0.56004 Carbon . 11.97 CO 2 C 0.272727 Ca00 3 C0 2 0.43995 Cerium 141.2 Chlorine . 35.37 AgCl Cl 0.24729 Ag Cl 0.32853 Chromium 52.45 Cr 2 8 Cr 0.81419 CrO 8 1.18581 200 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. Atomic Weight. Found. Sought. Factor Cobalt . . . 58.60 Co CoO L27116 Copper . . . 63.13 Cu CuO 1.25261 CuS 1,25309 Didymium 145.0 Erbium . . . 166.0 Fluorine . 19.06 CaF 2 F 0.48853 Gold .... 196.2 Au Au 2 O 8 1.12202 Hydrogen . . 1 H 2 H 0,11136 Iodine. . . . 126.54 Agl I 0.54031 Ag I 1.17546 Iron .... 55.88 Fe FeO 1.28561 Fe 2 3 1.42842 Lanthanum . . 138.5 Lead .... 206.39 PbO 2 Pb 0.86605 PbO 0.93303 PbCl 2 1.16289 Lithium . . . 7.01 LiCl Li 0.165408 Li 2 0.35370 Li 8 PO 4 Li 0.18156 Li 2 O 0.38824 LiCl 1.09764 Magnesium . . 23.94 Mg,PA Mg 0.21614 MgO 0.36024 Manganese . . 54.8 Mn 8 O 4 Mn 0.72029 MnO 0.93007 Mn 2 8 1.03496 Mn0 2 Mn 0.63192 MuO 0.81596 Mn 2 O 8 0.90798 MnS0 4 Mn 0.36383 MnO 0.46979 Mn 2 O 8 0.52277 TABLES FOR CALCULATION OF ANALYSES. 201 Atomic Weight. Found. Sought. Factor. Mercury . . . 199.8 Hg Hg,0 1.03994 HgO 1.07988 HgCl 1.17703 Hg 2 S 1.08003 HgS 1.16006 Molybdenum 95.9 MoS 3 Mo 0.49989 Nickel .... 58.6 Ni NiO 1.27116 Niobium . 93.7 Nitrogen . . . 14.01 Pt N 0.14411 NH 8 0.17497 NH 4 0.18526 Osmium . 195 Palladium 106.2 Phosphorus . 30.96 Mg 2 P 2 O 7 P 0.27952 P 2 Q 5 0.63976 Platinum . . . 194.43 Pt PtO 2 1.16417 Potassium 39.03 Pt K 0.40129 K 2 0.48848 KC1 0.76495 K 2 S0 4 0.89389 Rhodium . . . 104.1 Rubidium . . 85.2 Ruthenium . . 103.5 Selenium . 78.87 Silicon 28 SiO 2 Si 0.46729 Silver .... 107.66 Ag Ag 2 1.07412 AgCl 1.32853 Sodium 22.99 NaCl Na 0.39393 Na 2 0.53067 Na 2 SO 4 1.21488 Strontium 87.3 SrS0 4 Sr 0.47673 SrO 0.56389 202 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. Atomic Weight. Found. Sought. Factor. Sulphur . A , 31.98 BaSO 4 s 0.13744 ! "'? S0 8 0.34322 ->*' '-'./' SO 4 0.41181 Tantalum" . . 182 . Tellurium . . 127.7 Thallium . . . 203.7 T1 2 Tl 0.9623 Thorium . 231.96 Tin .... 117.35 Sn SnO 2 1.27201 Titanium . . . 50.25 TiO 2 Ti 0.61154 Tungsten . 183.6 W0 3 W 0.79316 Uranium . . . 239.8 UO 2 U 0.88249 U 8 O 8 1.03916 Vanadium 51.1 V 2 O 5 V 0.56154 Yttrium . . . 89.6 Zinc .... 64.88 Zn ZuO 1.24599 ZnS 1.49291 Zircon . . . 90.4 Zr0 2 Zr 0.73904 REAGENTS. 203 REAGENTS. POTASSIUM OXALATB. The crystallized potassium oxalate of commerce always contains de terminable quantities of iron and lead. To purify it, one part of the salt is dissolved in three parts of water in a porcelain dish, and ammonium sulphide is added drop by drop, as long as a precipitate forms. The solution is now heated on the water-bath till the precipitate settles, and filtered through a plaited filter. To decompose the slight excess of ammonium sulphide, a current of air is conducted through the solution till it is perfectly colorless, and no longer gives a reaction with sodium nitroprusside. The separated sulphur is allowed to settle, and the clear solution siphoned off. AMMONIUM OXALATE. The same impurities are present as in potassium oxalate. The salt is purified by precipitating the hot saturated solution with ammonium sulphide. It is heated over a naked flame till the precipitate coheres together, and filtered hot by the use of a hot-water funnel. The greater part of the ammonium oxalate crystallizes from the filtrate on cooling. The solution is poured off, and the crystals dried by placing them in a funnel stopped with asbestos, and connecting with a filter- pump. 204 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. OXALIC ACID. The impurities are similar to those of the alkali oxalates ; it is purified by repeated recrystallization. AMMONIUM SULPHATE. This salt is purified in the same way as ammonium oxalate. SODIUM SULPHIDE. The crystallized sodium sulphide of commerce is not only exceedingly impure, but is not inonosulphide at all, but a mixture of polysulphides and sodium Irydroxide. The presence of the latter explains that of alumina, which is always found in abundance. If commercial sodium sulphide is used, its solution must first be completely saturated, without access of air, with hydrogen sulphide gas. It is better, however, to prepare the substance, in which case the process is as follows : Sodium hydroxide purified by alcohol is dissolved in water to a solution of 1.35 sp. gr. The solution is divided into two equal parts, and one half, with exclusion of air, saturated with the purest possible hydrogen sulphide gas till the volume ceases to increase. The hydrogen sul- phide is purified by passing it through a wash-bottle of water, and several tubes filled with cotton or wadding. When completely saturated, the solution is filtered from the pre- cipitate formed, and mixed with the other half of the sodium hydroxide solution. Hydrogen sulphide is again passed into the mixture, with exclusion of air, and it is filtered again. The nearly colorless filtrate is evaporated in a capacious platinum or porcelain dish, over a strong free flame as quickly as possible. It boils without bumping if a platinum spiral is REAGENTS. 205 placed in it. As soon as a thin crystalline pellicle forms on the surface, the boiling is stopped, and the solution poured while hot into small flasks with well-ground glass stoppers which must be filled full. It is best to completely exclude the air by melted paraffine. For the separation of antimony and tin, the solution should have a sp. gr. of 1.22-1.225. ALCOHOL. The alcohol used for washing metals must be free from acid, and, as nearly as possible, absolute. It is left standing in a large flask, for twelve hours, over quicklime, and then distilled off on a water or steam bath. The distillate must leave no residue on evaporation. 206 QUANTITATIVE ANALYSIS BY ELECTROLYSIS. ANALYTICAL RESULTS. There follow some of the most important of the many determinations which have been made in the author's labora- tory within the past few years, many of them by students. Determination of Iron. Solutions of the chloride and of the sulphate were employed. The determination of iron in crystallized iron ammonium sulphate gave 14.21, 14.22, 14.25, and 14.23 ft iron. Calculated 14.28 J6. Determination of Cobalt. Pure metallic cobalt (prepared from cobalt oxalate) was dissolved in hydrochloric and in sulphuric acid, converted into the double oxalate, and electrolyzed. Results were, 99.98, 99.99, 99.97, 99.99 . To test the method given on p. 112, weighed quantities of antimony tersulphide were heated with anhydrous sodium ANALYTICAL RESULTS. 209 thio sulphate, and the antimony determined electrolytically alter oxidation of the polysulphide with hydrogen peroxide. Results were, 71.40, 71.57, 71.54, 71.57, and 71.64 88 26.61 39.82 33.57 26.59 39.71 34^09 19.80 30.49 49.71 19.77 30.15 49-76 1 - (Ludwig.) CHEMISTRY. QUALITATIVE QUANTITATIVE OEGANIC INOKGANIC, ETC. A MANUAL OP QUALITATIVE CHEMICAL ANAL- YSIS. By C. R. Fresenius. Translated into the New System, and newly edited by Samuel W. Johnson, M.A., Prof, of Theo- retical and Agricultural Chemistry in the Sheffield Scientific School of Yale College, New Haven. New edition, revised and enlarged with a beautifully colored frontispiece showing spectra. Eighth edition Svo, cloth, $4 00 A bHORT COURSE IN QUALITATIVE CHEMICAL ANALYSIS. With the new notation, by Prof. I. M. Crafts. Revised with additions. By Prof. Chas. A. Schaeffer, of Cornell Institute. Sixth edition 12mo, cloth, 1 50 A SYSTEM OP INSTRUCTION IN QUANTITATIVE CHEMICAL ANALYSIS From the last English and German editions. Edited by Prof. O. D. Allen, of the Sheffield Scientific School of Yale College, with the co-operation of Prof . Sam'l W. Johnson, of the same. This is an entirely new edition from new plates, the New Notation and Nomenclature being employed throughout. QQO pages Tenth edition .................... ....... 8vo,. Cloth, ^ w._ QUANTITATIVE CHEMICAL ANALYSIS. (* \ By T. E.Thorpe, Prof, of Chemistry, Glasgow. Eighth^ ' : edition ..................................... ... 18mo, cloth, QUANTITATIVE ANALYSIS. !;.: The Student's Guide in Quantitative Analysis, designed as.>an X - aid to Fresenius' larger work. By H. Carrington : Bolton. Ph.JX '-/ With many original methods by the author. Third edition* p 1 50 QUANTITATIVE CHEMICAL ANALYSIS TRO LYSIS. According to Original Methods, by Dr. Alexander Classen, of the Koyal School of Technology at Aachen. Authorized translation from the second revised and enlarged German edition, by Prof. Wm. Hale Herrick, of Penna. State College. Svo, cloth, 2 60 AN ELEMENTARY TEXT-BOOK OP CHEMISTRY. By William G. Mixter, Professor of Chemistry, Sheffield Scientific School, Yale College, New Haven. Third revised edition ........................................ 12mo, cloth, 1 50 A HAND-BOOK FOR SUGAR MANUFACTURERS AND THEIR CHEMISTS. Containing Practical Instruction in Sugar-House Control, the Diffusion Process, Selected Methods of Analysis, Reference Tables, etc., etc. Bv Guilford L. Spencer, A.C., of the TJ. S. Department of Agriculture ............ 12mo, morocco flaps, 2 00 SUGAR ANALYSIS, ETC. A Hum! -Book in Schools of Chemical Technology, Etc. By F. C. \Vkvhiiiann. School of Mines .............. ' Svo cloth, 2 50 BOCKS AND SOILS. Their Origin, Composition, and Characteristics ; Chemical, Geological, and Agricultural. By Horace Edward Stockbridge, Ph.D., Professor of Chemistry and Geology in the Imperial College of Agriculture, Sapporo, Japan. Chemist to the Hokkaido Cho. With 13 full page plates .8vo, cloth, $2 50 A LABORATORY GUIDE IN CHEMICAL ANALYSIS. Third edition, entirely rewritten and revised by Prof. David O'Brine, of Colorado State Agricultural College. . . .8vo, cloth, 2 00 INTRODUCTION TO ORGANIC CHEMISTRY. Translated from the German of A. Pinner, Prof, of Chemistry at Berlin, with additions and revisions by Prof. P. T. Austen, of Rutgers College. A Text -book for the use of Colleges and High Schools. Second edition 12mo, cloth, 2 50 TEXT-BOOK OF INORGANIC CHEMISTRY. A short Text-book of Inorganic Chemistry, by Dr. Hermann Kolbe, Prof, in the University of Leipzig. Translated and edited by Prof. T. S. Humpidge, of the University of Wales. With a colored Spectra and numerous wood engravings. (Over 600 pages) 12mo, cloth, 1 50 CHEMICAL LECTURE NOTES. Containing Notes and Observations on Topics that Often Give the Student More or Less Trouble. By Prof. P. T. Austen. 12mo, cloth, 1 00 CHEMICAL REACTIONS, Being a thorough discussion of Chemical Reactions, which occur in the ordinary course of Qualitative Analysis and Gen- eral Elementary Chemistry. By E. Drechsel. Translated, and specially adapted for the use of American students and teachers, by N. F.Merrill, Ph.D. Third edition, 12mo, cloth, 125 LABORATORY CALCULATIONS. And Specific Gravity Tables. By John S. Adriance, A.B., Fel- low of the Chemical Society. Second edition. . . .12mo, cloth, 1 25 WATER SUPPLY. Considered mainly from a Chemical and Sanitary standpoint. By Prof. Wm. Ripley Nichols. With numerous plates. Third edition 8vo, cloth, 2 50 A TEXT-BOOK OP PHYSIOLOGICAL CHEMISTRY. Ly Olof Hammarsten, Professor of Medical and Physiological Chemistry in the University of Upsala. Authorize'd Transla- tion from the second Swedish edition, and from the author's enlarged and revised German edition. By John A. Mandel, Assistant to the Chair of Chemistry, etc., in the Bellevue Hos- pital Medical College and in the College of the City of New York. First edition 8vo, cloth, 4 00 LECTURE-NOTES ON THEORETICAL CHEMISTRY. By Ferdinand G. Wiechmann, Ph.D., Instructor in Chemical Physics and Chemical Philosophy, School of Mines, Columbia College. First edition 12mo, cloth, 2 50 PUBLISHED AND FOR SALE BY JOHN WILEY & SONS, 53 E. 1 Oth St., New York. *** Mailed and prepaid on the receipt of the price. ' UNIVERSITY OF CALIFORNIA LIBRARY THIS BOOK IS DUE ON THE LAST DATE STAMPED '"BELOW JAf 30m-6,'14