K A TEXT-BOOK OP INORGANIC CHEMISTRY. BY DR. A. F. HOLLEMAN, Professor Ordinarins in the University of Amsterdam; Emeritus Profess&i Ordinarius in the University of Groningen, Netherlands, and Fellow of the Royal Academy of Sciences, Amsterdam. ISSUED IN ENGLISH IN COOPERATION WITH HERMON CHARLES COOPER. FOURTH ENGLISH EDITION, COMPLETELY MWISED. TOTAL ISSUE, SIXTEEK THOUSAND. NEW YORK JOHN WILEY & SONS, LONDON: CHAPMAN & HALL, 1913 x Copyright, 1901, 1902, 1905, 1908, 1911, BY HERMON C. COOPER. First, Second, and Third Editions entered at Stationers' Hall. lUE SCIENTIFIC PRESS ROBERT DRUMMOND AND COMPANY BROOKLYN, N. Y. PKEFACE TO THE FOUETH EDITION. THE present edition represents a thorough revision of the work by the Dutch author and the American collaborator. It profits by the author's experience with the frequent editions in other languages but is independent in composition. Very many of the descriptive portions have been rewritten, notably those on the sulphur oxides and acids, rare gases, nitrogen oxides and acids, sodium hydroxide and carbonate, radio-active elements and platinum, as well as the sections on thermo- chemistry, colloids and the iron-carbon system, while the sub- jects of stability and the reality of molecules and atoms furnish new material. The chapter on metal-ammonia compounds is reprinted as approved by Professor WERNER for the third edition. Notwithstanding the appearance of differential formulae in the book, it is believed that a student who is unfamiliar with the calculus should have little difficulty in understanding the meaning and use of such formulas, provided he is willing to take the author's word for the solutions of the equations. Independent students may well be cautioned against regard- ing any text-book as infallible. Even in a book with a world market, such as this one enjoys, undergoing many revisions by the author and by collaborators in other nations, and being frequently reviewed critically by the journals, there will, probably, always be some textual errors and some passages whose lucidity could be improved. Readers can therefore render great service by reporting all unsatisfactory passages to t|j| publishers. Thanks are herewith expressed to my colleague, Professor H. MONMOUTH SMITH, for constant aid in detect ing: errors. References in the text to " ORG. CHEM." refer to the companion volume of this work, HOLLEMAN'S " Text-book of Organic Chem- istry," translated by WALKER and MOTT. H. C. COOPER. SYRACUSE UNIVERSITY, October, 1911. CONTENTS. \ Light-face figures refer to pages; heavy -face figures to paragraphs. PAGE INTRODUCTION (1-5) 1 PHYSICAL AND CHEMICAL PHENOMENA (6) 3 CHEMICAL OPERATIONS (7) 5 THE ELEMENTS (8) 7 Oxygen (9, 10) 9 Law of HENRY, 11; Oxidation, 12; Analytic and synthetic methods, 13. Hydrogen (11-13) 13 Oxyhydrogen blowpipe, 15; Detonating-gas, 15; Reduction, 16. THE CONSERVATION OF MATTER (14) 16 Water (15-19) 17 Physical properties, 20; Natural water, 20; Composition of water, 22. COMPOUNDS AND MIXTURES (20) 25 PHENOMENA ACCOMPANYING THE FORMATION OR DECOMPOSITION OF A COMPOUND, 27. EXPLANATION OF THE CONSTANT COMPOSITION OF COMPOUNDS; ATOMIC THEORY (21-23) 27 Law of constant composition, 27; Atoms, 28; Molecules, 28; Law of multiple proportions, 29; THE ATOMIC WEIGHTS OF THE ELEMENTS, 29; CHEMICAL SYMBOLS AND FORMULAE, 30. STOICHIOMETRICAL CALCULATIONS (24) . . . 31 Chlorine (25-35) 33 Catalytic action, 34; Hydrogen chloride, 37; Acids, bases and salts, 39; Composition of hydrochloric acid, 41; Law of GAY-LUSSAC, 43; AVOGADRO'S hypothesis, 44; Molecular weights, 45; RULES FOR DETERMINING MOLECULAR AND ATOMIC WEIGHTS, 47; Kinetic theory, 47; General gas equation, 48; THE REALITY OF MOLECULES AND ATOMS AND THEIR ABSOLUTE WEIGHT, 49. Ozone (36, 37) $. 50 Formula of, 52; Allotropism, 53. Hydrogen Peroxide (38, 39) .* 53 Status nascendi, 54. MOLECULAR WEIGHT FROM THE MEASUREMENT OF THE DEPRESSION OF THE FREEZING-POINT AND ELEVATION OF THE BOILING-POINT (40-43) . 57 Semi-permeable membranes, 57; Osmotic pressure, 58; PFEFFER'S experiments, 59; Isotonic solutions, 62; Formula of hydrogen peroxide, 65. v vi CONTENTS. Bromine (44, 45) 65 Hydrogen bromide, 67. Iodine (46-48) 69 Hydrogen iodide, 71. DISSOCIATION (49-51) 72 Reversible reactions, 73; Equilibrium, 73; Reaction velocity, 75; Law of chemical mass action, 75; Unimolecular and bimolecular reactions, 76. Fluorine (52, 53) 79 Hydrogen fluoride, 81. Compounds of the halogens (54-62): with each other, 83; with oxygen, 83. NOMENCLATURE (63), 92. Summary of the halogen group (64), 93. . ELECTROLYTIC DISSOCIATION (65, 66) 94 Ionic equilibrium, 98; Strength of acids and bases, 99; Hydrolysis, 101. Sulphur (67-93) 102 Allotropic modifications, 104; THE TRANSITION POINT, 106; "STA- BLE," "METASTABLE," AND "LABILE," 108; THE PHASE RULE OF GIBBS, 109; Hydrogen sulphide, 115; Solubility product, 119; Hydrogen persulphide, 120; Compounds of sulphur with the hal- ogens, 121; VALENCE, 122; Compounds of sulphur with oxygen, 124; Oxygen acids of sulphur, 131; Volumetric analysis, 146. Selenium and Tellurium (94, 95) 148 Selenium, 148; Tellurium, 150. Summary of the oxygen group, (96) 151. THERMOCHEMISTRY (97-104) 152 Law of HESS, 153: CHEMICAL AFFINITY, 156; THE DISPLACEMENT OF EQUILIBRIUM, 160; PASSIVE RESISTANCES, 161. Nitrogen (105-130) 162 The atmosphere, 165; Argon, helium and companion elements, 170; Compounds of nitrogen and hydrogen, 174; Compounds with the halogens, 179; Hydroxylamine, 180; Compounds with oxygen, 181; Oxygen acids, 187; Derivatives of the nitrogen acids, 195; Other nitrogen compounds, 198. Phosphorus (131-154) . . 199 Hydrogen compounds, 204; Halogen compounds, 209; Oxygen com- pounds, 211; Acids, 212. Arsenic (155-164) 221 Hydrogen arsenide, 223; Halogen compounds, 226; Oxygen com- pounds, 226; Oxy-acids, 228; Sulphur compounds, 229; Sulpho- salts, 230. Antimony (165-169) 231 Hydrogen antimonide, 232; Halogen compounds, 233; Oxygen com- pounds, 234; Sulphur compounds, 236. CONTENTS. vii PAGE Bismuth (170-174) 236 Summary of the nitrogen group (175), 239. Carbon (176-189) 241 Allotropic forms, 241; Molecular and atomic weight, 246; Com- pounds with hydrogen, 248; Compounds with oxygen, 249; Other carbon compounds, 255; The flame, 256. Silicon (190-196) 261 Hydrogen silicide, 262; Halogen compounds, 263; Oxygen com- pounds, 265; Silicic acids, 266; COLLOIDS, 268. Germanium (197) 273 Tin (198-202) 274 Stannous compounds, 276; Stannic compounds, 279. Lead (203-206) 281 Oxides, 283; Halogen compounds, 285; Other lead salts, 286; Sum- mary of the carbon group (207), 287. METHODS OF DETERMINING ATOMIC WEIGHTS (208-212) 288 Law of DULONG and PETIT, 289; Law of NEUMANN, 291; Law of MITSCHERLICH, 292; EXPERIMENTAL DETERMINATION OF EQUIVA- LENT WEIGHTS, 293. THE PERIODIC SYSTEM OF THE ELEMENTS (213-221) 296 Construction of a system of the elements, 303; Ascertaining atomic weights, 305; Prediction of properties of elements, 307; Correct- ing atomic weights, 308; Graphic representation, 308. Lithium and Sodium (222-226) 310 Lithium, 310; Sodium, 311; Oxides and hydroxides of, 312; Salts of, 315; (Soda, 319). Potassium (227-231) 322 Oxygen compounds, 323; Salts, 324. Rubidium and Ccesium (232) '. 329 Summary of the group of alkali metals (233), 330. Ammonium Salts (234) 331 SALT SOLUTIONS (235-239) 335 ACIDIMETRY AND ALKALIMETRY (240, 241) 350 Indicators, 352. Copper (242-244) 353 Cuprous compounds, 355; Cupric compounds, 358. Silver (245-247) ' 359 Compounds, 362; Photography, 364. Gold (248-251) 367 Testing of gold and silver, 369; Aurous compounds, 369; Auric com- pounds, 370; Summary of the group (252), 371. Beryllium and Magnesium (253-255) 372 Beryllium, 372; Magnesium, 374; Magnesium salts, 375. Calcium, Strontium, and Barium (256-262) 376 Calcium, 376; Oxides and hydroxides of, 377; Salts of, 379; Glass, v iii CONTENTS. 384; Strontium, 386; Barium, 387; Summary of the group of the alkaline earths, 388. SPECTROSCOPY (263-265) 389 THE UNITY OF MATTER (266) 395 Radio-active Elements (267) 397 Zinc (268, 269) 407 Cadmium (270) 410 Mercury (271-274) 410 Amalgams, 412; Mercurous compounds, 413; Mercuric compounds, 414; Summary of the zinc group (275), 417. ELECTROCHEMISTRY (276-281) 418 Boron (282, 283) 432 Halogen compounds, 433; Oxygen compounds, 433. Aluminium (284-287) 436 Compounds of, 437. Gallium, Indium, and Thallium (287, 288), 441. Summary of the group (289), 442. The Rare Earths (290) 443 Titanium, Zirconium, and Thorium (291) 446 Vanadium, Niobium, and Tantalum (292) 448 Chromium Group (293-299) 449 Chromium, 449; Chromous compounds, 450; Chromic compounds, 450; Chromates, 452; Molybdenum, 455; Tungsten, 456; Ura- nium, 457; Summary of the group, 458. Manganese (300, 301) 458 Manganic acid and permanganic acid, 460. Iron (302-308) 463 Iron-carbon system, 466; Ferrous compounds, 473; Ferric com- pounds, 474. Cobalt and Nickel (309-312) 477 Cobalt, 477; Nickel, 479. Platinum Metals (313-316) 481 Ruthenium, 482; Osmium, 483; Rhodium, 483; Iridium, 484; Palladium, 484; Platinum, 485. Metal-ammonia Compounds. WERNER'S EXTENSIONS OF THE NOTION OF VALENCE (317-318) .486 INORGANIC CHEMISTRY. INTRODUCTION. 1. Chemistry is a branch of the natural sciences, the sciences which deal with the things on the earth and in the outside universe. The knowledge of these things is obtained by observation with our senses, this being the only means we possess. It is well to understand, therefore, that we know not the things themselves, but simply the impressions which they make upon our sense-organs. When we see an object, we perceive, in reality, only the effect on our retina; if we feel the object, it is not the body itself but the excitement of the sensory nerves of touch in our fingers that we are made aware of. Hence it may be fairly asked whether the objects of which we are cognizant are really just as we perceive them, or whether they even exist at all outside of our person. The natural sciences leave this problem out of consideration its solution is the task of speculative philosophy. In reality they are not concerned with the objects, which in themselves we cannot know, but with the study of the sensations that we receive. The sensations take the place of the objects, and we regard them as such. 2. The Scientific Investigation of Things. What is to be understood by the term? In the first place, a most accurate description of the objects. From a study of this it is found that many objects resemble each other to a greater or less degree, and it is therefore possible to make a classification, i.e. an arrangement of like objects into groups and a separation of the various groups from each other. By the descriptive method we are finally able to divide the natural sciences into Zoology, Botany, Mineralogy and Astronomy. 3. In the second place, scientific investigation includes the 2 INORG AttlC CHEMISTRY. [ 3- study of the relations which the objects bear to each other; in other words, the study of phenomena. The heavenly bodies move towards each other; water turns to ice on cooling; wood burns when heated. It is the task of the natural sciences to accurately observe and describe such phenomena, i.e. to ascertain in what way the heavenly bodies change their relative positions, what conditions affect the freezing of water, what becomes of the burn- ing wood, under what conditions it can burn, etc. The description of the phenomena leads to a different division of the natural sciences than the description of objects, viz., a divi- sion into Physics, Chemistry and Biology, the latter being the study of vital processes, and including Physiology, Pathology and Thera- peutics. 4. The human mind, in pursuing the scientific study of nature, does not feel contented with the accurate description of objects and phenomena; it seeks also for an explanation of the latter. The various attempts at explanations constitute the most im- portant part of science. When, for instance, we see that a ray of light in passing through a piece of Iceland spar is split up into two other rays of different properties, we strive to account for the phenomenon. When copper is heated in the air, it turns into a black powder; the question again arises, why this thing is so. In, searching for an explanation of the phenomena we thus endeavor to penetrate deeper into the essence of things than is possible by direct observation. Although the phenomena themselves are found to be unchangeable, our explanation of them may be modi-' fied as our knowledge increases. The transformation of copper into a black powder on heating in the air was formerly explained by the supposition that something left the metal; subsequently, when the phenomenon was better understood, by assuming that the copper takes up something from the air. Scientific investigation pursues in general, then, the following course: A phenomenon is observed and studied as carefully as possible. Thereupon an explanation of it is sought. A hypoth- esis is set up. From this conclusions can be formed, some of which can be tested by experiment. If the latter really leads to the expected result, the hypothesis gains in probability. If it is subsequently found to explain and link together a whole serie of phenomena, it becomes a theory. 6.] PHYSICAL AND CHEMICAL PHENOMENA. The nineteenth century was an era of great prosperity for scientific inquiry. For numerous phenomena explanations have been found which possess a great degree of probability. Still it cannot be denied that the present theories penetrate only a little into the real essence of things, and the investigator very soon stumbles upon questions whose explanation does not at present even seem to be a possibility. The chemical process that goes on when copper to retain our former example is heated in the air is well known. However, the deeper question, why the action takes place just so and not otherwise, or why the resulting powder is black, still awaits a satisfactory answer. 5. We observed in the preceding paragraph that the natural phenomena are found to be unchangeable. The movement of the planets, for example, still takes place in the same manner as in the times of the Ptolemies; whenever water turns to ice the same increase of volume is to be observed; the crystal form of common salt, whenever and wherever examined, is invariably the same; from the burning of wood the same products are always obtained ; the microscopic structure of the leaves of one and the same plant is never found to vary. This general principle finds its expression in the phrase, constancy of natural phenomena. Every one is con- vinced of its truth, and it is tacitly accepted as the basis of every natural scientific investigation. If, for example, one has measured the angles which the faces of a soda crystal form with each other, he considers it certain that all soda crystals must show the same angles, at whatever time or place they may be measured. If it has once been determined that pure alcohol boils at 78 under normal pressure, it is forthwith assumed that this must be the case with all alcohol, no matter how it may be obtained or when and where it may be tested. PHYSICAL AND CHEMICAL PHENOMENA. 6. It was stated above (3) that the description of phenomena leads to a division of the natural sciences into Physics, Chemistry and the study of vital processes (Biology) . In defining the province of Chemistry Biology may be left out of consideration; however, it is desirable to compare the field of Chemistry with that of Physics. In general it may be said that Physics deals with the 4 INORGANIC CHEMISTRY. [ 6- temporary, Chemistry with the lasting, changes of matter. By matter or substance we understand the objects without reference to their form. Iron, marble, sand and glass are kinds of matter, or substances, independent of their external shape. A couple of illustrations may make this conception of temporary and lasting changes clear. (a) A platinum wire glows when held in a colorless gas-flame. On removal it cools off and no change is visible. This is a physical phenomenon; the change, the glowing, is of a temporary sort. So soon as the cause of the change is removed, the wire resumes its original condition. When some magnesium wire is held in the flame, it burns with the emission of a brilliant light and turns into a white powder, which is wholly different from the substance magnesium. Here a lasting change has occurred; we have to do with a chemical phenomenon. (6) Again, we may take two white crystallized substances, naph- thalene and cane-sugar, and heat each separately in a retort with receiver. The naphthalene at first melts; on continued heating it begins to boil, then distils over and condenses in the receiver. The distilled naphthalene resembles the undistilled in every respect. The substance has, as a result of heating, undergone physical changes melting, change to vapor and, finally, return to the solid state. The cane-sugar behaves differently. Here also a melting is observed at first, but soon the sugar turns darker; a brownish liquid distils over; a peculiar odor is noticeable and at last there remains in the retort a charred, porous mass. The cane-sugar suffers a lasting change on being heated. In this case we have a chemical change, (c) As a third and last example we may consider the behavior of a metallic wire on the one hand and that of acidulated water on the other, when an electric current passes through them. The wire displays other properties so long as the current is on. If the latter ceases, the wire returns to its original condition. This is a physical action. In the acidulated water, however, the current induces an evolution of gas, and this gas arising from the water has properties entirely different from those of the water. A lasting change in the substance has occurred; a chemical action has taken place, A sharp distinction between physical and chemical phenomena is often as will be seen later very difficult to make. 7.J CHEMICAL OPERATIONS. 5 CHEMICAL OPERATIONS. 7. In order to avoid repetitions it seems advisable at this point to describe briefly some of the commonest chemical operations. Solution. When sugar, salt or saltpetre, for example, is put into water, the solid substance disappears and its taste is taken on by the water. The substance has dissolved in the water. There is a definite limit to the solubility of each of these, for, if the tem- perature is kept constant and more of the substance is gradually added, a point is finally reached when the water will take up no more. The solution is then saturated. The solubility of most solids increases with the temperature. Moreover it is very differ- ent with different substances, varying all the way from solubility in all proportions to imperceptible solubility. Thus cane-sugar is dis- solved in large quantity by water, while sand is practically insolu- ble in it. Liquids can be either miscible in all proportions (water and alcohol) or only partially soluble in each other. When, for instance, water is shaken with a sufficient quantity of ether and allowed to stand, two liquid layers are formed; the water has dissolved some ether and the ether some water. In most cases the solubility of liquids in each other also increases with the tem- perature. In the case of gases solubility decreases with rising temperature. Separation of a Solid and a Liquid. This may be accomplished by filtration. A funnel is lined on the interior with "filter-paper " and the mixture poured upon it. The solid is retained on the paper while the liquid passes through. Decantation is a less com- plete method of separation, since more liquid remains with the solid by this method than by filtration. However, it is evident that neither method affords a really complete separation. This can only be accomplished by washing, i.e. by replacing the por- tion of the liquid which remains between the solid particles by another liquid. If the liquid of the mixture be a salt solution, pure water is very effective. It is obvious that by repeating the washing several times the salt solution can be wholly removed. Suppose that 1 c.c. salt solution remains between the particles of the solid and that 9 c.c. water is then added. The solution is thus reduced to one-tenth of its original concentration. If 1 c.c. of this dilute solution again remains with the solid and another 6 INORGANIC CHEMISTRY. [ 7- 9 c.c. water is added, the concentration is then 10"~ 2 , or one- hundredth of the original concentration; after six! such operations it would be only 10~ 6 , or one millionth of the original, so that the separation is practically complete. Crystallization. If a solution is saturated in the warm and 'is then allowed to cool, the dissolved substance frequently separates out in the crystallized state. Advantage is often taken of this for purifying crystallizable substances. Distillation (Fig. 1). This operation is frequently made use of in working with liquids. The liquid is placed in a flask or a retort and heated to boiling. The escaping vapor is cooled to FIG. 1. DISTILLATION. a liquid in a condenser. The latter consists of a sufficiently wide tube encased in a jacket, through which water flows to keep the inner tube cold. The condensed liquid is collected in the receiver. It is readily seen how volatile substances can be separated from non-volatile ones by distillation, e.g. water from salt, since the former distil over and the latter remain in the distilling-flask. However, liquids of different volatility can also be separated in this manner. Take, for example, a mixture of alcohol and water. The more volatile constituent, alcohol, passes over for the most part in the early stage of the operation; towards the end the less volatile, water. If the two distillates are collected separately, an approximate separation results. A few repetitions of this so-called fractional distillation bring about a practically complete separation in many cases. 8.] THE ELEMENTS. Sublimation. Certain solids, e.g. camphor, when heated (at ordinary pressure), turn to vapor without melting. If this vapor comes in contact with a cold surface, the substance is deposited in the solid, crystallized state. It is evident that we have here another method of separating some substances. THE ELEMENTS. 8. When a substance ( 6) is subjected to various influences, such as heat, electricity, or light, or is brought in contact with other substances, it is very often split up into two or more dis- similar components. As an example let us take gunpowder. Water is added and the whole is stirred well and gently warmed; after a while it is filtered, and that which remains on the filter is found to be no longer gunpowder, for it is unexplosive. On evapo- rating the filtrate a white crystalline substance, saltpetre, remains. The undissolved part is dried and then shaken with another sol- vent, carbon disulphide. After a time the mixture is filtered, as before, and there is left on the filter a black mass, consisting of charcoal powder. The carbon disulphide of the filtrate evaporates and leaves yellow crystals of sulphur. Thus we see that, by suc- cessive treatment with water and carbon disulphide, gunpowder can be separated into three substances, viz. carbon, sulphur and saltpetre. The two former are incapable, even when subjected to all the agencies at our command, of division into different com- ponents. Not so with saltpetre, for when the latter is heated strongly a gas is given off in which a glowing wooden splinter is at once ignited. When the evolution of gas ceases, a substance remains which gives off red fumes on treatment with sulphuric acid, something that saltpetre does not do. Saltpetre can evidently be broken up still farther by heating. If we subject all sorts of substances to a successive treatment with reagents of the most different kinds, we finally discover cer- tain ones that cannot be resolved into simpler substances by our present means. Such substances are called elements. Although the number of substances, according to 6, may be considered as infinitely great, experience has taught that the number of elements is small. There are about eighty. As our methods of examination improve, it may quite possibly 8 INORGANIC CHEMISTRY. [ 8 be found that the substances which the chemist of to-day regards as elements have no right to the name. Therefore, when we use the word "element/' it is to be regarded as a relative term, dependent on the extent of our knowledge and the means' at our command. In the history of chemistry some cases are to be found where sub- stances, once believed to be elements, were subsequently decom- posed. The exact number of elements cannot be definitely stated, be- cause, on the one hand, not all the substances that possibly exist may be within our reach,* and, on the other hand, it is doubtful whether certain substances now regarded as elements cannot be divided by means already known. On the inside of the back cover will be found a list of the elements now known. As may be seen from this list, the metals are included in the elements. Together with them we find a number of other sub- stances, as oxygen, sulphur, phosphorus, etc., that are classed under the term non-metals, or metalloids. To the latter class belong many very important substances, e.g., oxygen, an element that combines with almost all others, causing what is called combustion. Oxygen is present in a large amount in the air. Another non- metal is carbon, which is present in all organized substances, and is therefore a constituent of every animal and plant. Sulphur, which burns with a blue flame, giving off a pungent odor, and chlorine, a greenish-yellow gas of very disagreeable odor, which combines readily with most metals, are also non-metals. The elements occur in very unequal proportions in the part of the earth accessible to us. Oxygen, which occurs in air, in water, and in the solid part of the earth's crust, is very preponderant, composing approximately 50% of these portions of the earth which have been investigated. The elements silicon, aluminium, iron, calcium, carbon, magnesium, sodium, potassium, and hydrogen, together with oxygen, make up 99% of the earth's crust. There remains, therefore, only 1% for all the other elements. Some of these are quite common, e.g., lithium, but they almost always. * Of the interior of the earth only a very small part is known. If we think of the earth as about the size of an orange, the deepest mine-shafts would not even penetrate the thin yellow exterior layer of the orange skin. 9.] OXYGEN. 9 occur in very small quantities. Others, like niobium and tantalum, are found in relatively very small amounts and in isolated places, With the aid of spectroscopy ( 263-265), it has been ascer- tained that the heavenly bodies contain most of the elements found in our earth, and also some others. OXYGEN. i 9. Under ordinary conditions of temperature and pressure, oxygen is a colorless and odorless gas, whose most noticeable property is its ability to set glowing substances on fire with the evolution of much light and heat. A glowing splinter of wood, for example, when introduced into an atmosphere of oxygen, begins at once to burn brightly. This action is ordinarily used as a characteristic test for the identification of oxygen. This gas can be obtained in various ways. There are many substances which are known to evolve oxygen on heating. (1) Mercuric oxide, when heated strongly in a retort (Fig. 2), yields oxygen, which can be collected by means of a delivery-tube FIG. 2. PREPARATION OF OXYGEN FROM POTASSIUM CHLORATE. opening under the mouth of a cylindrical receiver filled with water. The inside of the retort becomes covered with drops of mercury. (2) The same apparatus can be used in making oxygen from potassium chlorate (chlorate of potash), as well as from potassium nitrate (saltpetre), potassium permanganate, and many other sub- stances. The preparation of oxygen by heating potassium chlorate is a method frequently used in the laboratory. 10 INORGANIC CHEMISTRY. [ 9- Some substances give off oxygen when heated together with others, as in the following cases : (3) Potassium dichromate or manganese dioxide, when heated with sulphuric acid ; (4) Zinc oxide, when heated in a current of chlorine. The atmospheric air consists principally of oxygen and nitrogen. The following method for separating these gases was employed by LAVOISIER in 1774. He introduced some mercury into a retort A (Fig. 3) with a long, doubly-bent neck that opened under a bell- jar P filled with air and resting in a dish R of mercury. He then FIG. 3. ABSORPTION OF OXYGEN BY MERCURY. heated the retort steadily for several days, keeping the mercury almost boiling. As a result, a part of the air in P disappeared, and the gas remaining was found to possess other properties than air it was nitrogen. At the same time the mercury had been partially transformed into a red powder, mercuric oxide. On heating the latter more strongly oxygen was obtained. Oxygen in now prepared from liquid air (cf. 109) . The physical properties of oxygen, besides those already men- tioned, are as follows: Its specific gravity, assuming the density of air to be 1, is 1.10535. A liter of oxygen at and 760 mm. Hg pressure weighs 1.4290 g. Oxygen can be liquefied; the difficulties in obtaining it on a large scale in the liquid state have -now been completely overcome. Apparatuses for lique- fying oxygen have been constructed by HAMPSON and by LINDE, a description of which is to be found in text-books of physics. The critical temperature of oxygen is -118, and its critical pressure 50 atmospheres. Liquid oxygen 10.] OXYGEN. 11 has a specific gravity of 1.124 (based on water) and a boiling-point of - 182.95 at 745.0 mm. pressure. Its color is light blue. It can be preserved for some time at ordinary pres- sure, with the aid of a so-called vacuum-flask (Fig. 4.) The latter is a vessel enclosed in an air-tight jacket, the space be- tween the walls being evacuated. 100 1. water at dissolves 4.89 1. oxygen. The gas is also somewhat soluble in alcohol and in molten silver. When the silver solidifies, the oxygen a volume about ten times that of the metal suddenly escapes from solution, caus- ing peculiar elevations on the surface of the silver ("spitting" of silver). We remarked above ( 7) that the solubility of gases in liquids diminishes with increasing tem- perature. A very remarkable law expresses the relation that exists between the solubility of a gas and its pressure, namely, the solubility is propor- tional to the pressure. This is the law of HENRY. FIG. 4. VACUUM- Thus, when the pressure becomes a-fold, the solu- FLASK. bility also becomes a-fold. As the mass of a gas which is present in a certain volume is likewise proportional to the pressure, the law of HENRY can also be expressed thus : The volume of a gas dissolving in a certain quantity of a liquid is independent of the pressure. This law is rigid when the solubility of the gas is small; when the solubility is large, for instance 100 volumes in 1 of the liquid, its deviations are considerable. Still another formulation of this law is of value in understanding certain of its applications: The concentrations of the dissolved and undissolved portions of a gas bear a constant ratio to each other. By " concentration " is meant the quantity of the gas in grams per unit volume (cubic centimeter). 10. Among the chemical properties of oxygen the most promi- nent is its vigorous support of combustion. The following are interesting examples : Charcoal glows in air only moderately and without much evolu^ tion of light. In oxygen, however, it burns with a bright glow. Sulphur, which burns in air with only a small flame, burns in oxygen with an intense blue light. Phosphorus burns in oxygen 12 INORGANIC CHEMISTRY. [ 10- with a blinding white light. A steel watch-spring that has been heated to redness at one end and put into oxygen, burns with scintillation. Zinc also burns in it with a dazzling light. In all these and analogous cases the oxygen, as well as the burning mate- rial, disappears during the combustion, while new substances are formed. A lasting change therefore takes place and we have to do with a chemical process. The product of burning charcoal is found to be a gas that makes lime-water cloudy and is unable to support combustion; it is called carbonic acid gas. Sulphur also yields a gas; it has a pungent odor and is called sulphur dioxide. Phosphorus produces a white flocculent powder, phosphorus pent- oxide. When iron burns, a black cindery powder is formed, called "hammer-scale," because it composes the sparks that fly from the anvil. The question now arises as to what really occurs in the above cases. In the first place, it has been found that the weight of the product of combustion is greater than that of the substance burned. The increase in weight of the substance during burning can in many cases be easily demonstrated. For instance, a horseshoe magnet that has been dipped in iron filings may be hung on the lower side of a scale-pan and balanced by weights put in the other pan. The iron filings may be burned by passing a non-luminous flame under them a few times. On cooling, the scale-pan attached to the magnet sinks. In a similar way one may demonstrate the increase of weight in the burning of copper. In order to prove the increase of weight in a case where only gaseous products are formed, a candle may be burned and the combustion products, carbon dioxide and water vapor, collected by letting them pass over unslaked lime, with which both unite. Closer investigation has revealed the fact that the increased weight is due to the presence of oxygen, as well as the burned substance, in all combustion products. The latter are compounds of these substances with oxygen. The participation of oxygen in the burning of zinc, for example, may be proved by heating the combustion product, zinc white, in a tube and leading over it chlorine gas, whereby oxygen is driven oft 7 . The compounds of oxygen are called oxides, and the act of this combination is known as oxidation. When substances burn in the air, it is only the oxygen which combines with them. Nevertheless, the nitrogen of the air is heated 11.] HYDROGEN. 13 and thus takes a part of the heat evolved in the combustion. Therefore the temperature of a burning object cannot rise so high as in pure oxygen, and, since the emission of light increases very rapidly as the temperature rises, combustions in oxygen are for this reason much brighter than in air. There are two general methods of ascertaining what elements exist in a compound. According to the one method the compound is decomposed and the elements composing it thereupon deter- mined. This is the analytic method. According to the other, the synthetic method, the composition is found by combining different elements to form new substances. In the above-described experiment ( 8) of LAVOISIER the composition of the red powder is learned by decomposing it at high temperature, whereupon it separates into only mercury and oxygen. Inversely it was possible to obtain the red powder by heating pure oxygen and pure mercury together at a lower temperature. The former is an example of analysis, the latter of synthesis. HYDROGEN. ii. Hydrogen is a colorless and odorless gas that is rarely found on the earth in the free state. The gases of some volcanoes contain it and it can also result from processes of decay. In combination with other elements, however, hydrogen is very widely distributed and occurs in very large amounts (8). Hydrogen can be prepared in various ways. In the first place, hydrogen compounds can be broken up. (1) Water containing some dissolved electrolyte is decomposed by the electric current with the evolution of hydrogen at the negative pole (cathode). The ordinary methods of preparing hydrogen depend on the indirect decomposition of hydrogen compounds, i.e., their reaction with other substances. The following are examples of this sort: (2) The action of zinc on dilute sulphuric acid ( 89). This is the commonest method. For the preparation of hydrogen in the laboratory the apparatus shown in Fig. 5 is often used. A contains granulated zinc (or iron nails) and B dilute hydrochloric acid or sulphuric acid. When the cock C is opened the acid flows through D to the metal and the evolution of hy- drogen commences at once. The cock being closed again, the gas still 14 INORGANIC CHEMISTRY. [ 11- continues to come off and forces back the acid. This is facilitated by- changing the relative levels of A and B. (3) The action of zinc or aluminium fil- ings on caustic potash or slaked lime. (4) The action of sodium or potassium on water or alcohol. (5) Magnesium powder, when boiled with water, also evolves hydrogen, especially when some chloride of magne- FlG * 5t sium is dissolved in the water, because such a solution dissolves the magnesium oxide which forms on the surface of the metal. Likewise, red-hot iron decomposes water with the liberation of hydrogen (compare 305). i 12. The physical properties of hydrogen are these: It is the lightest of all known substances, its specific gravity (air= 1) amounting to only 0.06949. One liter of hydrogen at and 760 mm. Hg. pressure weighs 0.0899 g. Its lightness renders it useful for inflating balloons. It is very hard to liquefy, because its critical temperature lies only 30-32 above the absolute zero (273). On the other hand, the critical pressure is only 15 atmospheres. Liquid hydrogen is colorless. It boils at 252.5. Its specific gravity, with reference to water, is only 0.07 at its boiling-point and 0.086 at its freezing-point, being therefore considerably less than that of all other known liquids. DEWAB further succeeded in bringing hydrogen to the solid state by allowing the liquid to evaporate quickly at 30-40 mm. pressure. The melting-point of solid hydrogen is about 16 (absolute tem- perature). The heat of evaporation of liquid hydrogen is very high, being 200 cal. ; for this reason a flask containing liquid hydrogen soon becomes Covered with a layer of liquid air, which drops down and soon partially solidifies. Hydrogen is slightly soluble in water, 100 1. water dissolving 2.15 1. of the gas at 0. Alcohol takes up somewhat more. 13. Chemical Properties. Hydrogen does not unite with as large a number of elements as oxygen. At a higher temperature 13.] HYDROGEN. 15 it displays a strong tendency to unite with oxygen, burning with an almost colorless and a very hot flame to form water. This property serves for the identification of hydrogen gas. When a current of hydrogen is directed upon very finely divided platinum (spongy platinum or platinum black, 316), the hydrogen is ignited ( 25). The high temperature of the hydrogen flame is made use of in fusing platinum, quartz, etc. Such a flame is known as an oxyhydrogen flame. An apparatus (oxyhydrogen bloivpipe) like that represented in Fig. 6 is required for producing it. The hydrogen enters at W and passes out at a, where it is lit. Oxygen is blown into the flame at S. Thus the gases do not mix till they reach the flame, and the possibility of an explosion is avoided. FIG. 0. OXYHYDROGEN BLOWPIPE. A mixture of hydrogen and oxygen, especially in the proportion of 2 vols. H and 1 vol. O (detonating-gas), when ignited, turns instantaneously to steam; in other words, it explodes. This ex- periment can, however, be performed harmlessly by using a wide- mouthed cylinder of not too great dimensions. A loud report is heard in this case, because the steam at the moment of its forma- tion occupies a much larger volume at the high temperature of the combustion than the mixture of the original gases, and as a result the air is suddenly ejected with violence. When the explosion occurs in a closed vessel, no sound is heard (cf. e.g. Fig. 13, p. 25). The temperature to which detonating-gas must be heated to explode is found to be about 700. At a lower temperature com- bination between hydrogen and oxgyen also takes place, but not instantaneously, as in explosions; the lower the temperature, the slower the process. When, therefore, no change in cold detonating- gas is observed even in the course of several years, we must attribute the fact to the extraordinary slowness of the process at ordinary temperatures. A simple calculation will make this plain. BODENSTEIN observed that, when detonating-gas is heated at 509 for 50 minutes, 0.15 of the whole is changed to water. Now it is a general rule that, when the temperature sinks 10, a chemi- cal reaction becomes about twice as slow; at 499 it would thus 16 INORGANIC CHEMISTRY. [ 13- take 100 minutes till the 0.15 part of the gas had formed water. At the ordinary temperature, say at 9, it would be 50 X2 50 minutes, that is about 1.06X 10 11 years. The same can be said of all chemi- cal reactions. When we see that wood, sulphur, etc., burn quickly at higher temperatures, we must admit that oxidation takes place also at ordinary temperatures, though so slowly that we cannot perceive it. MOISSAN, however, succeeded in proving that charcoal at 100 and sulphur at ordinary temperatures are oxidized very slowly in a current of oxygen. Hydrogen is not only able to unite with free oxygen, but it also has the power to withdraw oxygen from many of its compounds. The action of hydrogen on a compound is called, in general, reduction. This action is often a very useful means of determining whether a compound contains oxygen, since the latter, if present, frill usually unite with the hydrogen to form water. Copper oxide may serve as an example of the application of this method. A little is placed in a tube, hydrogen is led over it, and heat is then applied ; one soon sees the black oxide change to red copper, and water depositing in drops on the colder parts of the tube. Many other oxides can be similarly reduced, e.g. iron oxide, lead oxide, etc. THE CONSERVATION OF MATTER. 14. The quantitative relationships in oxidizing and reducing processes, such as have been discussed in 13, i.e. the relations of the masses of the substances participating in the changes, may be used to elucidate a very important law. A definite amount of copper powder, for example, may be placed in a tube and the weight of the tube with the powder ascertained. Oxygen is then led over the copper at a high temperature. The apparatus should be so arranged that the volume of the oxygen which combines with the copper can be measured. When the oxidation process has proceeded for some time, the tube containing the oxidized copper is allowed to cool and then weighed. The weight is found to have increased, and the increase is just equal to the weight of the volume of oxygen used up. Thereupon hydrogen is passed through the tube with the copper oxide and heat applied. Here also arrange- ments should be made for measuring the volume of hydrogen con- sumed in reduction. The reduction is allowed to go on until all the copper oxide is transformed back to copper. When the tube and powder are subsequently weighed, they will be found to have re- 15.] WATER. 17 assumed their original weight. The water that forms can be absorbed by a substance like quicklime or concentrated sulphuric acid and weighed. It will be found equal in weight to the loss of weight of the copper oxide on changing to copper plus the weight of the consumed hydrogen. In these cases, therefore, the combined weight of the reacting substances before and after the reaction is the same. Copper + consumed oxygen weighs just as much as copper oxide; copper oxide -f consumed hydrogen weighs just as much as copper + water; and, finally, the regained copper weighs just as much as that origi- nally taken. The substances can be changed into different states, but their weight remains unaltered. This phenomenon is observed without exception in chemical actions, and we therefore accept as a law the statement that matter is indestructible, or that no matter can be lost or gained. This principle was introduced into chemistry by LAVOISIER (1743-1794;. The old Greek philosophers were already firmly convinced of the impossibility of producing or destroying matter. In all ages this belief has been the basis of philosophic thought. To LAVOISIER is due the credit of having demonstrated the practical application of the principle of the indestructibility of matter. He assumed that gravity is an inseparable attribute of all matter concerning which a great deal of doubt still existed and that the combined weight of the substances concerned must therefore be the same before and after a chemical reaction. The theory of knowledge teaches that the principle of the indestructi- bility of matter lies originally at the basis of our thinking. It is entirely incorrect to suppose that it was established by experimentation; on the contrary, we test the correctness of our experimental results by ascertain- ing in how far they conform to this principle. This can be easily under- stood in the above case of the oxidation and reduction of copper. In per- forming this experiment one finds that the weight of copper + oxygen is not exactly equal to that of the copper oxide formed. Even after several repetitions slight differences are still found. Because we feel that there must be absolute equality, we attribute these differences to imperfections in our instruments, and we consider our instruments improved if they enable us to approach nearer the complete equality of the weights before and after the experiment. Nevertheless, we are unable to really observe an absolute equality. WATER. 15. Water was regarded as an element for many centuries. Not until 1781 did CAVENDISH discover that, when a mixture of hydro- 18 INORGANIC CHEMISTRY. [15. gen and air or oxygen explodes, water is formed. Being, how- ever, a supporter of an erroneous theory ( 106), he failed to realize the importance of his discovery. LAVOISIER in 1783 repeated this experiment and comprehended it as a synthesis of water, as we still do to-day. With the aid of the apparatus pictured in Fig. 7, this synthesis can be easily demonstrated. The hydrogen is generated in the FIG. 7. COMBUSTION OF HYDROGEN. two-necked (WOULFP) bottle from zinc and sulphuric acid. In order to free the gas from water vapor, it is passed through the horizontal tube, which contains chloride of calcium or bits of pumice-stone soaked in sulphuric acid. The dry gas is ignited and, as it burns, water is gradually deposited on the walls of the bell -jar. A mixture of hydrogen and oxygen unites to form water when illu- minated with ultraviolet light. In addition to this direct synthesis from its elements there are other ways of obtaining water. For example, many compounds, such as the blue crystals of copper vitriol, give off water when heated. The formation of water by the action of hydrogen on oxygen compounds was illustrated ( 13) in the reduction of copper oxide. On the other hand, it is also produced by the action of oxygen on certain hydrogen compounds. This is seen, for example, in the burning of alcohol. Finally, water can result from the reaction of a hydrogen com- pound with one of oxygen. This is the case when ammonia gas ( 111) is led over hot copper oxide. 16. WATER. 19 8 The synthetic methods of preparing water, such as the above- named and many others, possess, however, merely theoretical importance. Even when water is wanted in a perfectly pure state, natural water is resorted to. This contains solids and gases in solution, which must be eliminated. Its purification is accom- plished by distillation. An apparatus well suited to this purpose is shown in Fig. 8. High pressure steam and electricity are often used for heating instead of the flame. Water is placed in the retort A, which rests over the fireplace, and boil- ed. The dissolved gases are first driven off; the hot steam follows, passing t hrough the dome B into the condensing coil (" worm ") C, which is cooled by water in the vessel Z). The condensed water, now pure, flows down into the bottle; the solid PIG> s. PURIFICATION OF WATER BY DISTILLATION. substances that were dissolved in the water remain in the retort. The cooler D is supplied with cold water through a tube, entering near the bottom, while the heated, and therefore specifically lighter, water flows out near the top. The steam thus meets with cooling- water of a lower temperature as it passes down the worm,and is in this way very completely condensed (principle of the counter-current). A single distillation is usually insufficient for the complete elimination of all gaseous and solid constituents. For this purpose the operation must be repeated in an apparatus of platinum (tin is less satisfactory) with a condensing coil of the same metal, and only the middle fraction collected. An excellent criterion for the purity of water is to be found in the measurement of its electrical resistance. Very pure water conducts the electric current scarcely at all. KOHLRAUSCH found the conductivity at 20 INORGANIC CHEMISTRY. [15- 18 of the purest water obtainable to be k = 0.038 X10~ 6 expressed in reciprocal ohms; by this is meant the conductivity of a body a column of which 1 cm. long and 1 cm. square in cross-section has a resistance of 1 ohm. The magnitude of the resistance of such water is better un- derstood by comparing it with resistance of copper. 1 cu. mm. of this water has at the same resistance as a copper wire of the same cross- section and 25 million miles long; it could be strung around the earth's equator one thousand times. The slightest traces of salts or even con- tact with the atmosphere cause a market increase in its conductivity. PHYSICAL PROPERTIES. 1 6. Water at ordinary temperatures is an odorless, tasteless liquid, showing no color in thin layers. On looking through a layer 26 meters thick, SPRING observed a pure dark-blue color. The thermometer-scale of CELSIUS is fixed according to the physical constants of water, its freezing-point being called and its boiling- point at 760 mm. pressure 100. These two points are dependent on the pressure. An increase of pressure lowers the freezing-point (0.0075 per atmosphere). This is the reason why ice melts under high pressure. Water possesses the very uncommon property of having a maximum of density (minimum of volume) at a definite temperature. The volume of almost all other substances increases with rising temperature, but here it diminishes up to 3.945, above which temperature water expands as heating continues. During the transformation of water to ice the volume increases considerably. One vol. water at yields 1.09082 vol. ice of the same temperature. The specific heat of water is greater than that of a vast majority of other substances. Its latent heat of fusion is 79 Cal., its latent heat of vaporization 536 Cal. Water is extensively used as a sol- vent. Numerous substances dissolve in it to a greater or less degree. There are many liquid substances that mix with water in all pro- portions, and many, also, which do not. (See 7.) The remarkable physical properties of water play a very important role in nature; this subject is extensively discussed in physics, meteor- ology, and geology. NATURAL WATER. 17. Water, as it occurs in nature, is by no means chemically pure. It may contain solid matter in suspension as well as sub- stances, either solid or gaseous, in solution. The purest natural water is rain-water. This has really passed through a natural process of distillation, the water on the earth's surface being vapor- 17.] NATURAL WATER. 21 ized by the sun's heat and condensed again by contact with colder portions of air, whereupon it falls in the form of rain. Neverthe- less it contains dust particles (in large cities more, of course, than in the country) and gases from the air, as well as traces of ammo- nium salts. Spring- and well-waters contain in 10,000 parts about 1-20 parts of solid matter, consisting largely of lime salts. Well-water that con- tains much lime is called hard ( 259). Well-water also contains some carbonic acid and air in solution, both of which give it its refreshing taste; distilled water tastes flat. Natural water is used extensively for drinking purposes. When it comes out of a soil that is contaminated by decaying organic matter, as is the case in many large cities, it is injurious to health, principally on account of the presence of bacteria. It can be freed from these by filtration through a PASTEUR-CHAMBERLAND porcelain filter (Fig. 9). This consists essentially of a hollow cylinder of porous porcelain (called a "candle") A, through whose walls the water is forced by its own pres- sure. The lower end of the candle opens into the nozzle. In large cities it has been found much more practicable to purify the well- or river-water at the central station and to pipe it thence to the various houses. Epidemic diseases have really decreased remarkably since the introduction of the methods of modern sanitary science. A water which contains so many substances in solution that it has a definite taste or a therapeu- tic effect is called a mineral water. There are very many kinds of mineral waters, differing accord- ing to the amount and kind of dissolved matter they contain. We distinguish between saline waters containing common salt, bitter waters with magnesium salts, sulphu- rous waters with sulphuretted hydrogen, carbonated waters with carbonic acid, chalybeate waters with iron, and many others. Detailed analyses of the mineral waters of numerous watering-places are accessible in works on balneology. Sea-water contains about 3% of salts, of which 2.7% is common FIG. 9. PASTEUR- CHAMBERLAND FILTER. 22 INORGANIC CHEMISTRY. [ 17- salt. A large number of elements, viz., about thirty, have been found in sea-water, although the most of them exist there only in extremely small quantities. It was stated above (16) that pure water is blue. The color of the rivers, lakes and seas varies, however, through many nuances from pure blue to brown. This variation is due principally to the presence of more or less brownish-yellow humous (marshy) substances or an extremely fine floating slime. Both conditions can produce a brownish-yellow color. It is easily seen how the combination of blue and yellow or brown may bring about the various blue, green or brown tints in natural waters. COMPOSITION OF WATER. 18. Decomposition. It was stated above that water can be obtained by direct combination of hydrogen and oxygen; inversely, it can be decomposed into these same elements. In the flask A (Fig. 10) some water is heated till it boils vigorously. A strong electric current is then sent through the wire a c b, so that the fine platinum wire c glows intensely. This heat partially decomposes the f ^ ? 'f 'II ^ ^ !M If, II" -'||. FIG. 10. DECOMPOSITION OF WATER BY GLOWING PLATINUM, water vapor into hydrogen and oxygen, which pass out through the tube d and are collected in the cylinder C. This gas mixture is nothing but the explosive mixture ( 13) of hydrogen and oxygen, as can be easily proved by applying a flame. 19.] COMPOSITION OF WATER. 23 Many metals decompose water on contact, the hydrogen being set free and the metal uniting with the oxygen. Potassium and sodium effect this decomposition at ordinary temperatures ( 11); iron, zinc and other metals require a higher temperature, iron, e.g., acting at red heat. 19. Let us now study the quantitative composition of water, i.e. determine the relative amounts of hydrogen and oxygen present. For this purpose both the analytic and synthetic methods can be used. (a) The Analytic Method. When an electric current is passed through water to which has been added a little sulphuric acid, the water is decomposed. If the gases evolved at the electrodes are collected separately, it is found that for every 1 vol. oxygen 2 vols. hydrogen are given off. A suitable apparatus for this experiment is shown in Fig. 11. Since 1 liter of hydrogen weighs 0.0899 g. and 1 liter of oxygen weighs 1.4296 g., both at and 760 mm. pressure, the weights of 2 vols. hydrogen and 1 vol. oxygen must bear to each other the ratio of 2X0.0899 : 1.4296, or 1 : 7.943. (6) The Synthetic Method. As early as 1820 the reduction of copper oxide by hydrogen was employed for this purpose by BERZELIUS; in 1834, also, by DUMAS and STAS. A weighed amount of care- fully dried copper oxide is heated in a current of hydrogen and water is formed, which is collected and weighed. The weight of the oxygen given up by the copper oxide is found from the difference between the weight of the copper oxide used and that of the resulting copper. The weight of the hydrogen contained in the water collected is therefore equal to the difference in weight of water and oxygen. The apparatus used for this experiment is represented in Fig. 12. In A the hydrogen is generated from zinc and dilute sulphuric FIG. 11. ELECTROLYSIS OF WATER. 24 INORGANIC CHEMISTRY. 19- acid. It is then passed through the permanganate solution in the wash-bottle B to free it from impurities, and also through the U- tubes C, D and E, containing calcium chloride, sulphuric acid and phosphorus pentoxide, respectively, for drying it. In F is placed the copper oxide, which is carefully weighed together with the tube. The water that forms is condensed in G, the U-tube H being attached to absorb any escaping water vapor. At the completion of the experiment, F, with its contents, is again weighed, likewise G and H; the differences in weight indicate the amount of water FIG. 12. SYNTHESIS OF WATER AFTER DUMAS AND STAS. formed. DUMAS and STAS found in this way that 100 parts (by weight) of water consist of 11.136 parts of hydrogen and 88.864 parts of oxygen, or, in other words, that the mass-ratio of these elements is 1:7.980, a relation which agrees with that ob- tained in (a) within the range of the unavoidable experimental error. Another synthetic method, which is especially adapted to the lecture-table, consists in mixing hydrogen and oxygen and deter- mining in what volume-ratio these gases unite. For this pur- pose an apparatus (Fig. 13) described by HOFMANN is best em- ployed. Hydrogen and oxygen in different proportions by volume are introduced into the arm of the U-tube, which can be closed by 20.] COMPOUNDS AND MIXTURES. 25 a stop-cock at the top : the cock is thereupon closed and the arm tightly stoppered with a cork. The mixture is then exploded by an in- duction spark, the volume of air en- closed on the other side acting as a cushion to moderate the severe shock on the mercury, which might otherwise break the apparatus. It is found that only when the volumes of hydrogen and oxygen bear to each other the ratio 2:1 does the entire gas mixture dis- appear, a slight coating of tiny drops of water appearing in its place on the inside of the glass. In case more hy- drogen or more oxygen than the ratio calls for is let into the tube, the excess is found to remain after the explo- sion. From these experiments, analytic and synthetic, it follows that water has a FIG. 13. HOFMANN'S constant composition; it consists of 2 vols. of hydrogen and 1 vol. of oxygen, or of 1 part, by weight, of hydrogen to 7.943 parts of oxygen. open AP- SYN- COMPOUNDS AND MIXTURES. 20. In water we have become acquainted with a substance which is different in many and important respects from the elements of which it is composed. We have further seen that the elements in it bear to each other a fixed relation by weight. Such substances are known in very large number. Copper oxide, mercury oxide, sulphuric acid, potassium chlorate, common salt, soda and many others already mentioned belong to this class. In each of these, no matter how obtained, we discover by analysis or synthesis a definite proportion between the elements composing it. Such sub- stances are called compounds. In addition to the characteristics mentioned difference of prop- erties from those of the elements and constant composition we find that the compounds also have constant physical properties. 26 INORGANIC CHEMISTRY. [ 20- Under the same pressure water always has the same melting-point and the same boiling-point, in whatsoever way it may have been obtained; salt always crystallizes in the same crystal system; soda, at a definite temperature, always requires the same amount of water for solution, etc. When elements or compounds are brought together without any chemical action on each other taking place, we have a mixture of these elements or compounds. The number of possible mixtures is, of course, unlimited. They are distinguished from compounds by the following characteristics : In a mixture the properties of the components reappear in many and important respects. Gunpowder, for example, is a mixture of sulphur, charcoal and saltpetre. The latter is soluble in water; sulphur dissolves in carbon disulphide; charcoal is insoluble in both. These properties are still evident in the constituents of gunpowder. In a mixture of sulphur and iron filings one can detect with a micro- scope the yellow grains of sulphur and the black particles of iron. The iron can be drawn out with a magnet; the sulphur dissolved out by carbon disulphide. If, however, a mixture of 7 parts iron and 4 parts sulphur is heated, a glow passes through the powder and a compound of both iron sulphide is formed, whose prop- erties are entirely different from those of its elements. It is non- magnetic and insoluble in carbon disulphide and under the micro- scope only a homogeneous scoriaceous mass is seen. The constituents of a mixture, since they still preserve their properties, can often be separated from each other by mechanical means, e.g. by the use of microscope and tweezers, by sifting, by treatment with solvents, by washing, etc. In a mixture the ratio of the constituents can vary in all pro- portions. There are, for example, many sorts of gunpowder, dis- tinguished from each other by the proportions in which their con- stituents are mixed. When 1 part sulphur and 100 parts iron, or, on the other hand, 1 part iron and 100 parts sulphur, are mixed, we have in either case a mixture of both elements, possessing hardly the same, but at least analogous, properties. Moreover, a mixture often has no constant physical properties. Water has a constant boiling-point; the boiling-point of a mixture of benzene and turpentine, however, rises gradually as the more volatile component, benzene, distils off. The melting-point of 21,] COMPOSITION OF COMPOUNDS. ATOMIC THEORY. 27 sulphur is constant and can be accurately determined; that of a mixture of tin and lead differs according to the proportion of the elements and is in many proportions not at all sharp, there being only a softening instead of real fusion. In the examples cited here the distinction between a compound and a mixture is well marked. There are, however, other instances where this is not the case and where it is therefore very difficult to know whether one is dealing with a compound or a mixture. We shall meet with many examples of this later. There is, however, one way whereby a compound can be distinguished from a mixture, viz., by ascertaining whether or not the substance, prepared in different ways, has a constant composition. PHENOMENA ACCOMPANYING THE FORMATION OR DECOMPOSITION OF A COMPOUND. The most common phenomenon of this sort is an elevation or depression of temperature, i.e. an evolution or absorption of heat (caloric effect). Sometimes the rise of temperature is so great that light is produced. A decomposition or a combination can be so violent that it causes an explosion. In other instances electricity may be produced by chemical action. All these facts may be com- prised in this statement: Chemical action results in a change in the energy-supply of the reacting substances. EXPLANATION OF THE CONSTANT COMPOSITION OF COMPOUNDS. ATOMIC THEORY. 21. It was stated that constant composition is the distinctive characteristic of a chemical compound. The proportions in which elements unite to form a certain compound are always the same. This Law of Constant Composition (definite proportions) was finally established by PROUST in the beginning of the nineteenth century, and at about the same time DALTON offered an expla- nation of it which is still accepted and may be considered as the foundation of theoretical chemistry. This explanation involves a hypothesis as to the constitution of matter. It is possible to regard matter as infinitely divisible; according to human conception the smallest particle that can really be obtained is still capable of division into an infinite number of others. However, even the ancients were of the opinion that there 28 INORGANIC CHEMISTRY. [21- must be somewhere a limit to the divisibility and that we must finally arrive at particles incapable of. further division, the atoms. In the fifth century B.C. there existed a school of philosophy, that of the Eleatics (so called from the city of ELBA), whose most prominent representative was PARMENIDES. He taught that everything that exists cannot be otherwise conceived than as unchangeable; trans- formation of the existent, which was thought to have never originated and to be at the same time unalterable, was held by them to be incon- ceivable. These theses they regarded in a certain sense as axioms, i.e. statements of truths which are accepted without proof. Daily experience teaches one nevertheless that transformation does occur in that which exists, a fact that led them to suppose that everything observed by men is merely appearance. Three theories were proposed in the same century which aim to form a bridge between the doctrine of the unalterable existent and the experience that points toward continuous change. These theories originated with EMPEDOCLES, ANAXAGORAS, and the Atomists, LEUCIPPUS and DEMOCRITES. The immutability of the existent is disposed of by ascribing it to extremely small unchangeable and indestructible particles ; every change is thought to depend on the movement of these smallest integral particles toward or away from each other. EMPEDOCLBS and ANAXAGORAS assume in this connection an infinite divisibility; the Atomists, on the contrary, regard the world as built up of indivisible particles, atoms, all of which consist of the same primordial substance but differ in form and size. Now DALTON has used this conception of the ancients regarding the atom to explain the fact that the combining weights are con- stant. The atoms of the various elements, he assumes, have dif- ferent weights; the atoms of the same element are alike in weight. A compound of two elements is therefore produced by the associa- tion of atoms of these elements. Such a combination of two or more atoms is called a molecule. It is obvious that these supposi- tions lead directly to the law of constant proportions; for, if copper oxide is formed by an atom of copper uniting with an atom of oxygen to make a molecule of copper oxide, its composition must, according to the above hypothesis, be constant. DALTON deduced another conclusion from his hypothesis, and confirmed the same experimentally. He observed that oxygen unites not only with one very definite amount of nitrogen oxide, but also with twice as much, not, however, with any intermediate 22.] THE ATOMIC WEIGHTS OF THE ELEMENTS. 29 amount, He also showed by the investigation of marsh-gas and oiefiant gas, both of which are made up of only carbon and hydro- gen, that the former contains twice as much hydrogen to a certain weight of carbon as the latter. It is readily seen how such observa- tions can be explained on the basis of the atomic theory; in one case 1 atom of carbon is in combination with n atoms of hydrogen; in the other with 2n atoms. The observations of DALTON were subsequently confirmed and extended, especially by BERZELIUS. The following statement is therefore now accepted as a law: When two elements combine to form more than one compound, the different weights of the one element which unite with one and the same weight of the other element bear a simple ratio to each other. This is the Law of Multiple Proportions. THE ATOMIC WEIGHTS OF THE ELEMENTS. 22. The absolute weight of the atoms is only approximately known (see 35). Nevertheless, their relative weights, i.e., the weights of the atoms of the various elements, when that of a certain element is arbitrarily fixed, have been determined in a variety of ways ( 208-210). These relative weights are known as atomic weights. It is now customary to take the atomic weight of oxygen as 16.00. The atomic weights of the remaining elements then have the values that are given in the table on the inside of the back cover of this volume. The acceptance of 16 as the atomic weight of oxygen has a historic reason. For a long time hydrogen was taken to be 1 ; it was believed that the ratio of the atomic weights of hydrogen and oxygen was 1:16. Inasmuch as the atomic weights of most elements are determined from the composition of their oxy- gen compounds, the basis is really O = 16 and not H= 1. This made no difference, so long as the proportion H:0=l:16 was con- sidered accurate. Even when the ratio was later found to be a different one (according to investigations of MOELEY and of W. A. NOTES the ratio 1:15.88 may now be regarded as very accurately determined), it was still the simplest plan to preserve O = 16 as the basis, since a change would necessitate a complete recalculation of all the atomic weights, and this necessity would 30 INORGANIC CHEMISTRY. [22- moreover recur as often as a new refinement of methods of inves- tigation brought about a change in the ratio H:O. A few years ago there was established a permanent inter- national commission whose duty it should be to revise the table of atomic weights critically every year. Those values are accepted as the " international atomic weights " which appear to be the most probable among the determinations that have been pub- lished. The atomic weights in the table are carried out to as many decimal places as may be accepted with certainty. For many purposes, however, it is sufficient to use round numbers, such as N=14, Br = 80, etc. Besides the atomic weights, we quite frequently use equivalent weigfits. These are the weights of the elements which combine with a unit amount of a certain standard element. One part of hydrogen combines, for instance, with 35.5 parts of chlorine and with 8 parts of oxygen. These amounts of hydrogen, chlorine and oxygen are equiv- alent to each other. The atomic weight is either equal to the equivalent weight or a multiple of it. CHEMICAL SYMBOLS AND FORMULAS. 23. The relative, or atomic, weights are expressed by symbols, that were introduced by BERZELIUS and are of great convenience in the representation of compounds and the formulation of chemical reactions. The symbols whose derivation is not at once apparent are taken from the Latin names of the elements; e.g., Sb from stibium, Au from aurum, Cu from cuprum, Hg from hydrargyrum, Pb from plumbum, Sn from stannum, Fe from ferrum, and Ag from argentum. A symbol stands not only for the element concerned, but also for the relative weight of an atom of that element. If the atomic weight of copper is 63.57 and that of oxygen 16.00, the symbol Cu indicates 63.57 parts by weight of copper, the symbol O 16.00 parts by weight of oxygen. It has been deter- mined that in copper oxide one atom of copper is combined with one atom of oxygen; copper oxide is therefore represented by the formula CuO, which expresses, first, that we are dealing with a compound of copper and oxygen, and, second, that 1 atom (63.57 parts by weight) of copper is united in it to 1 atom (16.00 24.] STO1CH10METRICAL CALCULATIONS. 31 parts by weight) of oxygen. Many compounds contain several atoms of the same element. This is indicated by placing the proper figure to the right of and below the symbol. Sulphuric acid, for example, contains 2 atoms of hydrogen (H), 1 atom of sulphur (S) and 4 atoms of oxygen (0) in the molecule. Its formula is, therefore, H 2 S04. Chemical actions can be very simply represented by the use of these formulae; thus, the decomposition of mercuric oxide into oxygen and mercury by HgO = Hg+0; that of potassium chlorate into oxygen and potassium chloride by KC10 3 = KC1 + 30; Potass, chlorate. Potass, chloride. the generation of hydrogen from zinc and sulphuric acid by In such equations the same atoms and the same number of each must appear on both sides, in accordance with the principle of the Indestructibility of Matter. STOICHIOMETRICAL CALCULATIONS. 24. If the formulae of the compounds are known the means of ascertaining these will be discussed in detail later and the atomic weights of the elements composing them also known, it is very easy to calculate the weights that enter into reaction in all chemical changes. A couple of examples may serve to make this clear. 1. It is required to know how many liters of oxygen at and 760 mm. pressure can be obtained by heating 1 kilogram of mercuric oxide. The atomic weight of mercury is 200, that of oxygen is 16; mercuric oxide, HgO, is, therefore, 200 + 16. Out of these 216 parts by weight of mercuric oxide 16 parts of oxygen can be ob- 32 INORGANIC CHEMISTRY. [24- tained by heating, i.e. from 1 kilo (=1000 g.) can be obtained =74.07 g. Since 1 1. oxygen at 0and 760 mm. pressure weighs 1.4296 g., 74.07 g. occupy a volume of Q =51.8 i j. .TT^yo 2. How much water can be formed from the hydrogen obtained by the interaction of 1 kg. zinc and the corresponding amount of sulphuric acid? The reaction of zinc and sulphuric acid is expressed by the equation Zn + H 2 SO 4 = ZnSO 4 + 2H and the combustion of hydrogen to form water by the equation 2H + O=H 2 O. From these equations it follows that the hydrogen formed by the action of 1 atom of zinc yields 1 molecule of water. For every atom of zinc we obtain, therefore, 1 molecule of water. The atomic weight of zinc is 65, the molecular weight of water 18; therefore 65 parts of zinc correspond to 18 parts of water. 1 kg. zinc must . , , 1000X18 yield - - = 276.9 g. 3. How many grams of potassium chlorate are necessary to pro- duce enough oxygen to oxidize 500 g. copper to copper oxide? The reactions concerned are KC1O 3 =KC1+3O and Cu+0=CuO. Hence 3 atoms of copper can be oxidized with the oxygen derived from 1 molecule of potassium chlorate. For every 3 atoms of copper 1 molecule of potassium chlorate must be consumed. The molecular weight of the latter substance is 39.10+35.6 + 3X16 = 122.56; the atomic weight of copper is 63.57; for every 122 56 63.57 parts of copper ^ = 40.85 g. potassium chlorate are o 500X40 85 therefore required. Hence 500 g. copper require - = bo .57 321.5 g. potassium chlorate. In most chemical computations gram molecules are employed, these being the molecular weights of the substances in grams. The abbreviation mole has been suggested by OSTWALD for this 25.] CHLORINE. 33 term. Thus " 1 mole " copper oxide means 63.57+16.00 = 79.57 grams of it. The molecular weight in milligrams is called a millimole. In the same way we may speak of a kilomole, etc. CHLORINE. 25. Chlorine does not occur free in nature, since it acts upon the most diverse substances at ordinary temperatures. In com- pounds, however, it occurs extensively. Common table salt is a compound of sodium and chlorine. Various other metallic chlor- ides are also met with in nature. . Chlorine gas can be obtained by the direct decomposition of certain chlorine compounds; thus: 1. By the electrolysis of hydrochloric, or muriatic, acid (i.e. a solution of hydrogen chloride, HC1, in water). Chlorine is given off at the positive pole (anode), hydrogen at the negative pole (cathode). The indirect decomposition of its compounds offers, as in the case of hydrogen ( 11), the most practicable methods of obtaining the element. They are all based on the oxidation of the hydrogen of hydrochloric acid, whereby water is formed and chlorine liber- ated. 2. Commercially, as well as in the laboratory, manganese dioxide, Mn0 2 , is frequently used as the oxidizing agent: Mn0 2 + 4HC1 = MnCl 2 + 2H 2 + 2C1. It is very often convenient to generate the hydrochloric acid from salt and sulphuric acid in the same vessel with the manganese dioxide. The two reactions thus proceed simultaneously: I. NaCl+H 2 S0 4 II. 4HC1 + MnO 2 = MnCl 2 + 2H 2 + 2C1. 3. Other commonly used oxidizing agents are chloride of lime and potassium dichromate; e.g. K 2 Cr 2 7 + 14HC1 = 2KC1 + Cr 2 Cl 6 + 7H 2 + 6C1. 34 INORGANIC CHEMISTRY. [25- 4. The oxygen of the air can also serve as the oxidizing agent: 2HC1+O=H 2 O+C1 2 . For this purpose a mixture of 60% of air and 40% of hydrogen chloride at about 430 is passed over porous bricks which are soaked with copper sulphate solution. About 70% of the hydro- gen chloride is converted into chlorine. This method, which is known as the Deacon process, is used commercially. The copper sulphate serves as a catalyzer. The progress of chemical changes is often modified by the mere presence of a substance which has the same chemical con> position after the reaction as at the beginning. Such a substance is termed a catalyzer and the action which it exerts is called catalysis, or catalytic action. The quantity of the catalyzer necessary to exert a perceptible influence is often very small, This is the case, for example, in the combination of hydrogen and oxygen in the presence of platinum as a catalyzer ( 13, p. 16). A minute trace of platinum sponge brought into contact with detonating-gas accelerates the combination to such a rate that the reaction takes place very quickly and can even become explosive. In the DEACON process a small quantity of copper sulphate suffices to bring into reaction unlimited quantities of hydrogen chloride and oxygen. At the temperature of ^430 there is practically no reaction between oxygen and hydrogen chloride without the catalyzer. That there must, nevertheless, be a reaction, although a very slow one, can be demonstrated by the same reasoning as in 13. The catalyzer therefore does not cause a reaction, but only accelerates it. OSTWALD compares its action to that of oil on the axles of a machine which move with very great friction. When oiled, the ma- chine will go much faster, notwithstanding that the force of the spring (here the energy of the chemical reaction) has not changed. A further point in the analogy is that the oil is not consumed. In most cases of catalysis it can be proved that the catalyzer takes part in the reaction but at the end of it reappears in its original condition. In the platinum catalysis of detonating-gas, 27.] CHLORINE. 35 for example, the metal unites with the oxygen, whereupon the resulting compound reacts with the hydrogen, giving water and metallic platinum. The phenomenon of catalysis is universal. OSTWALD thinks it probable that there is no kind of chemical reaction that cannot be influenced catalytically and that there is no substance, element, or compound, which cannot act as a catalyzer. Catalyzers may accelerate or retard reactions; at present, however, much more is known of the first than of the second kind. 26. Physical Properties. Chlorine is yellowish-green (hence its name, which is derived from jAc^pos, greenish-yellow) and has a disagreeable odor. Its specific gravity is 2.45, taking air as unity, or 35.46, based on O=16. 1 1. chlorine weighs, therefore, 3.208 g. at and 760 mm. pressure. At -34 it becomes liquid under ordinary pressure;, at 102 it solidifies and crystallizes. Its critical temperature is 146. Liquid and solid chlorine are yellow. Chlorine gas dissolves in about one-half its volume of water.- The aqueous solution bears the name " chlorine- water.". It can, therefore, not be collected over water, but a saturated salt- solution may be used, in which it is only slightly soluble. The most convenient way to fill a vessel with it is by displacement of air, the gas being conducted to the bottom, where it remains and drives out the air above, because the chlorine is denser. 27. Chemical Properties. Even at ordinary temperatures, chlo- rine combines with many elements and acts on many compounds. If perfectly pure chlorine is mixed with an equal volume of hydro-, gen, the two unite in direct sunlight, causing an explosion. If the chlorine is impure or the sunlight diffused, combination occurs slowly. When a hydrogen flame is introduced into chlorine gas, it continues to burn, with the formation of hydrogen chloride. Many metals combine with chlorine with the evolution of light, e.g. cop- per (in the form of imitation gold-leaf), finely powdered antimony, molten sodium, etc. The precious metals are in general quite resistive to chemical action. They are, however, attacked by chlorine and changed to chlorides, i.e., chlorine compounds. Gold, for instance, dissolves in chlorine-water, forming gold chloride. 36 INORGANIC CHEMISTRY. [27- Chlorine also unites readily with many non-metals, e.g. phos- phorus, which burns in it with a pale flame to phosphorus chloride. The tendency of chlorine to unite with hydrogen its so-called chemical attraction, or affinity, for the latter is so strong that chlorine abstracts the hydrogen from many hydrogen compounds in order to combine with it. A strip of paper dipped in turpentine burns with a sooty flame when introduced into an atmosphere of chlorine; the chlorine unites with the hydrogen of the turpentine and sets the carbon free. A burning candle continues to burn in chlorine, depositing soot (carbon) and forming hydrogen chloride. If sulphuretted hydrogen gas, H 2 S, is passed into chlorine-water, hydrochloric acid and sulphur are formed. Water is also decomposed by chlorine, oxygen being liberated : 2H 2 0+2C1 2 =4HC1 + O 2 . This reaction takes place under the influence of sunlight, but proceeds very slowly. It can be conveniently demonstrated as in Fig. 14. A retort is filled with dilute chlorine-water, inverted and exposed to the sunlight. After a few days a bubble of gas collects at the top of the retort, and. on investigation with a glowing splinter, it is found to be oxygen. FIG. 14. SLOW DECOMPOSITION OF WATER BY CHLORINE. Upon this decomposition of water depends the bleaching and disinfect' ing action of chlorine and those substances which generate chlorine. In bleaching, the coloring matters usually of an organic nature are oxi- dized by oxygen to colorless substances. Bacteria are killed by oxida- tion. Ordinary atmospheric oxygen does not produce these effects. Lit- mus, for instance, which is rapidly decolorized in moist chlorine gas, is 28.] HYDROGEN CHLORIDE. 37 totally unaffected by the air. The particularly energetic action of the oxygen that is produced from water by chlorine is explained by assuming that it exists in an atomic condition, the status nascens, regarding which more will be said later ( 38). Perfectly dry chlorine has no bleaching power. If water is saturated with chlorine at 0, crystals are deposited, of the composition C1 2 + 8H 2 O, chlorine hydrate. At a higher temperature these are wholly decomposed into chlorine and water. HYDROGEN CHLORIDE, HC1, and HYDROCHLORIC ACID. 28. Hydrochloric acid, of the formula HC1 ( 31), is a gas, occurring in nature in the free state, e.g. in the gases of some volcanoes. It forms an important, although small, part of the gastric juice of man and other animals. Some of its methods of formation have been already given ( 27), e.g., by direct synthesis from its elements under the influence of light. It is quite remarkable, however, that ultraviolet rays decompose hydrogen chloride even at ordinary temperatures. We saw also (I. c.) that hydrogen chloride is formed by the action of chlorine on hydrogen compounds. Moreover, it can also result from the action of hydrogen on some chlorine com- pounds, e.g., silver chloride, AgCl, and lead chloride, PbC^, when heated in a current of hydrogen, yield metal and hydro- chloric acid: AgCl+H=Ag + HCl. The ordinary method of preparation is by the action of a chlorine compound on a hydrogen compound, viz., that of salt (sodium chloride) on concentrated sulphuric acid: NaCl + H 2 S0 4 = NaHS0 4 + HC1. Sodium Sulphuric chloride. acid. This method is employed technically as well as in the labora- tory. 38 INORGANIC CHEMISTRY. [28- The above reaction takes place at ordinary temperatures. If the sulphuric acid is to be completely used up, i.e. if all the hydrogen of the sulphuric acid is to go off with the chlorine of the salt as hydrochloric acid, the temperature of the reaction must be raised (cf. also 226) : 2NaCl+ H 2 SO< -Na^O, +2HC1. 29. Physical Properties. Hydrogen chloride is a colorless gas with a pungent odor. Its critical temperature is +52.3; the critical pressure 86 atmospheres. Liquid hydrogen chloride boils at -83.7; the solid melts at -111.1. Specific gravity of the g as = 1.2696 (air = l); 1 1. HC1 at and 760 mm. pressure weighs 1.6533 gr. For obtaining hydrogen chloride in a pure state MOISSAN has elabor- ated a method which is generally applicable to gases, since low tempera- tures are easily attainable by means of liquid air. The freshiy generated gases contain in most cases moisture and other impurities. The gases are first dried by being passed through one or two wash-bottles placed in a bath of a lower temperature than 50. At that temperature the tension of water vapor is practically zero. The gases dried in this way are now condensed by strong cooling to the solid state. Air can then be pumped out of the vessel If the temperature is now allowed to rise, the solid mass melts first; the resulting liquid, when vaporized, gives the perfectly pure gas. The gas fumes strongly in the air, forming a cloud with the moisture of the air. It is very soluble in water, 1 vol. water at being able to absorb 503 vols. HC1 gas. The aqueous solution of the gas is called " hydrochloric acid," * also muriatic acid. It is manufactured commercially on a large scale ( 226). Hydrochloric acid is employed almost exclusively in the form of this aqueous solution. A solution saturated at 15 contains 42.9% HC1 and has a specific gravity 1.212; it fumes vigorously in the air. The ordinary pure " concentrated" or "fuming" muriatic acid of commerce usually has a specific gravity of 1.19 and contains about 38% HC1. * The gas itself is often called ''hydrochloric acid gas"." 30.J HYDROGEN CHLORIDE. 39 Hydrogen chloride does not obey the law of HENRY ( 9) in its behavior towards water, for its solubility in this liquid is not at all proportional to the pressure. The larger part of it is absorbed in water without reference to the pressure, and an increase of pressure causes only a small increase in the solubility. Such con- duct indicates that a change in the compound has occurred; just what this change consists in we shall soon have occasion to con- sider ( 65, 66). The saturated solution of hydrogen chloride in water gives off HC1 on warming. On distilling it a fraction is obtained that boils con- stant at 110 and contains 8 mols. H 2 O to 1 mol. HC1, corresponding to about a 20% solution of HC1. A solution of the same concentration and boiling-point results from distilling a more dilute hydrochloric acid, enough water boiling off to raise the concentration to the above value. 30. The chemical properties of hydrogen chloride are found to be quite different when it is in a perfectly dry condition, e.g. con- densed to a liquid, than when it is dissolved in water. In the former case it does not act on metals nor change the color of blue litmus. In the latter case just the contrary is true. Zinc, iron, and other metals, when dipped in the aqueous solution of hydrogen chloride, are vigorously attacked, hydrogen being given off. Blue litmus is turned red by the solution. Moreover, even dilute solu- tions taste sour. Now, there are a lot of substances that undergo a similar change of properties when they are brought in contact with water, and whose aqueous solutions possess about the same properties as those that are described here for hydrochloric acid. The nature of this change will be discussed later on ( 65). It should be stated here, however, that these substances have a common name. They are called acids. Acids have one or more hydrogen atoms that can be replaced by metals. The compounds of metals that are formed by such substitution are called salts. Salts can result not only from the direct action of metals on acids, but also from the interaction of acids and bases. The term " bases " includes compounds of the general type MOH, where M represents a metal. Most of them have an alkaline taste and turn red litmus 40 INORGANIC CHEMISTRY. [30- blue. When sodium is dropped into water, hydrogen is generated, and a base, sodium hydroxide, is formed: Na+H 2 O=NaOH + H. If this hydroxide is now treated with hydrochloric acid, sodium chloride and water are produced: NaOH + HC1 = NaCl + H 2 0. If we indicate an acid by the general formula AH and a base by MOH, the formation of salts from the interaction of the two may be represented thus: MOH+HA=MA+H 2 0. A third way of forming salts is by the action of an acid upon a metallic oxide, e.g. ZnO + H 2 SO 4 = ZnSO 4 + H 2 0. Zinc Sulphuric Zinc oxide. acid. sulphate. In general, the bases are built up from metals, the acids from metalloids. When hydrochloric acid is added to a solution of a silver salt, fok' instance to silver nitrate, a decomposition of this salt takes place according to the equation AgNO 3 + HC1 = HN0 3 + AgCl. Silver nitrate. Nitric acid. Silver chloride. The silver chloride is insoluble, and is precipitated as a white, curdy mass. In this reaction the hydrochloric acid has liberated the nitric acid from its salt. It is also possible to liberate a base from a salt by the addition of another base: AgN0 3 + NaOH = AgOH + NaNO 3 . Silver Sodium hydroxide. nitrate. Such reactions are called single, or simple, decompositions. 31.] COMPOSITION OF HYDROCHLORIC ACID. 41 Now it can also happen that two salts exchange their metals when brought together : Nad + AgN0 3 = AgCl + NaN0 3 , Sodium chloride, so that two other salts are obtained. Such a reaction between salts is called a double decomposition. We shall later have occasion to study the laws governing both of these decompositions. COMPOSITION OF HYDROCHLORIC ACID. LAWS OF GAY-LTJSSAC AND AVOGADRO. 31. The composition of hydrochloric acid is determined by the following experiments : (a) When strong hydrochloric acid (a more than 23% solution) is subjected to electrolysis in a suitable apparatus (see below) it is observed that equal volumes of hydrogen and chlorine are evolved. (6) Equal volumes of chlorine and hydrogen unite to form hydrochloric acid without leaving a remainder of either element. 2 vols. HC1 are formed. Since the weight of 1 vol. Cl is 35.46 (O=16), hydrochloric acid must consist of 1 part by weight of hydrogen combined with 35.46 parts of chlorine. In the electrolysis of hydrochloric acid sticks of charcoal are ordinarily used, because platinum, the substance employed in most other electrolyses, is attacked by chlorine. The apparatus of Fig. 11 is also impracticable, since the solubility of chlorine in water increases with rising pressure more rapidly than that of hydrogen, and equal volumes of both gases are therefore not obtained. In its place we use an apparatus suggested by LOTHAR MEYER (Fig. 15), by which the compression of the chlorine by a steadily rising column of liquid is avoided. In A hydrochloric acid is electrolyzed and the hydrogen and chlorine are collected in the cylinders BB, which are filled with a saturated sodium chloride solution. The collected gases are thus under diminished pressure. The combination of equal volumes of chlorine and hydrogen can be carried out in a thick-walled tube, that is filled with the gases and then exposed for a day to diffused sunlight. Since the success of the experi- ment requires the use of the exact proportions of chlorine and hydrogen 42 INORGANIC CHEMISTRY. 31- and their absolute purity, the gas mixture is prepared by electrolysis in the dark and exposed to the action of light immediately after the tube is filled. FIG. 15. ELECTROLYSIS OF HYDROCHLORIC ACID. The fact that hydrochloric acid gas yields a volume of hydrogen equal to half its own volume can also be shown in another way. When perfectly dry hydrogen chloride is treated with sodium amalgam a solution of sodium in mercury the sodium combines with the chlorine, setting hydrogen free. The volume of the latter is then found to be half as large as that of the hydrochloric acid taken. Hydrogen and chlorine thus unite in a very simple ratio by volume (1:1), and the volume of their product also bears a very simple ratio to that of the components (2:1:1). In discussing the composition of water ( 19) we already remarked that oxygen and hydrogen combine in a very simple ratio by volume, viz., 1:2. By carrying out this synthesis at a temperature above 100, so that the steam is not condensed to water, it is found, further, that the volume of resulting steam bears a simple ratio to the volumes of its components, viz., that 1 vol. O + 2 vols. H gives 2 vols. H 2 O. The following arrangement serves this jwrpose (Fig. 16). The explosive mixture is introduced into the closed arm B of the U-tube over mercury. B is surrounded by a glass jacket, through which the vapor of boiling amyl alcohol (generated in A), whose temperature is about 130, is passing. This vapor is condensed in C. As soon as the gas mixture has reached this temperature an induction spark is flashed through, and it is found that the volume of steam formed is two-thirds that of the mixture. 31. J COMPOSITION OF HYDROCHLORIC ACID. FIG. -DETERMINATION OF THE VOLUME RELATIONS BETWEEN STEAM AND ITS COMPONENTS. What was found above to be true for hydrochloric acid and for water is a general principle. Gaseous elements combine in simple proportions by volume, and the volume of the products formed in the gaseous state also bears a simple ratio to the volumes of the com- ponents. This law was discovered by GAY-LUSSAC in 1808. This law, together with the atomic theory of DALTON, leads to important conclusions. In order to investigate the matter, let us assume that the formula of hydrochloric acid is HC1; in other words, that an atom of hydrogen is in combination with an atom of chlorine. Since one volume of hydrogen unites with one volume of chlorine to form the compound, it follows from the above as- sumption that equal volumes of chlorine and hydrogen contain the same number of atoms. If the formula were otherwise, e.g. H n Cl OT , the numbers of atoms in equal volumes of hydrogen and chlorine would be in the ratio of n : m. In the synthesis of water 2 vols. of hydrogen and 1 vol. of oxy- gen yield 2 vols. of steam. If the formula of water be H 2n O p , the numbers of atoms in equal volumes of hydrogen and oxygen must bear to each other the ratio n:p. Given, therefore, the relative numbers of atoms in equal gas 44 INORGANIC CHEMISTRY. [ 31- volumes and the volume ratio in which the gases unite, we can determine the formula of the resulting compound. As to the number of atoms in equal gas volumes, there was at first much uncertainty. Since all gases behave exactly alike to- wards changes of pressure or temperature, it was reasonable to suppose that the number should be alike for all gases; but this was soon shown to be incorrect. In the synthesis of water 3 vols. (2 vols. H+ 1 vol. O) give 2 vols. of steam; hence the number of atoms per unit volume must be different for steam than for the uncom- bined elements. However, all difficulties were overcome by a hypothesis, which AVOGADRO enunciated in 1811, to the effect that equal volumes of all gases at the same temperature and pressure con- tain the same number of molecules. AVOGADRO further supposes that the molecules of oxygen, hy- drogen, chlorine, and other elements consist of two atoms. The union of hydrogen and chlorine is then explained thus : Out of a molecule of each, two molecules of hydrochloric acid are formed : H 2 + C1 2 =2HC1. 1 vol. 1 vol. 2 vols. The total number of molecules thus remains the same after the combination and, since the entire volume has suffered no change either, there must' be just as many molecules present in each of the two volumes of hydrochloric acid as in each of the volumes of hydrogen and chlorine. The combination of hydrogen and oxygen takes place thus: 2H 2 +0 2 =2H 2 0. 2 vols. 1 vol. 2 vols. Every molecule of oxygen has split up into its two atoms, and each of these unites with two hydrogen atoms. The number of water molecules becomes therefore twice as great as that of the oxygen molecules and equal to that of the hydrogen molecules; but, since the volume of steam is also double that of oxygen, there must be in equal volumes just as many water molecules as oxygen molecules and hydrogen molecules. 32. It follows from the above that AVOGADRO'S hypothesis is of importance hi two respects: (1) hi furnishing us a means of ascer- 32.] COMPOSITION OF HYDROCHLORIC ACID. 45 taining the relative weights of the molecules of gaseous substances; (2) in putting us in a position to form an idea of how many atoms there are in the molecules. Let us examine both points more closely. As to (1): Since equal volumes of gases under the same conditions contain the same number of molecules, the ratio of the weights of these volumes gives us at once the ratio of the molecular weights. If the specific gravity of steam is 9, based on O=16, and that of hydrochloric acid is 18.25, the ratio of the molecular weights of water and hydro- chloric acid is 9: 18.25. The determination of the specific gravity of gases and vapors, the vapor density, becomes therefore of the greatest importance to chemistry. The practical method of pro- cedure is described in ORG. CHEM., 12. For the determination of the specific gravity of gases, see 212. As to (2) : In order to understand how AVOGADRO'S hypothesis can furnish an idea of the number of atoms which the molecules oi elements and of compounds contain, let us return to the example of the synthesis of hydrochloric acid. 1 vol. hydrogen unites with 1 vol. chlorine to form 2 vols. hydrochloric acid. According to the above law there must be just as many molecules present in the two volumes of hydrochloric acid as there were molecules of hydrogen and chlorine together. It is evident that this is only possible in case the molecules of hydrogen and of chlorine divide into two parts. For, if the chlorine and the hydrogen molecules consisted of only one atom each, the volume of hydro- chloric acid could not, in accordance with AVOGADRO'S law, be double that of each of its elements, but would have to be equal to it. It therefore follows that an even number of atoms must be present in the chlorine and in the hydrogen molecules; whether or not this number is two, as AVOGADRO assumed, can evidently not yet be determined; we shall therefore represent the molecules of hydrogen and of chlorine by ~H.2x and C^j/. From the synthesis of water the same conclusion is reached in regard to the oxygen molecule: 2 vols. hydrogen unite with 1 vol. oxygen to form 2 vols. steam. In each of these two volumes of steam there must be, according to AVOGADRO'S law, just as many molecules present as in the one volume of oxygen. This is likewise impossible unless the oxygen molecule splits into two parts, each of which combines 46 INORGANIC CHEMISTRY. [ 33- with a molecule of hydrogen, so that we obtain H 2x O z as the for- mula of water and O 23 as that of the oxygen molecule. 33. The formulae for hydrochloric acid, for water and for the molecules of hydrogen, chlorine and oxygen can be fully established, if the values x, y and z are known. These can be ascertained gen- erally in the following way : x must be at least equal to 1 ; if this is the case, the molecule of hydrogen becomes H 2 . That a smaller number of atoms is impossible is shown by the synthesis of hydro* chloric acid. The vapor densities of a series of hydrogen com- pounds, as compared with that of hydrogen, are then determined, from which we can find their molecular weights, based on the hydrogen molecule as unity. Thereupon these compounds are analyzed and the amount of hydrogen calculated that is repre- sented in the different molecular weights. It will then be found that in no case is the amount less than half of that in a molecule of hydrogen. The following table gives some examples: Substance. Sp. G. (H=l). Quantity of H. Hydrogen chloride 18.25 0.5 Hydrogen bromide 40.5 0.5 ,Hydrogen sulphide 17 1 Ammonia gas 8.5 1.5 Methane 8 2 Ethylene 14 2 Water 9 0.5 Since, therefore, no compound contains less than half a molecule of hydrogen, the atomic weight of hydrogen must be half its molecular weight, i.e. the formula of the hydrogen mole- cule is H 2 . Similarly it is found that the oxygen molecule is O 2 , that of chlorine C1 2 ; in other words, that x, y, and z are all equal to 1. The following table illustrates the case of oxygen: Substance. Sp. G. (H = l). Quantity of O. Oxygen 16 16 Water 9 8 Sulphur dioxide 32 16 Nitric oxide 15 8 Carbon monoxide 14 8 Carbon dioxide . . 22 16 34.] DETERMINING MOLECULAR AND ATOMIC WEIGHTS. 47 RULES FOR DETERMINING MOLECULAR AND ATOMIC WEIGHTS. 34. When the atomic weight of oxygen is taken =16, its molecular weight is 2 X 16 =32. If the specific gravity of another gas based on oxygen= 16 is a, the molecular weight of this gas becomes 2a. The following rule has therefore been prescribed for the determination of the molecular weight. Determine the vapor density of the compound, based on oxygen=16, and multiply the result by 2; the product is the molecular weight. For determining the atomic weight the following holds good, according to 33 : Determine the composition of molecular amounts of as many compounds of the element as possible; the smallest amount of the element that is found in any instance is the atomic weight. AVOGADRO'S hypothesis has been confirmed from a physical stand- point. It is at present one of the principal laws of chemistry and physics. Let us briefly examine, among others, the physical arguments in its favor. The molecules of bodies, solids as well as liquids and gases, are in con- stant motion, the intensity of which increases and decreases with the temperature. In different substances at the same temperature there must be a definite relation between the intensities of the molecular movements. This relation has been successfully worked out from the theory in the case of gaseous substances. It has been shown that in all gases at the same tem- perature the mean kinetic energy of translation of a molecule is the same. The pressure which a gas exerts against the walls of the vessel is caused by the impact of the molecules. We will call the number of mole- cules in a volume of the gas n, the mass of each molecule m and their mean velocity u. It is then clear that the gas pressure the above explanation of its cause being accepted must be proportional to n and m. Moreover the pressure must also be proportional to u 2 , for if the velocity were increased the enclosing walls would receive more impacts from the molecules moving to and fro, and every impact would also become stronger. The gas pressure p is therefore proportional to the product nmu 2 ; the theory says that p=%nmu 2 , or n= ^T* In this expression mu 2 is twice the kinetic energy of translation of molecules, which is the same for all gases at the same temperature. If then p is made the same for the different gases,^^, or n, the number of molecules per unit volume, must be the same for all gases. The laws of BOYLE, GAY-LUSSAC and AVOGADRO (we refer to the expansion law of GAY-LUSSAC) can be expressed in a single 48 INORGANIC CHEMISTRY. [$ 3l _ comprehensive formula, which is worthy of note because of its frequent use in physical chemistry. The laws of BOYLE and GAY-LUSSAC are represented by the equation PV PV=RT, or ^-=#, in which P is the pressure, V the volume and T the absolute tem- perature, of the gas and R is a constant which depends on the quantity and the nature of the gas under consideration. The value of R, however, becomes the same for all gases, if molecular amounts of them (One mole each) are taken. For, according to AVOGADRO'S law, the volume of one mole of every gas is the same under the same pressure and temperature. In the 'above equa- tion, then, V is constant for all gases and, since we have already made P and T the same in each case, it is evident that R must have a constant value. In other words, if we deal with molecular amounts, the equation PV=RT becomes a general expression of the laws of BOYLE, GAY-LUSSAC and AVOGADRO. The value of R may be calculated as follows: Let us consider 1 mole oxygen at and 760 mm. pressure. Since 1 1. oxygen under these conditions weighs 1.4290 g., the volume V of 1 molis If a correction is applied because oxygen does not exactly follow the gas laws of BOYLE and GAY-LUSSAC, we obtain 22412 c.c. The pressure of 760 mm. mercury corresponds to a pressure of 1013.25 g. per sq. cm., i.e. P= 1013.25. At the absolute temperature is 273 (more strictly 273.09). Substituting these values in the above expression for R } we obtain PF_ 1013.25X22412 _ = T~ = 273.09 in c.-g. units. If the pressure is expressed in millimeters of mer- cury, R becomes 760X22412 273.09 The product PV also represents the external work which is 'done when a gas under constant pressure P increases its volume by V (on being heated, for instance) or when a gas being generated 35.] THE REALITY OF MOLECULES AND ATOMS. 49 under the pressure P comes to occupy a volume V. For, if we sup- pose that the gas is enclosed in a cylinder of 1 sq. cm. transverse section having at one of its ends a piston, the increase of the volume must cause a weight P to move through V cm. One calorie (gram- calorie) =4 1890 gram centimeters. If this is substituted in the equation PV = 83155T, the latter becomes PV = T , or, very Q -LOt/\J approximately, PV = 2T. This latter form also is a common one of expressing the combined gas laws. It gives the external work in calories that is done when 1 gram molecule of any given sub- stance is converted into the gaseous state at the absolute tem- perature T. Since 1 gram molecule of a gas has a volume of 22.4 1. at and 760 mm. pressure, 1 c.c. under these same conditions contains , or 0.0446, millimoles. THE REALITY OF MOLECULES AND ATOMS AND THEIR ABSOLUTE WEIGHT. 35. The law of AVOGADRO teaches that equal volumes of all gases contain the same number of molecules. If we take a gram molecule oi every gas, which, as we just saw, has a volume of 22.41 1., it follows at once that there must be the same number of molecules in every case. The number of molecules in 1 gram molecule of any given gas is thus a universal constant; it is often represented by N. The determination of this constant N, i.e., the absolute number of molecules contained in the gram molecule, has been worked out in recent years by several widely different methods, all of which have yielded approximately the same result, viz., 70X10 22 . Back in 1875 VAN DER WAALS in his famous treatise on the continuity of the gaseous and the liquid states, calculated the value of N to be between 40 and 90X10 22 , which is of the same order of magnitude as the present more accurate value. All the methods are of a physical character, so that a full description of them is inappropriate here; nevertheless, to show the diversity of these methods, we may mention that the above value of N has been obtained from (1) the law of VAN DER WAALS; (2) the Browniai> movement (the irregular movement which solid parti- 50 INORGANIC CHEMISTRY. [35- cles, having the dimensions of the order of a micron (/*= 0.001 mm.) or smaller, exhibit when they are suspended in a liquid. It has now been proved that these movements are due to the impacts of the molecules) ; (3) the diffusion velocity of dissolved substances ; (4) the refraction of light in the atmosphere, causing the blue color of the sky; (5) the electric charge of the ions (266); (6) the life period of radium (267); (7) the energy of the infra-red spectrum. The significance of these investigations for all departments of natural science is extraordinarily great. When we see that so many wholly independent methods lead to the same absolute number of molecules, 70 X 10 22 , in a gram molecule there is no room left for doubt of the actual existence of molecules. So long as we had only rough and discordant approximations of this number, we could accept the assumption that matter is built out of mole- cules and atoms as an exceedingly useful hypothesis, while yet doubting the real existence of atoms and molecules. The satis- factory proof of their actual existence has put the knowledge of matter in general on a secure foundation. OZONE. 36. As early as 1785 VAN MARUM observed that when an elec- tric spark passes through oxygen a peculiar " garlic-like " odor is given off, and a bright mercury surface is at once made dull. SCHONBEIN investigated this phenomenon more carefully, and found that it is due to the formation of a peculiar substance, which he called ozone. This proved to be oxygen existing in a special condition. The fact that it really consists of nothing but oxygen is shown by its formation from perfectly dry oxygen under the influence of electric discharges, e.g., induction sparks. The amount of ozone thus formed is nevertheless small. It is greater when silent discharges are used. As ozone is formed from oxygen by ultraviolet light, the formation of ozone by silent discharges may be caused by the ultraviolet light accompanying them. This is one of the best ways of obtaining ozone, although the maximum yield is only 5.6%. However, if the oxygen is cooled by liquid air and then submitted to the silent discharge at a pressure of 100 mm. Hg, it is wholly converted into ozone. Fig. 17 represents an apparatus constructed by BERTHELOT for the preparation of ozone at the ordinary pressure and tem- perature. The wide tube./, together with the supply-tube d 36.] OZONE. 51 FIG. 17. PREPARATION OF OZONE. and the exit-tube e, are sunk in a vessel of sulphuric acid, into which the pole of the inductor b is dipped. The other wire a of the latter ends in a tube c, which is slipped down inside / and is almost entirely filled. The silent discharge between the two bodies of sulphuric acid thus passes through a thin layer of oxygen and has a powerful ozoniz- ing effect. Ozone is formed iu many reactions, such as the slow oxidation of moist phosphorus, also in a small quantity, when hydrogen burns in an atmosphere of oxygen. The oxygen that is ob- tained by the electrolysis of dilute sulphuric acid always contains it. Ozone is also given off by the decom- position of permanganic acid that is set free in the reaction of potassium permanganate and concentrated sulphuric acid (cf. also 52). When oxygen is subjected to a very high temperature (e. g. flame temperature) it is partially converted into ozone, and the more so the higher the temperature (103). It is necessary, however, to cool down the ozonized gas very rapidly, because the velocity of decomposition of ozone is very great, especially at high temperatures. An instantaneous cooling can be accomplished by directing the flame (of hydrogen, carbon monoxide, acetylene or other gas) upon the surface of liquid air, which has a temperature of 180 . That the generation of ozone has no con- nection with the combustion, but that it is caused only by the high tem- perature to which the oxygen is raised by the flame, may be proved by the fact that an incandescent platinum wire or NERNST glower, dipped in liquid air also generates ozone. The formation of ozone is also observed when a rapid current of dry air or oxygen is allowed to impinge against a hot NERNST glower. When the air contains moisture almost no ozone is formed, the product being hydrogen peroxide (54). Physical Properties. At ordinary temperatures ozone is a gas ; it has a peculiar odor, which is one of the most delicate tests for its presence. One part of ozone can still be detected by its odor in 500,000 parts of air. In the liquid state it is indigo-blue. Ozone boils under normal pressure at 119. 52 INORGANIC CHEMISTRY. [36- Chemical Properties. Ozone is characterized above all by its ability to oxidize vigorously at ordinary temperatures, especially in the presence of moisture. Phosphorus, sulphur, and arsenic are oxidized to phosphoric acid, sulphuric acid, and arsenic acid, respectively, ammonia to nitric acid, and silver and lead to per- oxides; e.g., the metallic surface of silver, especially when heated to above 240, become blue when ozonized air is directed against it. Iodine is deposited by ozone from a solution of potassium iodide : 2KI + H 2 O + O = 2KOH + 21. Organic substances are strongly oxidized by ozone, hence no apparatus containing it should have connections of rubber. Dye- stuff solutions, like indigo and litmus, are decolorized (by oxida- tion) . Ozone effectively destroys micro-organisms, and is there- fore used successfully in the sterilization of drinking-water. The detection of ozone, especially in quantities too small to be recog- nized by the odor, is a difficult matter because several other oxidizing sub- stances, such as chlorine or bromine in the presence of water, the oxides of nitrogen, hydrogen peroxide and still others, give closely analogous reac- tions and furthermore, their smell at high dilutions somewhat resembles that of ozone ; hence it becomes necessary to first prove their absence. The tests for ozone are usually executed by moistening strips of filter-paper with the reagent and dipping them in the gas containing ozone. The reagents used for this purpose are lead sulphide and thallous hydroxide. The strips of paper are first moistened with dilute solutions of the nitrates of these metals and then exposed to hydrogen sulphide and ammonia fumes, re- spectively. Lead sulphide is oxidized by ozone to lead sulphate, thus turn- ing from black to white; thallous hydroxide, which is white, is converted to brown thallic hydroxide. However, these changes of color also occur with the other oxidizing agents mentioned. A strictly characteristic test for ozone is the violet color produced with an acetic acid solution of tetramethyl- p-p f - diamido-dipheriyl-methane (an organic compound.) Nitrogen dioxide gives a straw-yellow color, chlorine and bromine a dark blue, while hydrogen peroxide produces no coloration at all. Ozone is stable at ordinary temperatures, but is easily changed to oxygen on heating. It is slightly soluble in water. 37. Formula of Ozone. The formula of ozone has been determined by LADENBURG in the following way. A glass globe with two cocks was first weighed when filled with pure oxygen and then when containing ozonized oxygen. After reducing both weights to the normal temperature and pres- sure the globe in the latter case was found to be a mg. heavier. This increase of weight is due to the replacement 38.] HYDROGEN PEROXIDE. 53 of a certain number of oxygen molecules by the same number of ozone molecules. The volume that the ozone occupies in the gas mixture can be determined by absorbing it in turpentine. Suppose this to be v c.c., when reduced to normal pressure and temperature. The weight of this v. c.c. ozone can be represented by the weight of an equal volume of oxygen + a mg. and must be, therefore, (v X 1.43 + a) mg. ; 1.43 mg. being the weight of 1 c.c. oxygen at normal pressure and temperature. Hence the weight g of 1 c.c. ozone is 0X1.43 -fa 9= - In one of his experiments LADENBURG found a = 16.3 mg. and c\ f\n i> = 26.0 c.c., hence #=2.06 mg. 1 c.c. ozone thus weighs - 1.45 times as much as an equal volume of oxygen, or very nearly 1J times as much. The molecule of oxygen being O 2 , that of ozone must be represented by 63. In an oxidation by ozone the volume of the ozoniferous gas remains unchanged. Only the third atom in Os has oxidizing power, not all three atoms of the molecule. In ozone we have become acquainted with oxygen that is dif- ferent from the ordinary kind. This phenomenon is also seen in other elements; it is called allotropism. HYDROGEN PEROXIDE, H 2 2 . 38. This compound is usually prepared by treating barium per- oxide with dilute sulphuric acid: Ba0 2 + H 2 S0 4 - BaS0 4 -f H 2 2 . Barium Insoluble, peroxide. In a very concentrated state it can be obtained by direct distillation in vacuo of a mixture of sodium peroxide and sulphuric acid: NaA + H 2 S0 4 =Na 2 S0 4 + H 2 O 2 . Hydrogen peroxide is also formed in many other ways; e.g. together with ozone ( 36) in the slow oxidation of phosphorus; by the combustion of hydrogen, when the flame is cooled by a piece of ice. The formation of ozone has been often detected when hydrogen in the nascent state comes in contact with oxygen mole- cules. We suppose that in the moment just after hydrogen is set 54 INORGANIC CHEMISTRY. [33. free, its atoms have not yet united to form molecules, so that the individual atoms possess unusual chemical activity. This is the general conception of the status nascendi. Thus TRAUBE has observed the following instances of the production of hydrogen peroxide: Zinc filings, when shaken with water and oxygen or air, give hydrogen peroxide, since the zinc and the water generate a small quantity of hydrogen, which unites with the oxygen. Palla- dium-hydrogen behaves likewise when brought in contact with water and air. In this case it is the hydrogen released from the palladium that unites with the oxygen. Many metals, such as copper, lead and iron, yield hydrogen peroxide on being shaken with air and dilute sulphuric acid, for the same reason as in the case of zinc and water. Finally, the peroxide is formed in the electrolysis of water, when a current of air or, better, oxygen passes over the negative electrode (at which hydrogen is evolved). Hydrogen peroxide is also formed at very high temperatures from steam and oxygen ( 103); just as in the formation of ozone under the same conditions ( 36), a rapid cooling is necessary in this case also, else the compound decomposes. The formation of hydrogen peroxide in the combustion of hydrogen has been shown in the following way: A hydrogen flame was allowed to burn at the mouth of a bulb tube containing a little water. By means of a very rapid current of air the flame was blown into the bulb, causing a very sudden cooling of the mixture of steam and air. After a time the water in the bulb gave the tests for hydrogen peroxide. As a further analogy to the case of ozone it has been shown that the formation of hydrogen peroxide has no connection with the combustion, for on directing a fine stream of water upon an incandescent NERNST glower some hydrogen peroxide is generated in the water. Physical Properties. In the pure anhydrous condition hydrogen peroxide is a colorless, slightly viscid liquid, having a specific gravity of 1.4584 at 0, based on water at 4. (A density calculated on this basis is indicated by d.) It 'becomes solid at a low tem- perature and melts at 2. Chemical Properties. Hydrogen peroxide, when wholly free from impurities, especially from suspended particles of solid mat- ter, is rather stable and can be distilled in vacuo; when impure, it decomposes, however, into water and oxygen, as it also does in dilute solution. In the latter state it is more stable in the 38.] HYDROGEN PEROXIDE. 55 presence of traces of acid than in the presence of bases. It is an interesting fact that it decomposes rapidly in contact with powdered substances, apparently without acting upon them. Finely divided silver, gold, platinum (platinum black), and especially manganese dioxide decompose it with effervescence (due to escaping oxygen). Even rough surfaces have a disturbing effect; BRUHL observed, for instance, that a concentrated solution of hydrogen peroxide evolves oxygen when poured upon ground glass. All these actions must be regarded as catalytic accelerations of the ordinarily very slow decomposition of hydrogen peroxide. The effect of heat is here, as elsewhere, to accelerate the reaction; concentrated preparations, when warmed, often decompose so rapidly as to cause an explosion. The oxidizing action of hydrogen peroxide is an important chemical property. This is always due to the surrender of an oxygen atom, which effects the oxidation, while water remains. Lead sulphide, PbS, is oxidized by a weak solution of hydrogen peroxide' to lead sulphate, PbS04; sulphuretted hydrogen, H 2 S, is converted into water and free sulphur. Barium, strontium and calcium hydroxides, Ba(OH) 2 , Sr(OH) 2 and Ca(OH) 2 , are pre- cipitated by dilute hydrogen peroxide from their solutions as peroxides of the general formula MO 2 < ^aq. 1 The colorless solu- tion of titanium dioxide in dilute sulphuric acid is turned orange-red by hydrogen peroxide lemon-yellow by traces of it on account of the formation of yellow trioxide, TiOs. This is a delicate test for hydrogen peroxide. Other tests are found in the following oxidation reactions: Potassium iodide starch-paste is at once turned blue by hydrogen peroxide in the presence of a little ferrous sulphate, FeSO4. The ferrous sulphate carries the active oxygen of the hydrogen per- oxide to the potassium iodide. As a result two atoms of iodine are set free, the ferrous sulphate being oxidized at the same time. According to MANCHOT a higher oxide of iron is formed in this reaction. A very characteristic reaction is this: Chromic acid solution (H 2 CrO 4 ), when treated with hydrogen peroxide, is changed to a higher oxide (see 295) which is blue in aqueous solution and may be taken up by ether if shaken with the latter. This test is, however, less delicate than the two preceding ones. 1 Aq. (aqua), a frequently used abbreviation for water of crystallization or hydration. ** _ * 56 INORGANIC CHEMISTRY. [ 38- A third group of chemical effects of hydrogen peroxide depends on its reducing power. When silver oxide is introduced into a solution of hydrogen peroxide, a vigorous evolution of oxygen occurs, water and metallic silver being formed at the same time. Potassium permanganate solution loses its color when mixed with a hydrogen peroxide solution acidulated by sulphuric acid, oxygen being given off rapidly: 2KMnO 4 + 3H 2 S0 4 + 5H 2 O 2 = K 2 S0 4 + 2MnS0 4 + 8H 2 + 50 2 . The brown peroxide of lead, Pb0 2 , is reduced to reddish-yellow lead oxide, PbO. Ozone and hydrogen peroxide yield water and oxygen; when dilute, they are, however, able to exist side by side. There is a test for hydrogen peroxide, depending on its reducing power, which is even more delicate than those described above. A mixed solution of ferric chloride and red prussiate of potash has a red color. On the addition of hydrogen peroxide Prussian blue is precipitated. Traces of the peroxide turn the solution green. The reaction fails in the presence of free acid. The ability of so powerfully oxidizing a substance as hydrogen peroxide to act also as a reducing-agent can be explained as follows : One of its two oxygen atoms must be loosely joined to the mole- cule, since it is easily given up. All the substances which are reduced by hydrogen peroxide, ajso have one loosely held oxygen atom; silver oxide, potassium/permanganate, ozone and others give up their oxygen at rather low temperatures. It is therefore possible that the mutual attraction of the oxygen atoms, which tends to make them form oxygen molecules, is stronger than the force by which they are held in hydrogen peroxide on the one hand, and the respective oxygen compound on the other. Uses of Hydrogen Peroxide. The colors of old paintings are often restored by means of it. The darkening of them is due in many cases to the transformation of white lead sulphate, PbS0 4 , to black lead sulphide. The latter is readily oxidized by hydrogen peroxide back to white lead sulphate. Hydrogen peroxide is also of value in bleaching ivory, silk, feathers, hair, bristles and sponges. It is also important in analysis. For therapeutic purposes at 30% solution of hydrogen peroxide is pre- pared by MERCK which is perfectly pure and is obtained by vacuum dis il- lation from a more dilute solution. Before use it is strongly diluted. It 40.] DETERMINATION OF MOLECULAR WEIGHT. 57 has the advantage of not being subject to decompos'tion. The concentra- tion of a solution of hydrogen peroxide is generally expressed in the volumes of oxygen that it can evolve; thus, for a 3% solution it is ten volumes. 39. The composition of hydrogen peroxide was established by THENARD as early as 1818. He first concentrated it in a vacuum and then introduced a weighed amount of it, enclosed in a vial, into a graduated barometer- tube over mercury. The vial was then broken and its contents decomposed by heating the tube from without or allowing finely powdered manganese dioxide to rise in the tube. It was thus found that very nearly 17 parts of hydrogen peroxide by weight yield 8 parts of oxygen, water being also formed. One atom of oxygen (16 parts by weight) is therefore obtained from 34 parts of hydrogen peroxide, the remaining 18 parts forming water; in other words, hydrogen peroxide is 1 molecule H 2 O + 1 atom 0. The peroxide therefore contains one atom of oxygen to every hydrogen atom. Its simplest formula (the so-called empirical formula) is then HO. Whether this also expresses the molecule or whether the latter is a multiple of it, remains to be determined by finding the molecular weight, inas- much as every compound of the general formula (H0) n possesses the same composition, viz., 16 parts by weight of oxygen to 1 part of hydrogen. On account of the instability of this substance its vapor density cannot well be determined. It was therefore necessary, in finding its molecular weight, to follow another course, which is based on the properties of dilute solutions. In this manner the molecule of hydrogen peroxide was found to possess the formula H 2 O2. The method referred to is explained in the following sections. MOLECULAR WEIGHT FROM THE MEASUREMENT OF THE DEPRESSION OF THE FREEZING-POINT AND ELEVATION OF THE BOILING-POINT. 40. Certain membranes possess the peculiar property of allow- ing a solvent, e.g. water, to pass through, but not the dissolved substances. They bear the name "semi-permeable membranes.''^ This property appears to depend not so much on a sort of sieve action as upon the ability of the membrane to dissolve, or else to absorb or loosely combine with, the solvent on one side and release 58 INORGANIC CHEMISTRY. [40- it again on the other, while the dissolved matter remains behind. One of the ways of obtaining a semi-permeable partition is by dip- ping a porous cup such as is used in galvanic cells containing a solution of yellow prussiate of potash into a solution of bhie vitriol. A thin layer of copper ferrocyanide is thus formed in the wall of the cup, making it semi-permeable. If a dilute sugar solution, salt solution or the like be poured into such a cup and the cup placed in a dish of water, it will be found that the dissolved substance does not diffuse through this sort of a partition. The water goes through, however, for if the cup be closed with a perforated stopper through which a glass tube passes and then dipped deep enough under water so that the entire cup is submerged, the water will be seen to rise slowly in the tube till it reaches a definite height above the level outside. The pressure exerted by this column of liquid is called the osmotic pressure of the solution. If a tight-fitting piston were inserted in the cup, the force which one would have to exert on it to prevent the infiltration of the water would be equal to the pres- sure of the column of liquid, for the water continues to rise in the tube till the pressure of the column prevents the entrance of any more. According to researches of VAN'T HOFF the osmotic pressure of dilute solutions,' like the pressure of gases, obeys the law of BOYLE and the expansion law of GAY-LUSSAC. If the pressure exerted at a certain temperature by a kg. of a gas in a vessel be p, the pressure which na kg. of the gas at the same temperature exerts in the same vessel is np. The concentration, i.e. density, of the gas has been multiplied n-fold. If the osmotic pressure of a solution containing a per cent of a substance be determined and found to be p, the osmotic pressure will be np, if an na per cent solution of the same temperature be taken, i.e. if the concentration be n times as great. An investigation of the pressures which a gas of constant volume exerts at the absolute temperatures T\ and T 2 shows that these pressures bear to each other the ratio TI : T 2 . The same proportion is observed when the osmotic pressure of a solution of constant concentration is measured at the same absolute temperatures as above. 41.] DETERMINATION OF MOLECULAR WEIGHT. 59 41. The laws of osmotic pressure find experimental verification in measurements which were made by PFEFFER previous to VAN'T HOFF'S enunciation of the laws, PFEFFER investigated dilute sugar solutions and used an apparatus not unlike the one just described. The gas laws are expressed by the equation ( 34) PV=RT, (1) in which P represents the pressure, V the volume, and T the absolute temperature of a gas, while R is a constant. The volume, V, is inversely proportional to the concentration, according to the above definition; therefore ^ may be substituted for V, if C indicates the concentration. The above equation then becomes p C~ RT ' or, at a constant temperature, p = Const. This equation must also be applicable to osmotic pressure.. This was really the case in PFEFFER' s measurements of aqueous sugar solutions of different concentrations, as may be seen from the following brief table. The temperature varied between 13.5 and 16.1, and hence was not perfectly constant: I 1% 535 mm. 535 2% 1016 " 508 4% 2082 " 521 6% 3075 " 513 p The differences in the values of .^r must be ascribed to the variations of temperature and the unusual experimental difficulties which attend such measurements. From equation (1) it also follows, when V (or C) is a constant, that p -= Const. 60 INORGANIC CHEMISTRY. { 41- This conclusion, too, was demonstrated experimentally by PFEFFER in the case of sugar solution, as may be seen from the following table, A one per cent solution was used: P T P. r 510 287.15 1.78 520.5 288.5 1.80 544 305 1.78 567 309 1.83 VAN'T HOFF has further shown that the numerical value of the osmotic pressure is the same as that of the gas pressure; that is to say, when a definite amount of a substance in the gaseous state occupies a given volume, the gas pressure which it exerts is just as great as the osmotic pressure which would be produced if the same amount of substance were dissolved in a liquid making the same volume of solution. The measurements of PFEFFER also furnished experimental proof of this. He found that a 1% sugar solution at 7 exerts a pressure of of an atmosphere. If there is really equality between osmotic pressure and gas pressure, or, in other words, if the law of AVOGADRO for gases is also applicable to dilute solutions, the constant R of the equation PV=RT must have the same value for solutions as for gases. P in the above case was found to be f of an atmosphere, or |X 1033.6=689.0 g. A 1% sugar solution contains 1 g. sugar in 100.6 c.c. As the molecular weight of this substance is 342, the volume V which contains 342 g. is 7=100.6X342. T= 273 + 7 =280. Substituting these figures in PV R = -jr, we have #=84664. The close agreement of the two values of R (compare p. 48) proves the equality of gaseous and osmotic pressure. 42. It follows from the preceding that AVOGADRO 's law must also hold for dilute solutions. Assuming that an equal number of molecules of different substances are dissolved in equal volumes at the same temperature, we know from the equality of gas pressure and osmotic pressure that the various substances will exert the same osmotic pressure ; inversely, in equal volumes of solution hav- ing the same temperature and osmotic pressure there is the same number of molecules. This is a very important extension of AVOGADRO 's law. We are thus able not only to compare the weights of equal gas volumes at the same temperature and pressure and calculate therefrom the molecular weight, but we can apply the same principle to solutions, 42.] DETERMINATION OF MOLECULAR WEIGHT. 61 since we know that in solutions of the same temperature and the same osmotic pressure the quantities of the dissolved substances contained in equal volumes of solution are to each other as their molecular weights. Just as it is possible to ascertain the molecular weights of gaseous bodies from determinations of temperature, pressure, weight and volume, it is also possible to find those of substances in dilute solution by measuring the volume of liquid, the tempera- ture, the quantity dissolved and the osmotic pressure. The molecular weights of all substances that dissolve in some liquid or otfrer can be determined in this way, and, since the number of soluble substances is very large, there are not a few whose molec- ular weights were first determined in this way. In working out this method, however, there is a practical difficulty. The osmotic pressure is very hard to measure directly. This would render the whole method of little value, if it were not for the fact that the calculation only requires that it be known whether two solutions have the same osmotic pressure, not the absolute amount of the latter; the law of AVOGADRO simply requires the equality of volume, of temperature and of pressure (osmotic or gas), without regard for the absolute value of these factors (between certain limits). Now, it is easy to measure magnitudes which are proportional to the osmotic pressure, and from which it may be seen whether equality of osmotic pressure exists or not. These magnitudes are the depression of the freezing- point and the elevation of the boiling-point. An explanation of these terms is perhaps necessary: When a substance is dissolved in a liquid the maximum tension of the vapor is less above the solu- tion than above the pure solvent at the same temperature, for the particles of the dissolved body attract the molecules of the solvent, hindering the formation of vapor on the one hand, and, on the other hand, facilitating the return of vapor molecules into the liquid. This lowering of the vapor pressure necessarily causes a depression of the freezing-point and an elevation of the boiling-point, as may be proved by the following diagrams. In Fig. 18, abc represents the vapor-pressure curve of a solvent in the neighborhood of its freezing-point 6; the part ab gives the pressures for the frozen matter; the part be for the liquid solvent. This latter part is always more nearly horizontal than the former, as has been proved 62 INORGANIC CHEMISTRY. [42^ both experimentally and theoretically. The freezing-point of a liquid is that temperature at which the solid and liquid states can exist side by side indefinitely. This condition requires that the solid and the liquid substance have the same vapor tension. If, for instance, the vapor tension of the solid were greater than that of the liquid, we should have, at a constant temperature, the vapor given off from the solid condensing to a liquid and the former grad- FIG. 18. FIG. 19. ually turning into the latter. Inversely, if the vapor tension of the solid were less than that of the liquid, the entire liquid would, under similar conditions, solidify. The freezing-point b can thus be regarded as the intersection of the vapor-pressure curves ab and be of the solid and the liquid, re- spectively. Let us now consider the curve &V of a solution. Its vapor pressure is lower than that of the pure solvent, so its inter- section with the curve ab must lie more to the left, that is, its freezing-point is lowered. On the other hand, the boiling-point of a solution is that temperature at which the tension of its vapor equals one atmosphere. If Od in Fig. 19 represents this tension, a line dd f parallel to the axis of abscissas will intersect the vapor- pressure curve ac of the pure solvent at a lower temperature, than it will the curve a'c' of the solution. The latter must, therefore, have a higher boiling-point. 43. The connection between these magnitudes and the osmotic pressure will be better understood after the following considerations: 1. Solutions in the same solvent, separated by a semi-permeable parti- tion, can only be in equilibrium when they are isotonic, i.e. when they exert the same osmotic pressure. 43.] DETERMINATION OF MOLECULAR WEIGHT. 63 Let us imagine the solutions in an apparatus consisting of two cylinders that are connected by a tube containing a semi-permeable partition. In both cylinders the level of liquid is kept at the same height constantly by adding or removing some from time to time. The solution with the greater osmotic pressure will extract solvent from the other, for, because of the stronger pressure which the dissolved molecules exert upon the free surface of the liquid, the first solution will endeavor to increase in volume at the expense of the second. Equilibrium will be established so soon as the same pressure is exerted by the dissolved molecules upon the unit area of the free surfaces of the liquids from both sides of the semi-permeable partition; in other words, when the solutions are isotonic. 2. Isotonic solutions with the same solvent have the same vapor tension at the same temperature. \ 1 A B H FIG. 20. The proof of this statement lies in the contradiction to which the assumption that isotonic solutions have unequal vapor tensions leads. The accompanying diagram, Fig. 20, represents a closed vessel, that is separated by the semi-permeable partition HH into two parts, which contain the isotonic solutions A and. B. Near the top the two parts are connected with each other. Assuming that the vapor tension of A is greater than that of B, vapor must pass out of A and condense in J5; the result is that A becomes more concentrated, B more dilute, and they are no longer isotonic. In such a case, according to the first prin- ciple, the solvent would then begin to pass through HH from B to A. The assumption of perpetual motion which is thus made necessary can only be avoided by supposing that the vapor tension is the same. 3. Isotonic solutions with the same solvent have the same freezing- point. Let us again take the same apparatus, containing, in addition to the isotonic solutions A and B, a piece, (7, of the solvent in the solid state (Fig. 21). Let us also assume that A and C have the same vapor tension. We then have, according to definition (see 42), the tern- 64 INORGANIC CHEMISTRY. [ 43,. perature of the freezing-point of A. However, if A and B are isotonic, they have the same vapor tension. B will, therefore, have the same vapor tension as C. Hence B and C must also be at their only coexist- ence temperature, the freezing-point of B. At their freezing-points A and B } therefore, have the same temperature as C, i.e. they possess the same freezing-point. 4. Isotonic solutions with the same solvent have the same boiling-point. As we saw in 42, the boiling-point of a solution is that temperature at which the tension of its vapor equals one atmosphere. Two solutions with a common solvent, therefore, have the same vapor tension at their boiling-point. Now, it was shown above that solutions having the same temperature and vapor tension are isotonic. If these solutions have the same vapor tension (at their boiling-point) and are isotonic, they must also have the same temperature. Since, as has just been demonstrated, isotonism requires like freezing- points and boiling-points, it is evident the depression of the freezing- point and elevation of the boiling-point must be the same in isotonic solutions with the same solvent. In the depression of the freezing-point and the elevation of the boiling-point we thus have a means of deciding whether solutions are isotonic. Use is made of this fact for the determination of molecular weights in the following way: The freezing-point of a liquid, e.g. water, acetic acid, phenol, etc., is first determined. Thereupon a gram molecule of a substance whose molecular weight is known is dissolved in a given weight (hence also in a given volume) of the liquid. A depression of the freezing-point is observed. This depression will always be the same, no matter what the substance is that is dissolved in the liquid, providing that one gram molecule is dissolved in the same volume of liquid. The depression of the freezing-point for one gram-molecule of solute is thus a constant for the solvent. Now if we prepare a 1% solution of a compound whose molecular weight, M, is unknown and measure the freezing-point depression, A, we have A M= Constant. This formula is also applicable to the elevation of the boiling-point, as can be readily seen. M is the only unknown and can therefore be calculated. 44.] BROMINE. 65 When water is used as the solvent, the product of the depression A of the freezing-point of a 1% solution and the molecular weight M has been found from numerous observations to be 19. We have therefore for water For hydrogen peroxide, the depression of the freezing-point of a 3.3% aqueous solution was found to be 2.03. This would correspond to 2 03 -^-3- =0.615 for a 1% solution; hence A =0.615, from which it follows that 19 the molecular weight is Since the formula HO corresponds to a molecular weight of 17, H 2 2 to one of 34, and the latter number is the nearer to the molecular weight found by experiment, we conclude that hydrogen peroxide has the doubled empirical formula H 2 2 . The constants for the freezing-point depression (molecular depression) and for the elevation of the boiling-point (molecular elevation) of some compounds that are well adapted for these determinations are given in ORG. CHEM., 13. The freezing-point method for determining molecular weight is called the cryoscopic method, while the boiling-point method is known as the ebullioscopic method. Apparatuses for the easy and exact determination of the depression of the freezing-point and elevation of the boiling-point are described in ORG. CHEM., 14 and 15. BROMINE. 44. This liquid element does not occur free upon the earth because of its strong tendency to form compounds. In com- bination with metals it is found in the salts of sea-water. It was discovered in the latter by BALARD in 1824. Bromides occur in rather large amounts in the so-called Abraum-salze of the Stassfurt salt-mines, and also in considerable quantities in the brines of many salt wells, notably those of Michigan. In the neighborhood of Stassfurt, Germany, there are extensive beds of rock-salt (halite). Above the halite are found layers of other salts (called " Abraum-salze " because they have to be removed in order to get at the halite). These salts were formerly rejected as worthless, but they 66 INORGANIC CHEMISTRY. [ 44^ have since been found to be rich in potassium salts, bromides and other valuable minerals, so that the " waste salts " of former days are now the leading source of many commercially and scientifically important com- pounds. The purification of these Stassfurt salts is accomplished by solution hi water and partial evaporation of the latter. Various substances crystallize out, while the remaining liquid ("mother- liquor ") still contains the most soluble salts, among which is mag- nesium bromide, MgBr 2 . From this mother liquor the bromine is obtained by the use of chlorine, which sets bromine free from bromides, thus: MBr + Cl = MCI + Br . (M = Metal.) The process employed is an application of the principle of the counter-current (15). The mother-liquor is allowed to flow down through a tower filled with round stones, so that the exposed sur- face of the liquid is greatly enlarged. A current of chlorine is passed into the tower from below, and, as it rises, the gas is in con- stant touch with the bromide liquor, the most concentrated gas being in contact with liquor which has already yielded the greater part of its bromine, so that practically all the bromine is thus easily obtained. The bromine prepared in this way always con- tains a little chlorine, from which it is freed by distillation over finely powdered bromide of potassium. Another method, common in the United States, of obtaining the bromine from the mother-liquor is by distilling the latter with manganese dioxide (or potassium chlorate) and sulphuric acid, corresponding to the method of making chlorine ( 25). Still another method is to electrolyze the bromide solution and boil off the bromine. The bromine thus obtained still contains a little water. It is dried by shaking it with concentrated sulphuric acid and then dis- tilling again. Physical Properties. Bromine is a liquid at ordinary tempera- tures; it is the only element, excepting mercury, that displays this property. It solidifies at 7.3 and boils at 59. It is dark brown, and is transparent only in thin layers. At the temperature of 45.] HYDROBROMIC ACID. 67 liquid hydrogen (20.5 absolute) it becomes colorless; MOISSAN showed the same to be true of chlorine and fluorine as well. It is quite volatile at ordinary temperatures, giving off brown fumes of an extremely irritating and disagreeable odor, whence its name (/3pSj^os = stench). Sp. g. =3.1883 at 0. 100 parts of water dissolve 3.5 parts of bromine. The addition of potassium bromide to the water increases its solubility a little. Its vapor density is 79.96 (0 = 16). The chemical properties of bromine are completely analogous to those of chlorine, but the action of the former is less energetic. While, for instance, chlorine combines with hydrogen in the day- light at ordinary temperatures, bromine does not. Its affinity for many elements is, however, very strong. It reacts vigorously with phosphorus; and powdered arsenic and antimony take fire when sprinkled upon bromine. It is an interesting fact that of the two closely related alkali metals, potassium and sodium, the former reacts vigorously with bromine, while the latter does not react with it at all at ordinary temperatures. The bromine molecule consists of two atoms ; for, since its vapor density is 79.92 (see above), its molecular weight must be 159.84. Inasmuch as a gram molecule of no one of the very numerous bromine compounds contains less than 79.92 g. bromine, but often simple multiples of this quantity, its atomic weight is taken to be 79.92, based on 0=16. The molecule, therefore, contains 159.84 79.92 2 atoms. HYDROGEN BROMIDE, or HYDROBROMIC ACID, HBr. 45. This gaseous compound can be obtained by direct syn- thesis from its elements; for this purpose it is necessary to pass hydrogen, together with bromine vapor, through a hot tube con- taining platinum gauze. This is the most practical method of manufacturing it. Hydrobromic acid can also be obtained by the action of hydro- gen on bromine compounds. Silver bromide, AgBr, for example, is reduced by hydrogen at a high temperature to metallic silver with the formation of hydrogen bromide. 68 INORGANIC CHEMISTRY. [ 45- On the other hand, it is also formed by the action of bromine on hydrogen compounds. For this purpose numerous organic com- pounds can be used. For example, bromine reacts with naph- thalene, CioHg, at ordinary temperatures to form hydrogen bromide, somewhat impure, however, from the presence of organic sub- stances. Hydrogen bromide is also produced, together with free sulphur, when hydrogen sulphide 'is led into bromine under water: H 2 S+Br 2 = S+2HBr. Hydrobromic acid may also be prepared by the decomposition of a bromine compound with a hydrogen compound, phosphorus pentabromide, PBr 5 , and water being employed : PBr 5 + 4H 2 O = 5HBr Phosphoric acid. As phosphoric acid is not volatile, but the desired substance is, the two products of the reaction can be easily separated. Physical Properties. At ordinary temperatures hydrogen bro- mide is a gas. It can be condensed, by cooling, to a liquid which boils at 64.9 (under 738.2 mm. pressure), and, by still farther cooling, to colorless crystals, which melt at 88.5. It has a pungent odor and a sour taste. In contact with moist air it forms dense clouds, like hydrochloric acid ( 29). It is very soluble in water, 1 vol. water dissolving about 600 vols. at 10; its solu- bility is thus even greater than that of hydrochloric acid. Chemical Properties. Here, too, the acidic nature is strongly displayed. Various metals, such as zinc and magnesium, are acted upon by hydrobromic acid, forming a salt and free hydrogen. The most of its salts are soluble in water; silver bromide, however, is insoluble and lead bromide difficultly soluble. A very high temperature is required to decompose hydrogen bromide into its elements. The composition of hydrobromic acid can be determined in the same way as that of hydrochloric acid. Since its vapor density is 40.46, it has a molecular weight of 80.92. The atomic weight of bromine being 79.92 (O= 16), it follows that the formula of hydro- bromic acid must be HBr. Moreover, the dry gas can be decom- 46.] IODINE. 69 posed with sodium amalgam, whereby it is found that half of its volume consists of hydrogen; this confirms the above molecular formula. IODINE. 46. This element, a crystalline solid, was discovered by COUR- TOIS in 1812, but its elementary nature was first recognized in 1815 by GAY-LUSSAC. Like chlorine and bromine, it does not occur in the free state, but is frequently found in nature in combination with some metal. An important source of iodine compounds is the mother-liquor ( 44) remaining in the purification of Chili saltpetre; another, the ash of seaweeds, known in Scotland as kelp and in Normandy as varec, which contains iodides. The extraction of the iodine is accomplished either by passing chlorine into the solution or by distilling with manganese dioxide and sulphuric acid in the same manner as for bromine and chlorine. The commercial iodine is purified by warming it gently with the addition of a little potassium iodide, the iodine subliming in the pure state, free from traces of chlorine and bromine that may have been present. Finally, it is dried in a desiccator over sul- phuric acid. Physical Properties. Iodine forms tabular crystals of a dark gray metallic lustre. Its specific gravity is 4.948 at 17. It melts at 114.2, and boils under 760 mm. pressure at 184.35. Its vapor is characterized by a beautiful dark-blue color, which gave the element its name (io tdrjs = violet). In water it is only slightly soluble enough, however, to color the water yellow. It dissolves easily in a solution of potassium iodicje, the latter being turned brown. In various other liquids, such as alcohol, ether, carbon disulphide and chloroform, iodine is also easily soluble. It is a peculiar fact that the alcoholic and the ethereal solutions are brown, while the solutions in carbori disulphide and chloroform are violet; other solvents, e.g. benzene, give solutions of an inter-* mediate color. The explanation of this diversity of color is that in the brown solutions the iodine has formed a compound with the solvent, whereas in the violet solutions, which have very nearly the same color as iodine vapor, the element exists in the free state. 70 INORGANIC CHEMISTRY. [46- This conclusion is reached in various ways; one is from the fact that the addition to the violet iodine solution of a small amount of a liquid that dissolves iodine with a brown color does not alter the freezing-point of the solution. The number of molecules free to move has thus not been changed by this addition; in other words, the iodine has united with the added liquid. The vapor density of iodine is 8.72 (air = 1) at about 600. As the temperature rises, it grows steadily smaller, however. At 1500 we find it is reduced to almost half of what it is at 600. Later ( 49) we shall have occasion to discuss this phenomenon, known as dissociation, which has been observed with many substances. 47. The chemical properties of iodine resemble very strongly those of chlorine and bromine. Its affinity for other elements is in general weaker, however, than that of the two halogens men- tioned. It combines with metals, e.g., mercury, directly to form salts (iodides) . A characteristic test for iodine is the intense blue coloration which it imparts to starch solution; the slightest traces of iodine can be thus detected. The blue color disappears on boiling and reappears on cooling, provided the boiling was not too prolonged. The blue substance formed from iodine and starch is not a com- pound, but is to be regarded as a product of the absorption of iodine by starch; for the quantity of iodine taken up by starch from the solution in KI depends largely on the concentration of the solution and does not become constant, even when a large excess of solution is present. It is the reaching of a constant ratio that characterizes chemical combination. The molecule of iodine, investigated by the same method as was employed with bromine, is found to consist of two atoms at 600; hence the formula is I 2 . Above 1500 it must contain only one atom, for the vapor density is only half as great. 48.] HYDROGEN IODIDE. 71 HYDROGEN IODIDE, or HYDRIODIC ACID, HI. 48. This compound can be obtained by direct synthesis from its elements, and that is really the best method for preparing it in a perfectly pure state. For this purpose hydrogen and iodine vapor are conducted together over heated platinum-black, which accelerates their combination. Hydrogen iodide can also be obtained by the reaction of iodine with hydrogen compounds. Organic hydrogen compounds are preferable, especially colophonium and copaiva oil. This method is also used for the laboratory preparation of the gas, but the hydrogen iodide thus obtained is more or less adulterated with organic substances. When iodine acts on hydrogen sulphide water hydrogen iodide is formed, sulphur being liberated ( 45). As an example of the action of hydrogen on an iodine compound, we may mention the reduction of silver iodide by hydrogen, from which hydrogen iodide results. Finally, the action of an iodide on a hydrogen compound is illustrated by the decomposition of a phosphorus iodide, Pis or PI 5 , by water. As was explained in 45, it is possible to use phosphorus, iodine and water. This method, with some variation or other, is the preferable one for the preparation of hydriodic acid. According to GATTERMANN, it is best to first add yellow phosphorus (4 g.) in very small pieces to 44 g. iodine, and then decompose the result- ing compound with a little water. In order to remove the free iodine from the hydrogen iodide formed, the gas is allowed to pass over red phosphorus. The decomposition of the halogen salt by sulphuric acid is even less available for the preparation of hydriodic than for hydrobromic acid, since the former is more easily decomposed by sulphuric acid than the latter. Physical Properties. Hydrogen iodide is a colorless gas, whose specific gravity is 62.94 (H = l). It fumes strongly when exposed to the air, and possesses an acid reaction and a pungent odor. At and 4 atmospheres pressure it condenses to a colorless liquid, which boils at 34.14 under a pressure of 730.4 mm. The melt- ing-point of the solid is 50.8. Hydrogen iodide is very soluble 72 INORGANIC CHEMISTRY. [48- in water; 1 vol. H 2 at 10 dissolves 425 vols. HI. This solution fumes strongly, and turns dark brown after a time, because of the liberation of iodine. Chemical Properties. Hydrogen iodide has all the charac- teristics of an acid. With metals it forms salts (iodides), hydrogen being given off. These are almost all soluble, with the exception of the iodides of silver, and mercury. Lead iodide is slightly soluble at ordinary temperatures. In addition to its acidic character, hydriodic acid possesses another property, which is not found in hydrochloric and hydrobromic acids. Since it splits up readily into hydrogen and iodine, it can act as a strong reducing-agent, especially at high temperatures. It has already been remarked that the aqueous solution of the gas turns brown, iodine being set free by the oxidizing action of the air; this change is greatly aided by the influence of light. At a high temperature hydrogen iodide is decomposed into H2 and 12, as is shown by the appearance of the violet iodine vapor. In organic chemistry, particularly, frequent use is made of the reducing power of this acid. Formula of Hydriodic Acid. The vapor density of this sub- stance has been found to be 62.94. Its molecular weight is there- fore 125.88. The atomic weight of iodine being 126.92 (0=16), it is seen that the molecular weight corresponds very closely to the formula HI =127. 92, and no other formula is possible. DISSOCIATION. 49. When hydrogen iodide is subjected to a slow increase of temperature, it commences to decompose at a definite temperature slightly above 180 into hydrogen and iodine vapor. As the heat- ing continues, the decomposition grows gradually greater till a point is finally reached when the gas mixture contains only the individual elements. If the mixture is then slowly cooled, the same stages are passed through in inverse order, so that the degree of decomposition at any one temperature is found to be the same, no matter whether the temperature was approached from above or below, it being only necessary that the particular temperature should be maintained for a sufficient length of time in both cases. The phenomenon just described is to be observed with a great 49.] DISSOCIATION". 73 many substances. It is called dissociation, and was first studied in 1857 by H. SAINTE-CLAIRE DEVILLE. The degree of decomposition of hydrogen iodide is, as we have seen, always the same for a definite temperature. It necessarily follows from this that if one starts with the uncombined elements, hydrogen and iodine, and heats them together long enough at a certain temperature, a gas mixture of exactly the same compo- sition will be formed as that resulting from the decomposition of the hydrogen iodide at the same temperature. This is confirmed by experiment; e.g. equivalent amounts of iodine and hydrogen were heated in a sealed vessel by exposing it to the vapor of boiling sul- phur (445). The amount of hydrogen iodide finally produced was 79.0%, i.e. 21.0% of the gas mixture remained uncombined. Again, when a like vessel filled with hydrogen iodide was heated to the same temperature, it was found that 21.5% had decomposed a figure very close to that obtained in the preceding experiment. Such reactions, which lead to the same result, no matter whether we start with the one set of substances (H2 + I2) or with the other (2HI), are called reversible reactions. When the final stage is reached, the sets, or " systems," are said to be in equi- librium with each other. If we have a system of substances A+B+ . . . , which is par- tially changed into another system P+Q+ . . . , the equilibrium between the two systems is expressed by the sign <=; thus: We saw that, in the preparation of hydrogen iodide, platinum black is used because it accelerates the combination. Neither this nor any other catalyzer, however, changes the proportional extent to which combination takes place. For instance, experiments have shown that at 350 18.6% of the hydriodic acid is decomposed when no platinum- black is present, and that in the presence of this catalyzer the decom- position reaches 19%; these two figures are alike within the limits of experimental error. There is a theoretical reason why this must be so. If the catalyzer influenced the equilibrium, we could realize combination and decomposition by alternately adding and removing the catalyzer. Under constant conditions of temperature the system would absorb energy in one instance and evolve it in the other. The energy obtained in the latter case could be used to do work. Work would thus be gained 74 INORGANIC CHEMISTRY. [49- from a system remaining at constant temperature, but according to the principles of thermodynamics this is impossible, since the production of work is always accompanied by a fall of temperature. The question now arises, how such an equilibrium comes about, and why the decomposition of a compound that begins at a cer- tain temperature does not complete itself. To this the kinetic theory of gases furnishes a satisfactory answer. According to this theory the molecules of gases are constantly in motion. While a constant mean velocity of the molecules may be assumed to exist for every temperature, the velocities of the individual molecules must be considerably different, because of their very frequent collisions with each other. The atoms of a molecule must also be supposed to be capable of changing their respective positions, for the repeated collisions of the molecules displace the atoms from their positions of equilibrium. These movements of the atoms are the more violent the greater the velocity of the molecules. It is easy to conceive that they may at last become so violent as to throw the atoms out of their sphere of mutual attraction. The molecule is thus broken up. In a body of gas at a definite temperature this will> however, only occur in those molecules whose velocity is above certain limits; hence we see a reason for partial decomposi- tion. The explanation of partial combination is exactly analogous. The atoms set free from the molecules of the elements enter into the spheres of mutual attraction, and if their velocities are not great enough to resist the attraction, the different atoms unite. In the case of the formation and decomposition of hydrogen iodide this can be conceived as follows: Two HI molecules meet in such a state of atomic movement that the H atoms enter the spheres of attrac- tion of each other, and the I atoms likewise, so that H 2 and I 2 are formed. On the other hand, these molecules H 2 and I 2 may again meet in such a way that each H atom enters the sphere of attraction of One of the I atoms, whereupon two HI molecules are formed. According to the above the state of equilibrium is to be ac- counted for by supposing that in the unit of time just as much passes from the one system into the other as vice versa. Until the state of equilibrium is reached, the amounts which pass from one system into the other in the unit of time are unlike. 50.] DISSOCIATION. 75 In order to define more clearly this condition of equilibrium we must introduce the concept of reaction velocity. By this term we understand the number of moles transformed from one system into the other in the unit of time. Suppose that in the unit of volume (one liter, for instance) there are a moles of a substance A, which can undergo a chemical change into a substance B. If in the unit of time (say one minute) moles of A are converted into J5, the reaction velocity S will be expressed by . Incase, n there are originally only Ja moles of A per liter, experiments have shown that the number of moles converted per minute is zr . We thus perceive that the reaction velocity is proportional 11 to the number of gram-molecules per liter, or, in other words, to the concentration. This is a principle of very wide application 4 , it is ordinarily called the law of chemical mass action. It finds a general expression in the equation in which K is a constant factor, the reaction constant, or velocity constant. 50. Let us assume that the molecules of a compound are dis- sociated by heat into two others, the process being expressed by in which A, B and C represent single molecules. Of the sub- stance A, a gram-molecules were originally present, but in the course of a definite time, t, x gram-molecules have undergone the above decomposition. The problem is to express the reaction velocity at any moment. At the beginning (during the first minute) the reaction velocity is proportional to a; after the time t, when the concentration has fallen to a x, the amount converted during the succeeding minute will be proportional to a x. The reaction velocity thus constantly diminishes. This being the case, it is evident that s=k-a and s' =k(a x) are not the true expressions, respectively, for the reaction velocity in the first minute and in the minute following the time t. They would be correct, if the con- centration remained constant during these minutes instead of 76 INORGANIC CHEMISTRY, [50- diminishing, as it does in reality. However, we can approach the real velocity by considering not one minute but a very small fraction of this unit of time, which we will call At; the smaller At is taken the less is the concentration change. Supposing that the quantity of A which is transformed in this very short period Ax At is Ax, the expression -j- must be very close to the real velocity, because it indicates the quantity transformed in a unit of time so small that the concentration scarcely diminishes during it. We approach the true velocity nearer and nearer according as we take At smaller and smaller) and when it is made infinitely Ax small -T- becomes the exact expression of the velocity. It is customary to express such infinitely small quantities by the letter d, thus: . The mathematical expression for the velocity at a time t, when the concentration is a x, thus becomes *--*(-,), .. K being the velocity constant. The above reaction is termed unimolecular. When two (like or different) molecules react with each other, the reaction is called bimolecular; it may be represented by the equation A+B=C or =C+Z> + ... The equation for the reaction velocity is different in this latter case. Assuming that originally a gram-molecules of A and b gram- molecules of B take part in the reaction, and that x gram-molecules of A and of B are decomposed at the end of the period t, there must be, respectively, ax and b x of the two substances present at this moment. The reaction velocity will then be proportional to the product of these quantities, thus: in which K f is again a constant ; for suppose that there were only one molecule of A present; the possibility of its reacting with a 51.] DISSOCIATION. 77 molecule of B would then be proportional to the number of mole- cules of B. When there are a x molecules of A this possibility becomes ax times as large. It is assumed in the above that the temperature is constant. We shall see in 104 that this factor has a great influence on the reaction velocity. 51. The manner of expressing the condition of equilibrium is now plain. Assuming that the reaction velocity of the one system is S and that of the other S', equilibrium must exist when S=S'. The state of equilibrium may therefore be defined as that state in which the reaction velocities of both systems have become equal. Let us apply these considerations to the dissociation of hydro- gen iodide. This may be expressed by If a gram-molecules HI per unit volume are present originally and f of these are decomposed after a given period, the reaction velocity at this moment (since in this case a = b, and the reaction is evi- dently bimolecular) is S=C(a-x) 2 , C being a constant. x From the x gram-molecules of hydrogen iodide gram-mole- cules of hydrogen and an equal amount of iodine have been formed. The velocity of formation of HI from H 2 and I 2 is therefore ex- pressed by the equation in which C' is a constant. Accordingly equilibrium will exist when 4C in which K is substituted for -. 78 INORGANIC CHEMISTRY. [ 51- This equation may be written in a slightly different way, since in gases the number of molecules per unit volume is proportional to the pressure ( 31). Assuming that at a given moment the pressure of the hydrogen iodide still present is p, that of the hydro- gen is p lf and that of the iodine vapor p 2 , the equilibrium constant, K, can be represented by the expression It has been ascertained that the dissociation is less if the hy- drogen iodide was originally mixed with hydrogen or with iodine vapor. The necessity of this being true follows immediately from the above equation, for the addition of these gases amounts to an increase of pi or p 2 . UK is to remain constant, p, or in other words the mass of undissociated hydrogen iodide, must increase. We see also from the equation that the same increase of H2 or of 12 must have the same influence on the equilibrium. A further conclusion from this equation is that, when p, pi and p 2 are in- creased n-fold, i.e. when hydrogen iodide undergoing dissociation is compressed or expanded at a constant temperature, the degree of dissociation must remain unaltered, since (np) 2 p 2 This, too, is confirmed by experiment. In the dissociation of hydrogen iodide the gas volume does not change, since two molecules (2HI) yield two molecules (H 2 and I 2 ). In all such cases the degree of dissociation must be independent of the volume, because an increase or decrease in the latter causes changes in the concentration of the reacting gases which are pro- portional to each other, and hence the factor representing the concentration falls out of the equation. Hydrogen iodide is also decomposed by light. It is a peculiar fact that this dissociation is unimolecular (HI = H + 1) , while that caused by rise of temperature is bimolecular. This may be demonstrated by the following very general method. When the reaction is unimolecular, the equation for the velocity of decomposition is ^=K(ax). When, 52.] FLUORINE. 79 however, it is bimolecular (2HI = H 2 + I 2 ), the equation becomes jT=K(a z) 2 . With the help of integra be solved for K' } from the first we find T=K(a z) 2 . With the help of integral calculus these equations can where log e is the natural logarithm, and from the second : a(a-x) If now we determine x for various values of t, the values of K can be calculated; they must be constant. If this constancy appears in (1), the reaction is unimolecular, if in (2), bimolecular. FLUORINE. 52. This element was first isolated from its compounds by MOISSAN in 1886. It occurs in nature chiefly in combination with calcium as fluor spar, CaF2, and in certain rare minerals. The great difficulty in obtaining it in the free state is due to its very great affinity, which makes it unite with other elements even at ordinary temperatures. As yet it has only been success- fully prepared by the electrolysis of pure anhydrous hydrofluoric acid in which potassium fluoride has been dissolved to make the liquid a conductor. The manner in which MOISSAN accomplished this is interesting. A mixture of about 200 g. anhydrous hydrofluoric acid and 60 g. hydrogen potassium fluoride is introduced into a copper U-tube (Fig. 22) of a capacity of about 300 c.c., which has two lateral exit-tubes. The open ends of the U-tube are closed with stoppers FF, made of fluor spar and wrapped in very thin sheet platinum. The cylindrical electrodes it of platinum-iridium pass through the stoppers and are held in place by the copper screws EE, which fit tightly to the ends of the U-tube, with the help of a band of lead P. During the electrolysis the apparatus is kept at the constant tem- perature of 23 (by boiling methyl chloride). The free fluorine, which is given off as a gas at the positive electrode, is first passed through a platinum vessel that is cooled by a mixture of solid carbon dioxide and alcohol, in order to condense the acid fumes which were carried 80 INORGANIC CHEMISTRY. [ 52- over with it. The last traces of the acid are removed by conducting the gas through two platinum tubes containing sodium fluoride, which absorbs the hydrofluoric acid. The free fluorine gas was collected by MOISSAN in a platinum tube, whose two ends were closed with plates of fluor spar so that one could look through. FIG. 22. PREPARATION OF FLUORINE BY ELECTROLYSIS. (AFTER MOISSAN.) Later MOISSAN found that perfectly pure fluorine attacks glass but very slowly, so that the gas may be collected in glass vessels. Physical Properties. Fluorine is a gas with a very pungent odor and a greenish-yellow color, which is somewhat paler than that of chlorine. As a liquid it boils at 187 and is bright yellow. It can be condensed in a glass vessel. When cooled by liquid ^hydrogen it freezes to a white mass, that melts at 223. The specific gravity of the gas is 19 (O=16), that of the liquid 1.14 (water =1). Chemical Properties. Of all the elements now known fluorine has the strongest tendency to form compounds. It combines with hydrogen in the dark at ordinary temperatures in an explo- sive manner. MOISSAN demonstrated this with the help of the above apparatus by reversing the electric current while fluorine 53.] FLUORINE. 81 was being generated; thus a mixture of hydrogen and fluorine was formed, which at once exploded. As low as 252.5, solid fluorine unites with liquid hydrogen immediately, producing a flame. Finely divided carbon ignites instantaneously in fluorine gas, form- ing CF 4 . With sulphur, red phosphorus, lime and other sub- stances fluorine reacts vigorously even at 187. Fluorine com- bines with most metals instantly and violently; it does not unite with oxygen, even when it is heated with the latter to 500 or mixed in the liquid state with liquid oxygen at 190. The alkali metals (potassium and sodium) and the alkaline-earth metals (calcium, strontium, barium) take fire in fluorine gas at ordinary temperatures with the formation of fluorides. Finely divided iron glows faintly in it. Copper be- comes covered with a layer of copper fluoride, CuF2, which pro- tects it against farther corrosion ; hence the possibility of employ- ing this metal for fluorine generators. Gold and platinum are not attacked by fluorine, a rather striking fact, since these metals are acted on by chlorine, which otherwise displays a weaker chemical affinity. Fluorine reacts readily with hydrogen compounds; e.g. water is decomposed by it at ordinary temperatures into hydrofluoric acid and strongly ozonized (as high as 14% by volume) oxygen. It sets chlorine free from potassium chloride, forming potassium fluoride. The molecule of gaseous fluorine is expressed by the formula F2. Its vapor density being 19, the molecular weight is 38. Inasmuch as no fluorine compound contains less than 19 g. fluorine per gram-molecule, but frequently a multiple of this amount, the atomic weight of fluorine becomes 19 and its molecular formula Y& HYDROGEN FLUORIDE, or HYDROFLUORIC ACID, HF. 53. This compound was discovered by SCHEELE in 1771 upon heating together fluor spar and sulphuric acid : CaF 2 + H 2 SO 4 = CaS0 4 + 2HF. This is still the usual method of preparing the substance. A mixture of powdered fluor spar and dilute sulphuric acid is distilled in an apparatus of platinum or lead, since glass is instantly attacked by hydrofluoric acid. The distillate is an aqueous solu- 82 INORGANIC CHEMISTRY. [53- tion of the acid, which for the above reason must be preserved in bottles of lead or caoutchouc. By direct synthesis from its elements ( 52) hydrofluoric acid may also be obtained. Another method is by the action of hydro- gen on a fluorine compound; e.g. silver fluoride, when heated in a current of hydrogen, gives hydrogen fluoride. Still other methods are "by the action of fluorine on hydrogen compounds ( 52) and by the direct decomposition of certain com- pounds, such as hydrogen potassium, fluoride, KF-HF, which splits up on heating into the two fluorides. This last reaction is made use of when anhydrous acid is sought. Physical Properties. Anhydrous hydrofluoric acid is a color- less liquid at ordinary temperatures. It boils at 19.5 and solidifies at - 102.5. Sp. g. (H= 1) = 0.9879 at 15. It has an extremely pungent odor and. is very poisonous when inhaled. It is very soluble in water. Chemical Properties. The aqueous solution of hydrogen fluoride, the " hydrofluoric acid" of commerce, possesses entirely the character of an acid; it evolves hydrogen with most metals, the precious metals, however, and also lead, being unaffected by it. The fluorides of the metals are, in general, soluble in water; some, however, such as those of copper and lead, dissolve with difficulty, while those of the alkaline earths (Ca, Sr and Ba) are insoluble. It is a peculiar characteristic of the alkali fluorides that they are able to combine with a molecule of the acid, forming double fluorides like that described above, KF-HF. This characteristic is probably due to the fact that in aqueous solu- tion the molecule of hydrofluoric acid is IT 2 F 2 . The formation of such double molecules is often observed for acids (especially organic acids) . It is called association. Thus liquid water consists in all probability of H 4 O 2 molecules. The most important property of the gas for practical purposes is that it attacks glass (cf. 193). As a result it finds extensive use in etching glass. Glass may be etched in two ways with a solution of the gas or with the gas itself. In the first case the etching is shiny and transparent; in the second dull. The glass object is covered with a coat of wax in which the figures or letters which one desires etched on the glass may be drawn with a stylus. Then the object is either dipped in dilute 54.] HYDROFLUORIC ACID. 33 hydrofluoric acia for a while or set over a leaden dish which contains a mixture of sulphuric acid and calcium fluoride kept slightly warm by a low flame. Only the places where the coating was removed are attacked, so that, when the latter is subsequently dissolved off (by turpentine or alcohol), the etch-figure is visible. MOISSAN has proved that glass is also attacked by perfectly drv hydrofluoric acid gas. The formula of hydrofluoric acid gas is HF, which can be deter- mined in exactly the same way as was done for the analogous chlorine and bromine compounds. Compounds of the Halogens with each other. 54. The halogens, or salt-formers, i.e. the elements fluorine, chlorine, bromine and iodine (so-called because they form salts with metals by direct combination), can unite with each other to form rather unstable compounds. In general the most stable of these compounds are those whose component halogens show the greatest dissimilarity. Iodine unites with fluorine to form a compound IF 5 , which can exist even in the gaseous state. There is also a BrF 3 ; chlorine and fluorine, however, do not combine with each other. Chlorine and bromine at low temperatures give an unbroken series of mixed crystals ( 212, 2), but form no compound. With iodine chlorine gives two compounds, IC1 and IC1 3 . It depends on the quantity of chlorine present as to whether the former or the latter is obtained. IC1 is a reddish brown oil that eventually yields crystals melting at 24.7; it boils at 101.3. Water decomposes it into iodic acid, iodine and hydrogen chloride. It exists in two modifications. IC1 3 crystallizes in long yellow needles and on fusing dis- sociates almost completely into IC1 and C1 2 . In a small quantity of water it dissolves almost unchanged ; but a larger quantity of water decomposes it partially into hydrogen chloride and iodic acid. Bromine and iodine give only one compound, BrI, which is considerably dissociated in the liquid as well as in the gaseous state. Oxygen Compounds of the Halogens. With the exception of fluorine, the halogens are known to form various oxygen compounds, having the common property, of instability, i.e. of being easily decomposed. Most of them can combine with water, forming acids. Oxides which show this latter property are called acid anhydrides. The acids which are thus formed from the halogen oxides contain each but one hydro- gen atom, and this can be replaced by a metal. Acids contain- ing one hydrogen atom which can be thus substituted are called monobasic. 84 INORGANIC CHEMISTRY. [ 55- HYPOCHLOROUS OXIDE. CHLORINE MONOXIDE, C1 2 0. 55. This compound can be prepared by passing chlorine bver dry mercuric oxide at a low temperature: 2HgO + 2C1 2 = C1 2 O + HgO HgCl 2 . Hypochlorous oxide is a brownish-yellow gas at ordinary tem- peratures. It can be condensed by strong cooling to a dark-brown liquid, which boils at + 5. It is an extremely dangerous sub- stance, especially in the liquid state, since the slightest mechanical disturbances make it explode vigorously, breaking up into its elements. It is possible to distil it without decomposition, only when everything with which it comes in contact is entirely free from dust (organic matter). It acts upon sulphur, phosphorus and compounds of carbon with explosive violence. The composition of this compound was determined by BALARD in the following way: He introduced 50 vols. of the gas into a tube over mercury and decomposed it by gently warming. He thus obtained a mixture of chlorine and oxygen which occupied somewhat less than 75 vols. After the chlorine was removed by caustic potash, 25 vols. remained, i.e. 50 vols. chlorine were present, the slight difference which was observed being ascribable to the fact that a little chlorine had united with the mercury in the tube, 1 voL hypochlorous oxide yielded therefore 1 vol. chlorine and J voL oxygen. This indicates the formula 2C1 2 O=2C1 2 2 vols. 2 vols. 1 vol. The vapor density of the compound was found to be 3.03 (air = 1), or 43.63 (O = 16). Its molecular weight therefore becomes 87.26, corresponding to the formula C1 2 O (201 = 71; O = 16; sum = 87). HYPOCHLOROUS ACID, HC10. 56. When chlorine monoxide, C1 2 0, is passed into water, it is absorbed; the solution contains hypochlorous acid: =2HC1O. 56.] HYPOCHLOROUS ACID. 85 This compound is known only in aqueous solution. Its com- position is studied in its salts! The same aqueous solution can also be obtained by adding finely powdered mercuric oxide to chlorine water. HgO + 2C1 2 + H 2 = HgCl 2 + 2C10H. Soluble. Upon distillation a pure aqueous solution of the acid is obtained. Still another method of preparing the acid solution is to lead chlorine into the solution of a base, e.g. potassium hydroxide, at the ordinary temperature, whereupon a salt of hypochlorous acid (hypochlorite) is formed: 2KOH +C1 2 = KC1 +KC10 +H 2 O. By carefully treating the hypochlorite with the equivalent amount of nitric acid the hypochlorous acid is set free and can be separated from the salts by distillation. When concentrated, the aqueous solution of hypochlorous acid has a golden color. It is unstable; only dilute solutions can be distilled without decomposition. It oxidizes vigorously, breaking up into oxygen and hydrochloric acid: 2C1OH = 2HC1 + O 2 . On the addition of hydrochloric acid all the chlorine of both compounds is set free: HC1O + HC1=C1 2 + H 2 O. The hypochlorites act just like the free acid, since the presence of very weak acids, e.g. the carbonic acid of the air, serves to liberate hypochlorous acid. They are therefore extensively employed as bleaching agents ( 27). A solution of potassium hypochlorite (eau de Javelle) is used for this purpose, but chloride of lime (" bleaching powder/ 7 258) deserves particular notice. The latter is obtained by treating lime with chlorine at ordinary tern- 86 INORGANIC CHEMISTRY. [ 56- peratures. The bleaching action of hypochlorous acid is twice as great as that of the chlorine which it contains would be, if the latter were to act in the free state : 2C1+H 2 O = 2HC1+O and 2C10H = 2HC1 + 20. However, it should be remembered that two atoms of chlorine were necessary to form the one HC1O molecule: 2KOH + C1 2 = KC1 + KC1O + H 2 O. On shaking an aqueous solution of hypochlorous acid with mercury a brownish-yellow precipitate of mercuric oxychloride, nHgO HgCl 2 , is formed, which is insoluble in hydrochloric acid. Chlorine water, on the other hand, when shaken with mercury, gives white mercuric chloride, HgCl 2 (sublimate). These reactions enable us to distinguish between the two substances. In a dilute aqueous solution of chlorine we have the following equilib- rium: as is shown by the facts that the solution reacts distinctly acid toward litmus and that the hypochlorous acid can be separated from the hydro- chloric acid by distillation. The difference in the action of chlorine water and a solution of hypo- chlorous acid on mercury is due to the fact that in the above equilibrium the system C1 2 + H 2 is by far the predominant one. CHLORINE DIOXIDE, C10 2 . 57. This gas is formed when potassium chlorate, KClOa, is treated with concentrated sulphuric acid. Chloric acid is at first set free and this decomposes as follows: 3HClp 3 = HC10 4 + 2C10 2 + H 2 0. Chloric Perchloric acid. acid. Chlorine dioxide is a dark-yellow gas. It can be condensed to a liquid, which boils at 9.9 and solidifies at 79 to a yellow crystalline mass. It has a peculiar otlor resembling chlorine and burned sugar. Chlorine dioxide is extremely explosive; warming, jarring or contact with organic substances causes it to explode with vio- lence. Light slowly decomposes it. 57.] CHLORINE DIOXIDE. 37 The following experiments give one an idea of the vigor with which it causes oxidation. (1) When finely powdered sugar is mixed carefully with potassium chlorate and a drop of concentrated sulphuric acid is added, the whole mass bursts into flame. The chlorine dioxide set free makes the sugar burn at ordinary temperature. (2) Place a few pieces of yellow phosphorus and some crystals of potassium chlorate under water and allow a few drops of concentrated sulphuric acid to flow down on the two substances. The phosphorus at once burns under water with a brilliant light. Chlorine dioxide is soluble in water. Such a solution can be easily prepared by floating a little porcelain cup in a large crystal- lizing-dish with a flat brim and containing 220 c.c. water, putting into the cup 12 g. potassium chlorate and adding a cooled mixture of 44 c.c. concentrated sulphuric acid and 11 c.c. water. The crystallizing-dish is then covered with a glass plate. The chlorine dioxide evolved dissolves in the water, forming a yellow solution. When a base is added to a chlorine dioxide solution, a chlorite ( 58) and a chlorate are formed: 2KOH + 2C1O 2 = KC1O 2 + KC10 3 + H 2 0. Pot. chlorite. Pot. chlorate. This reaction proceeds very slowly m dilute solution. The composition of chlorine dioxide was determined by GAY- LUSSAC as follows: He allowed the gas to flow through a capillary tube with three bulbs. By heating the part of the tube in front of the bulbs he decomposed the gas, the action being non-explo- sive in so narrow a space. Thus there was obtained in the bulbs a mixture of oxygen and chlorine in the same proportions as they are contained in the compound. The chlorine was absorbed by potash and the residual gas (oxygen) was passed over into a measuring-tube. The capacity of the bulbs being known, it was possible from these data to calculate the volume ratio of oxygen and chlorine. It was found that 2.00 vols. of the oxide yield .987 vol. chlorine and 2.063 vols. oxygen. The combining ratio is very close to that of 1 : 2, represented by the formula C1O2 : 2C1O 2 = C1 2 + 20 2 . 2 vols. 1 vol. 2 vols. This formula is also confirmed by the vapor density, which was found to be 34.5 at 10.5, while the formula C10 2 demands 35.5+2X16- _ 88 INORGANIC CHEMISTRY. [58- CHLOROUS ACID, HC10,. 58. This acid is unknown in the pure state. Its sodium salt is formed by the action of sodium peroxide solution on a chlorine dioxide solution: 2C10 2 + Na 2 2 = 2NaC10 2 + O 2 . The silver salt, AgC10 2 , is a yellow crystalline powder, as is also the lead salt, Pb(C10 2 ) 2 ; they are both difficultly soluble in water, and break up even on warming to 100 in an explosive manner. The anhydride of chlorous acid, corresponding to the formula C1 2 O 3 , is not known. CHLORIC ACID, HC10 3 . 59. The chlorates of potassium or barium are the usual start- ing-points for the preparation of chloric acid. When dilute sulphuric acid is added to the solution of the barium chlorate, barium sulphate is precipitated and a dilute solution of chloric acid is obtained, which may be filtered off from the sulphate and dried in a vacuum desiccator over concentrated sulphuric acid. In this way a 40% solution of the acid may be obtained. On concentrating it any farther, decomposition takes place, oxygen being evolved and perchloric acid formed. The concentrated acid is a powerful oxidizing agent; wood or paper ignites when brought in contact with it. It oxidizes hydrochloric acid, chlo- rine being given off; further sulphuretted hydrogen, sulphurous acid and others, even in dilute solution. The following reaction is very characteristic of chloric acid. When indigo solution is added to a dilute solution of the acid, the former is not decolorized ; however, on the addition of a little sulphurous acid the color dis- appears, since the chloric acid is thereby reduced to lower oxides. The salts are all soluble in water, that of potassium being somewhat difficultly so, however. The composition of chloric acid was ascertained by STAS from an analysis of silver chlorate. An accurately weighed amount of the latter was reduced by a solution of sulphurous acid to silver chloride and this was filtered off and weighed. Since he knew from previous investigations the exact composition of silver chloride, the analysis of the silver chlorate was complete. STAS found thus that silver chlorate consists of 60.] PERCHLORIC ACID. 89 Silver 56.3948% Chlorine 18.5257% Oxygen 25.0795% Total ........................ 100.0000% The atomic weight of silver is 107.88; that of chlorine 35.46; that of oxygen 16.00. We then find that the ratio of the atoms in this salt is 56.3948 18.5257 25.0795 107\88~~ ~^46~- ' ~l6M~~ i.e. very close to 1:1:3, from which it follows that the empirical formula of the salt is AgClOa, that of the acid itself PERCHLORIC ACID, HC10 4 . 60. This compound is obtained by distilling potassium per- chlorate with an excess of sulphuric acid of 96-97.5% in vacuo: KC10 4 + H 2 S0 4 = KHS0 4 + HC10 4 . Pure perchloric acid boils at 39 under a pressure of 56 mm. Hg, and has a specific gravity of D 4 22 = 1.764 at 22. It is a colorless liquid which does not solidify on being cooled with solid carbon dioxide and alcohol (about 80). It decomposes slowly, taking on a dark color. With water it forms different hydrates; the best known of them is the monohydrate, HC1O 4 -H 2 0, which melts at 50; with more water a thick oily liquid is formed, similar to concentrated sulphuric acid. Such an acid contains 71.6% HC1O 4 ; it distils without change in composition at 203 and has a specific gravity of 1.82. The dilute solution of the perchloric acid is stable. In the concentrated state perchloric acid is a very strong oxidizing agent. When a little is dropped on wood or paper, these ignite with explosion. Very painful flesh-wounds are pro- duced by it. When dilute, it does not, however, release its oxy- gen nearly so readily as chloric acid. It can, for example, be gently warmed with hydrochloric acid without giving off chlorine, and it is not reduced by sulphurous acid. By these facts and by its yielding no chlorine dioxide with sulphuric acid it may be distinguished from chloric acid. 90 INORGANIC CHEMISTRY. [ 60- The salts of perchloric acid, perchlorates, are all solu- ble in water; that of potassium and especially that of rubidium are, however, very difficultly soluble in cold water. The composition of perchloric acid has been determined, as in the case of chloric acid, by the analysis of a salt, in this instance the potassium salt. A weighed amount of the latter is heated to drive off all the oxygen. The loss in weight indicates the amount of the latter. The analysis of the remaining potassium chloride, KC1, shows the amounts of potassium and chlorine. From these data it is found, in the same manner as with chloric acid, that the empirical formula of the salt is KC10 4 , that of the acid, therefore, HC10 4 . Chlorine heptoxide, C1 2 O 7 , is the oxide corresponding to perchloric acid: 2HC10 4 -H 2 0=C1 2 7 . It may be obtained by slowly adding perchloric acid to phosphorus pent- oxide cooled below 10. By distillation on a water bath the oxide is obtained as a colorless liquid, which boils at 82. It is more stable than the other oxides of chlorine ; it neither attacks paper nor acts on sulphur or phosphorus in the cold. OXYGEN COMPOUNDS OF BROMINE. 61. Although no compounds with oxygen alone are known, there are two oxygen acids, viz., hypobromous and bromic. Hypobromous acid, HBrO, can be obtained in the same way as HC1O, namely, by shaking up bromine water and mercuric oxide together. The dilute solution can be distilled in vacuo, and has properties entirely anal- ogous to those of hypochlorous acid. Bromic acid, HBrO 3 , can be obtained from the barium salt with sul- phuric acid or from the silver salt with bromine-water : 5AgBr0 3 + 3Br 2 + 3H 2 = 5AgBr + 6HBr0 3 . Insol. It is also formed when chlorine is passed into bromine-water: Br 2 + 5C1 2 + 6H 2 O - 2HBr0 3 + 10HC1. It corresponds in its behavior with chloric acid. Many reducing-agents, such as hydrogen sulphide and sulphurous acid, are able to extract all its oxygen. Most of its salts are difficultly soluble in water. When heated, they give up all their oxygen. 62.] OXYGEN COMPOUNDS OF IODINE. 91 OXYGEN COMPOUNDS OF IODINE. 62. When iodine is introduced into a cold dilute solution of caustic potash or soda, a colorless liquid is obtained, which has other properties when fresh than it has later. When freshly prepared it decolorizes indigo solution and iodine is liberated on the addition of very weak acids. *Later on these two properties disappear. It is therefore to be supposed that a hypo-iodite KIO is first formed, and that this is changed slowly to KI and KI0 3 . At the boiling-point the change takes place almost instantly. Iodine pentoxide, I 2 O 5 , is the anhydride of iodic acid, since it can be obtained by heating this acid to 170, 2HI0 3 =H 2 + I 2 5 , and yields the same acid when dissolved in water. It is a white crystalline substance, which breaks up into its elements at 300. Iodic acid, HIO 3 , is prepared by the oxidation of iodine with nitric acid, or, better, with nitrogen pentoxide. 3I 2 + 10HNO 3 = 6HI0 3 + 10NO + 2H 2 0. Nitric acid. Nitric oxide. Iodic acid is crystalline and easily soluble in water. It is a power- ful oxidizing-agent, setting free chlorine from hydrochloric acid, for example. 2HI0 3 + 10HC1 = I 2 + 5C1 2 + 6H 2 0. It reacts instantaneously with hydriodic acid, all the iodine of both compounds being precipitated: 5HI+HIO 3 =3H 2 O+6L The salts of this acid, the i o d a t e s , are in general not very soluble in water; however, those of the alkali metals dissolve rather easily. On heating iodic acid with concentrated sulphuric acid oxygen is evolved and the compound I 2 4 , iodine dioxide, is formed. This is a lemon-yellow, crystalline powder that breaks up into its elements above 130. With hot water it reacts quickly to form iodine and iodic acid: 5I 2 4 +4H 2 0=8HI0 3 +I 2 . Periodic acid, HIC>4 + 2H 2 O, is formed by the action of iodine on perchloric acid : HC1O 4 + 1 + 2H 2 = HI0 4 2H 2 O + Cl. It is a colorless crystalline solid that is entirely decomposed at 140 into iodine pentoxide, oxygen and water ( 145). 92 INORGANIC CHEMISTRY. [ 63. NOMENCLATURE. 63. The system of naming the various halogen oxygen-acids is a general one, which is also used for the acids of other elements. The best-known acid usually has the suffix -ic, e.g. chloric acid,* phosphoric acid, sulphuric acid, etc. Acids that contain more oxygen have in addition the prefix per-, thus perchloric acid and persulphuric acid. Acids containing 'less oxygen have the suffix -ous, e.g. chlorous llcid, sulphurous acid, phosphorous acid, etc. Those which contain still less oxygen have the suffix -ous and also the prefix hypo-, e.g. hypochlorous acid, hyposul- phurous acid and hypophosphorous acid. The names in use in pharmaceutical chemistry (see the National Phar- macopoeia) follow the Latin. Thus we have Acidum sulphuricum (sul- phuric acid) and Acidum sulphurosum (sulphurous acid). The names of the salts of the best-known (-ic) acids end in -ate, e.g. potassium chlorate, -sulphate, -phosphate. The salts of the -ous acids have the ending -ite, as potassium chlorite, -sulphite, -phosphite. The salts of hypo- -ous acids are called hypo- -ites; thus sodium hypochlorite, -hyposulphite, -hypophosphite. The names of the anhydrides correspond to those of their acids. < In naming oxides the name of the element with or without the ending -ic is used, unless there is more than one oxide. Where there are two oxides, the name of the one with the more oxygen ends in -ic, that of the other in -ous, e.g. mercuric oxide, arsenic oxide, mercurous oxide, arsenious oxide. An oxide with less oxygen than the -ous compound is given the prefix hypo-, and one with more than the -ic oxide the prefix per-, as in the case of acids, thus hypochlorous oxide, lead peroxide. In some cases, for the sake of euphony, the suffix is added to the Latin instead of the English stem, as cuprous, ferric, etc. For historical reasons many names now in use do not conform to this system. In some instances the oxide first discovered took the suffix -ic, and those subsequently discovered were named accordingly, as in the case of the nitrogen oxides (119). It is not uncommon to speak of oxides of the general formula M 2 O 3 as sesquioxides. 64.J SUMMARY OF THE HALOGEN GROUP. 93 A much more rational system is to indicate the number of atoms of oxygen by the Latin or Greek numeral, e.g. chlorine protoxide, or monoxide, iodine pentoxide, etc. SUMMARY OF THE HALOGEN GROUP. 64. It is evident from the foregoing descriptions that the properties of the halogens and their compounds possess great similarity among themselves. A closer study reveals the fact that the increase of atomic weight is accompanied by a gradual change of physical and chemical properties. For example, let us notice the physical properties: F. Cl. Br. I. 19 35 46 79 92 126 92 Melting-point 223 -102 7 + 113 Boiling-point. 187 - 33 + 63 + 200 Sp. g. (liquid or solid). . . Color 1 . 14(liquid) f palegreen- 1.33 greenish- 3.18 brown. 4.97 violet- ( ish-yellow. yellow black It is seen that the values of the physical constants increase on the whole with the atomic weight. The purely metalloid char- acter of the first three is also found in iodine, although in the case of the latter an external characteristic of metals, viz., metallic lustre, is at once noticeable. The affinity for hydrogen decreases as the atomic weight increases. We saw that fluorine combines with this element even in the dark and at very low temperatures in an explosive manner; iodine unites with hydrogen directly only at a high temperature and the compound is easily decomposed by heat. Inversely, the oxygen compounds are the more stable the higher the atomic weight of the halogen. While a halogen of low atomic weight displaces a halogen of higher atomic weight from its hydroger compound, e.g. HI + C1=HC1+I, the halogen with higher atomic weight can on the other hand replace one with lower in its oxygen compounds, setting that other one free: 94 INORGANIC CHEMISTRY. [ 65- ELECTROLYTIC DISSOCIATION. 65. In 30 it was stated that the properties of an aqueous solution of hydrogen chloride differ widely from those of the dry gas. It was also stated there that many other substances undergo a similar change of properties when they are dissolved in water. We may now consider the nature of this change. If we investigate the freezing-point depression of the aqueous solution of an acid, base or salt of known concentration, we find that the depression does not correspond with that calculated from the accepted molecular weight ( 43). The freezing-point depres- sion and the boiling-point elevation are both greater than they should be. A 1% sodium chloride solution would, for example, 19 be expected to show a depression of -^-^=0.325, the molecular Oo.O depression for water ( 43) being 19, i.e. AM =19, and the molec- ular weight of sodium chloride 58.5. In reality, however, the depression is found to be 1.9 times as great, namely, 0.617. As the osmotic pressure is proportional to the freezing-point depres- sion ( 42), it must also be greater than the calculated amount. The fact at once occurs to us that gases, to which dissolved substances have been found to show close analogy, also exhibit a similar phenomenon. In numerous instances the pressure exerted by a definite weight of gas occupying a definite volume at a definite temperature is greater than the calculation indi- cates. This is but another phase of the observation that the vapor density of some gases is abnormally low at certain tem- peratures ( 47). This is explained by assuming a breaking up of the gas molecules; the number of particles free to move about is thus increased and accordingly the pressure becomes greater. This phenomenon is known as dissociation ( 49). In the case of abnormal osmotic pressure we are led to a similar explanation by assuming that the molecules are split up into several independent particles. A difficulty arises, however, when we try to conceive the nature of this division. In solu- tions of salts in water it would be possible to assume a hydrolytic separation, i.e. into free base and free acid (p. 104), which would necessarily be complete in dilute solutions of the salts of strong acids and bases, inasmuch as the osmotic pressure of such solutions in concentrations of T V normal ( 93), for instance. 65.] ELECTROLYTIC DISSOCIATION. 95 amounts to double the calculated pressure. There are, however, serious objections to such a hypothesis. In the first place, it has never yet been possible to separate such a solution by diffusion into the free base and free acid which it is supposed to contain. A second and still more serious objection is that an acid or base in an aqueous solution by itself exerts an osmotic pressure greater than that calculated. Here, however, hydrolytic dissociation is impossible. The question as to the real nature of the division has found its answer in a consideration of the relation which exists between the abnormal osmotic pressure and the transmission of the electric current. ARRHENIUS observed that only those substances which conduct the electric current in aqueous solution, namely, acids, bases and salts, show the above-mentioned abnormalities in osmotic pressure. When these substances are dissolved in another liquid than water, the resulting solution is a non-conductor, but at the same time its osmotic pressure again assumes the normal. These facts enable us to perceive the connection between the apparently disconnected phenomena of abnormal osmotic pressure and elec- trolytic conduction. In order to understand this relation it is necessary to know the usual explanation of electrolytic conduction. Let us take hydrochloric acid as an example. Perfectly dry hydrochloric acid gas is a non-conductor, as is also perfectly pure water. However, when the gas is dissolved in water, a solution is obtained which transmits electricity very well. Evidently a certain reaction must have resulted from the mixing of the water and the hydrogen chloride. We were led to surmise this above ( 29), when it was found that this gas solution does not obey HENRY'S law. Since during the transmission of the current the hydrogen chloride is broken up into hydrogen and chlorine while the water remains unchanged, it must be assumed that the hydrogen chloride molecules are the ones which have undergone a change. The phenomena of electrolytic conduction now find their com- plete explanation in the assumption that the* change which the hydrochloric acid underwent consisted in a separation of its mole- cules into electrically charged atoms (ions) ( 267). This separa- tion may have been complete or partial, the extent depending upon the concentration among other things. When a current passes through the solution, the negatively charged chlorine ions (the 96 INORGANIC CHEMISTRY. 65- anions) are drawn toward the positive electrode (anode); they become electrically neutral on contact with the latter and escape from the liquid. Similarly the positively charged hydrogen ions (cations) wander toward the negative electrode (cathode). In this way conduction goes on, the undivided molecules having no part in it. This division of the molecules is known as electro- lytic dissociation, or ionization. The existence of free ions in the solution of an electrolyte is demonstrated by OSTWALD in the following manner. The tube abed, Fig. 23, is nearly filled with dilute sulphuric acid. The & " m FIG. 23. narrowed portion be is about 40 cm. long. A rod of amalgamated zinc is lowered into a to serve as the positive electrode, while a platinum wire is fused into d at p for a negative electrode. If connection is made with a battery of ten accumulators, there is an immediate evolution of hydrogen at p. The passage of the current through the liquid results in the formation of zinc sul- phate around the bar in a: Zn+H 2 SO 4 = ZnS0 4 +H 2 . Now if this hydrogen has to pass through be to p, it must cover the 40 cm. in a very brief space of time. However, it has been shown both by investigations which cannot be described here and by calculus that this migration would take many hours. The hydrogen appearing at p as soon as the circuit is closed cannot, therefore, come from a; the most natural explanation is to sup- pose that there are already free ions in the neighborhood of p and that they are discharged by the current and given off from the liquid as free hydrogen. TOLMAN has shown that, when a long tube containing a solu- tion of an alkali iodide is rotated as the spoke of a wheel at 3000- 65.] ELECTROLYTIC DISSOCIATION. 97 5000 revolutions per minute, the outer end becomes negative with respect to the inner end. The solution must therefore contain positive and negative components which can move independently of each other. The iodine ions, being much heavier than the alkali ions, would naturally accumulate at the outer end. In order that this hypothesis of dissociation into ions may also account for abnormal osmotic pressure, it must be assumed that the ions are independent particles, just as free to move as the molecules are supposed to be. The number of freely moving particles in the same volume is thus increased. Hence, whether the amount of ionization is calculated from the electrical con- ductivity or from the osmotic pressure, the result should be the same according to the above hypothesis. This is found to be the case. Supposing that every molecule yields n ions by the dissociation and that the dissociated portion of the whole number of molecules is r, the number of freely moving particles is The osmotic pressure must therefore be [l + ?-(n 1)] times as great as in the case of undivided molecules. If this pressure p is p in the latter case, then wherefore rt r>- (1) From the electrical conductivity we are able to find the value of r i n the following way: As the dilution becomes greater, the molecular con- ductivity increases. By this term we mean the specific conductivity of the solution multiplied by the number of liters in which a gram-molecule of the substance is dissolved As the dilution is gradually increased, the molecular conductivity approaches a definite limit. Since the conductivity is only due to the d i s s o c i a t e d molecules, it may be assumed that, when this limit is reached, all the molecules are broken up into ions. If the molecular conductivity for infinite dilution is represented by A^ and that for the dilution v (1 gram-molecule in v liters) by X V} it is evident that The following table shows the agreement of the values calculated 98 INORGANIC CHEMISTRY. [65- by the two methods. The values opposite r were calculated from the observed freezing-point depressions and those opposite re from the con- ductivities of the salt solutions. The concentration throughout is 1 g. per liter. KC1. NH 4 C1. KI. NaNOs. f 82 88 90 82 T . . 86 84 0.92 0.82 66. Ionic Equilibrium. In a case of electrolytic dissociation we have an equilibrium to deal with, namely, that between the un- dissociated molecules on the one hand and the ions on the other. In the case of a monobasic acid this equilibrium may be repre- sented by where A' (cation). is the acid radical (anion) and H* the hydrogen ion For a base we have MOH<=M'+OH'. We may apply here the equilibrium equation deduced in 49. Given a gram-molecules of AH per unit-volume, of which x are divided into two ions each, then the equilibrium is represented by a-x=Kx 2 . From this equation it necessarily follows that the dissociation is diminished by the introduction into the solution of a substance with like ions (just as the addition of hydrogen or iodine reduces the dissociation of hydrogen iodide gas, 50). This effect (which is called the " common ion effect ") may be produced on a salt in solution by the addition to the solution of a salt of the same base or a salt of the same acid. The equation then becomes a x = K-x(x+p), p being the concentration of the added ion. K can only remain constant provided x diminishes. It also follows that the degree of dissociation depends on the con- centration. If the latter be increased n-fold, we have from the above equation n(a-x)=Kn 2 x 2 , or (a -re) -K-n-x 2 . 66.] ELECTROLYTIC DISSOCIATION. 99 If n is > 1, x must diminish, i.e. the ionization decreases with increasing concentration. If n is <1, x must increase, i.e. the ionization increases with the dilution. When n is infinitely small, we have a=x } in other words, at infinite dilution the ionization is complete. We are now able to give another definition of acids and bases than that of 30. Acids are those substances which give H-ions in aqueous solution; bases under the same condition give OH- ions. All the properties of acids, bases and salts are closely con- nected with the degree of their ionization, among others that which is indicated by the rather vague term strength of an add or base. As early as the eighteenth century it was observed that an acid can sometimes expel another acid from its salts. On adding hydrochloric acid to sodium carbonate, for instance, sodium chloride is formed and carbonic acid given off. The same is true of bases. When a solution of caustic soda is added to a solution of iron chloride, iron hydroxide is precipitated and sodium chloride is also formed. The acid or base that can expel another from its salts was considered by BEKGMANN (1735-1784) to be "stronger" than the one expelled. Experience has taught that those acids and bases are strong- est which are the most ionized for the same dilution. Hydro- chloric acid is, for example, stronger than hydrofluoric acid. At a dilution of one gram-molecule per liter the former is almost completely (about 80%) split up into ions, the latter only 3%. It was remarked above ( 30) that acids turn blue litmus red, and bases red litmus blue. It is only natural to seek the cause of . these common properties of acids on the one hand and bases on the other in that which all acid solutions have in common, namely, hydrogen ions, and in that which all solutions of bases have in common, namely, hydroxyl ions. The reactions be- tween acids, bases and salts in aqueous solu- tion are almost invariably reactions be- tween their ions. We shall explain this later in many instances; the following example may suffice for the present. When dilute solutions of a base and an acid are mixed, we have a salt solution ( 30). In order to understand what reaction has 100 INORGANIC CHEMISTRY. [ 66. taken place we must know that in dilute solution most salts are almost wholly split up into ions. Water itself, however, is split up only in an extremely small amount. In the equilibrium H 2 O<=IT+OH', there is thus very little of the system on the right-hand side. The amount of the ionization of water has been determined in various ways, which cannot be taken up here, but are discussed in text-books of electrochemistry. The results of the different methods agree well and show that the concentration of hydrogen, or hydroxyl ions, is very nearly 1.0X10~ 7 ; i.e., 1 g. H-ions and 17 g. OH-ions are contained in ten million liters of water. Now, when a base and an acid are mixed we have together in the solution M' + OH' and A' + H'. Of these ions M' and A' can exist freely side by side; but not so with H' + OH', for these must unite to form water according to the above equilibrium. In the forma- tion of the salt we therefore have only the H' and OH' ions uniting, producing undissociated molecules of water. It is now easy to understand also why a strong acid (i.e. one almost completely ionized) expels a weak (slightly ionized) acid from its salts. To use an example, suppose we add to a liter of a sodium fluoride solution, containing one mole of the salt, a similar solution of hydrochloric- acid. In the mixed solution we have the ions H' + d'+Na' Since the equilibrium HF<=H' +F is conditioned on the presence of only 3% of H* Ions and F ions, there is a large excess of these ions in the liquid and almost all of them must unite with each other, while the Na* and Cl' ions remain free; in other words, hydrofluoric acid and sodium chloride (dissociated) are formed. It also becomes manifest that the old notion, once very gen- erally held, that the stronger acid expels the weaker one from its salts completely is incorrect. When the expelled acid or base escapes from the solution as a gas or is precipitated, the expulsion may in- deed seem to be complete; we shall examine the case more thor- oughly in 73. We can now go a step farther. It was stated above that water 66.] ELECTROLYTIC DISSOCIATION. ^' ; M01 is partially ionized, though only to a very slight extent. Suppose that we dissolve in water a salt of a strong base and an extremely weak acid, such as potassium cyanide, for instance. As such a salt is highly ionized, we have in the solution the ions H* + OH'+K' Since the acid HCN is very weak, there will be too many CN' ions in the liquid to satisfy the demands of the equilibrium equation H' + CN'<=HCN; hence some of the H' ions of the water will unite with CN' ions. At the same time, however, an excess of OH' ions is created in the liquid; for, inasmuch as potassium hydroxide is a strong base, they do not unite with the K* ions. The water, which originally reacts neutral, because hydrogen and hydroxyl ions are present in equal numbers and are mutually compensating in regard to their action on litmus, thus comes to have an alkaline reaction by the solution in it of potassium cyanide. We therefore see that salts of this nature are partially split up by water into free base (K' + OH') and free acid (undissociated HCN). This phenomenon is called hydrolysis. We shall meet with it frequently in the sequel. When ARRHENIUS presented the doctrine of electrolytic dis- sociation in 1887, it met with much opposition. It was seen that its effect would be to produce a veritable revolution of many previously accepted views. Compounds such as hydrochloric acid, sodium nitrate and others, which had ever been considered as the most stable, were to be supposed according to the theory of ARRHENIUS to break up as soon as they dissolve in water. It also seemed nonsensical that we should have to assume the exist- ence of free potassium and iodine in a solution of potassium iodide, for example, since potassium produces hydrogen and potassium hydroxide as soon as it touches water and since a KI solution is colorless, while iodine solutions are brown. So far as the first point is concerned, it should be noted that it is the solutions of these same strongly ionized compounds which are chemically the most active, a fact which indicates rather a loose than a firm union of the constituent atoms in the molecules. In regard to the example of potassium iodide solution and other cases, care must be taken not to confuse atoms and ions. The 102 A CHEMISTRY. 66- solution of potassium iodide retaining our illustration contains neither free potassium nor free iodine but ions of potassium and ions of iodine. The atoms, however, must possess an altogether different energy supply than the ions, whose electric charges are very heavy, as can be proved by different methods. It is this energy supply on which the properties of bodies depend ; and since this is apparently much different with the ions than with the atoms, it is perfectly natural that the latter should display other proper- ties than the former. SULPHUR. 67. Sulphur was known to the ancients. It occurs free in nature, principally in the vicinity of active or extinct volcanoes. Sicily is its most important locality, closely followed by Louisiana in the United States, but large quantities are also found in other parts of the United States and in Iceland, Japan, and Mexico. FIG. 24. DISTILLATION OF SULPHUR. It is separated from the accompanying rock, or matrix, by fusion. 67.] SULPHUR. 103 In Louisiana this is accomplished by the FRASCH process, whereby hot water under pressure is forced through pipes sunk through the ground to the sulphur deposit, thus melting the sulphur, which, in a molten form, is forced up to the surface by compressed air. The crude sulphur thus obtained is still impure. It is refined (Fig. 24) by distillation. After being melted irr B it is let down into the cast-iron cylinder A, which is heated to a temperature above the boiling-point of sulphur. The vapor is conducted into a large brick chamber, equipped with a safety valve for the release of air. If the distillation is conducted so slowly that the temperature of the chamber does not exceed the boiling-point of sulphur, the latter is deposited in the form of a fine powder, called "flowers of sulphur" just as water vapor, when suddenly cooled below 0, turns to snow. Rapid distillation, however, yields a layer of liquid sulphur on the floor. It may be let out through the opening C and cast into slightly conical wooden molds. This is the roll sulphur, or the roll brimstone, of commerce. Besides occurring in the free state sulphur is also found in numerous compounds, from some of which it is obtained, e.g. pyrite, or iron pyrites, FeS 2 , which yields sulphur on heating: 3FeS 2 = Fe 3 S 4 +2S. Many other compounds of the element with metals, the sul- phides, occur in nature, e.g. galenite (lead sulphide), zinc blende (sphalerite, zinc sulphide), stibnite (antimony sulphide), cinnabar (mercury sulphide), realgar and orpiment (arsenic sul- phides) and chalcopyrite (copper pyrites, copper and iron sul- phide). Sulphur also occurs in the natural sulphates, of which gypsum (CaS04 + 2H 2 O) is the most important. It is also found in the organic world as a constituent of the albuminoids. Physical Properties. Sulphur is known in various modifications. At ordinary temperatures the stable form is a yellow crystalline solid; melting-point, 119.25. A little above its melting-point sulphur is a mobile yellow liquid. With a continued rise of tem- perature it becomes much darker in color and very viscid; at 180 it can no longer be poured; at a higher temperature, espe- cially above 300, it again becomes mobile, the dark color remain- ing; at 448 it boils, producing an orange-colored vapor. At 500 the vapor is red; above this temperature it becomes clearer again. During cooling these phenomena reappear in inverse order. At -80 sulphur is colorless. 104 INORGANIC CHEMISTRY. [ 67- Sulphur is insoluble in water and difficultly soluble in alcohol and in ether; it is easily soluble in carbon disulphide and in sulphur monochloride, S 2 C1 2 . 100 parts CS 2 dissolve 46 parts S at 22. The molecular weight of this element, more than that of any other, depends on the temperature. Below the boiling-point the molecular formula is Sg, according to the determination of the boiling-point elevation in carbon disulphide (boiling-point 46) and the freezing-point depression of fused naphthalene (melting- point 80). In the gaseous state the density (air= 1) varies from 7.937 at 467.9 to 2.23 at 860 and then remains constant even as high as 1800, indicating that at the lowest temperatures sulphur vapor consists of Sg molecules, and above 860 of only 82 molecules. Above 1800 the molecule 82 begins to dissociate into its atoms; at 2000 and 0.5 atmosphere pressure the dissociation has reached about 45%, according to an investigation of NERNST. 68. Allotropic Modifications. At least four solid forms are known, while in the liquid state there are two more. The solid allotropic conditions can be divided into crystallized and amor- phous. As for the former, sulphur is dimorphic, forming rhombic as well as monoclinic crystals. The former are transformed into the latter on heating ( 70). Rhombic sulphur can be obtained in beautiful crystals by allowing a solution of sulphur in carbon disulphide or chloroform to slowly evaporate. Monoclinic sulphur is easily obtained in the following manner: Some sulphur is fused in a large crucible and allowed to cool slowly until a crust forms on the surface. The crust is then broken through and the liquid sulphur poured out; the sides of the crucible are found to be covered with long, yellow transparent needles. In the course of a few hours these become opaque and brittle, however, and crumble at the slightest touch to a powder, which is found to consist of rhombic crystals (cf. 71). Amorphous sulphur may be either soft and soluble in carbon disulphide or powdery and insouble in this liquid. The soluble kind results from the decomposition of certain sulphur compounds. When hydrogen sulphide water is exposed to the air, sul- phur slowly separates in the form of a white powder. The polysulphides (CaSn, K 2 Sn, etc.) yield, when decomposed by an acid, a cloudy milk-like liquid, which is found to contain extremely fine particles of amorphous 68.] SULPHUR. 105 sulphur. In either case there is always formed in addition to the soluble variety some insoluble (in CS 2 ) sulphur. The insoluble form may be best prepared by heating sulphur to near its boiling point and then pouring it in a fine stream into cold water; thereby a semi-solid plastic modification is formed, which becomes brittle after a time. By extraction with carbon disulphide the soluble modification is removed and a yellow powder remains, which is the amorphous modification, insoluble in that liquid. The relative quantity . of this latter modification depends only on the temperature at which the sulphur was heated. The higher the temperature, the greater the yield; on heating at 440, for example, the yield is 30.3%. It is a very curious fact that amorphous sulphur is not formed if a few bubbles of ammonia gas or of carbon dioxide are first passed into the heated sul- phur and that, on the contrary, the introduction of air will restore the ability to form amorphous sulphur. Probably traces of sulphur dioxide are necessary for the formation of the amorphous state. Further, it has been proved that molten sulphur, no matter whether it can produce the amorphous modification by rapid cooling or not, has in both cases the same physical properties, such as specific gravity, boiling-point, solu- bility, etc. In order to explain this fact, the conduct of sulphur on heating must be considered. As has already been stated, sulphur when heated above its melting-point is at first a mobile liquid. When the temperature reaches 160 the liquid very soon becomes viscid. If the temperature is maintained for some time at 158-160, the molten mass separates into two liquid layers, a mobile one and a viscid one. In other words, sul- phur can form two liquid modifications, which are only partially miscible. At every temperature an equilibrium establishes itself in the molten sulphur between these modifications. When rapidly cooled the viscid form continues to exist, for the reason that its transition velocity is strongly diminished, and gives the amorphous modification on solidi- fying. When cooled slowly the viscid modification changes gradually into the mobile liquid form and the resulting solid does not contain amorphous sulphur. That amorphous sulphur is not obtained by rapidly cooling molten sulphur which is completely free from sulphur dioxide, must probably be attributed to a very strong catalytic retardation of the transition of the viscid form into the mobile one by traces of sulphur dioxide. In the light of the above facts concerning the behavior of sulphur we can understand why its melting-point is dependent on its previous history. Sulphur that has been heated melts lower than otherwise. This is due to the presence of amorphous sulphur and the consequent lowering of the melting-point, the same as by a foreign substance. The molecular weight of the amorphous insoluble sulphur is also S 8 . 106 INORGANIC CHEMISTRY. [e9- 69. Chemical Properties. Sulphur combines directly with many elements, not only metals but also metalloids. It has been already stated ( 10) that it burns with a blue flame when heated in air or in oxygen. The halogens and hydrogen unite with it directly. Powdered iron and sulphur, when mixed and heated, combine energetically, producing great heat ( 20). Copper takes fire in the vapor of boiling sulphur. When mercury and sulphur are rubbed together in a mortar, black mercuric sulphide, HgS, is formed. The sulphur compounds of the metals are called sul- phides. THE TRANSITION POINT. 70. As stated in 68 sulphur can crystallize in two modi- fications, rhombic .and monoclinic. These modifications can be readily transformed into one another. The peculiar phenomena connected with this transition deserve a closer study. At ordi- nary temperatures sulphur is rhombic and remains so till the temperature 95.4 is reached, above which there begins a slow but complete transformation into the monoclinic variety. Inversely, when the monoclinic modification is subjected to a temperature below 95.4, a complete change into the rhombic form occurs. At the temperature named the two modifications are equally stable and can exist side by side in any proportions for an indefinite period; above it only the monoclinic, below it only the rhombic, form can exist permanently. Such phenomena are not infrequent. The temperature at which the one system passes into the other is called the transition point, also point of inversion. This transi- tion point possesses great analogy with the melting-point. Just as ice, for example, is changed into water above and water into ice below 0, so in a system of substances possessing a transition point only one system is stable below that point, above it only the other. ^The theoretical explanation of both phenomena is exactly the same. Let us consider a body, ice for example, at temperatures slightly below its melting-point, and represent graphically in the diagram OTP (Fig. 25) the values of the vapor tension corre- sponding to different temperatures. The result is the line marked ice in the figure. This vapor-tension curve, if prolonged through and beyond the melting-point, is found to bend sharply at the 70.] THE TRANSITION POINT. 107 latter and take a new direction. This deflection is very slight in the case of ice and water; it can be nevertheless experimentally detected; it is much more evident with benzene and many other substances. By carefully cooling water it can be made to remain liquid even under 0; such a liquid is said to be supercooled. The vapor tension of this supercooled water is greater than that of ice at the same temperature and the curve representing the former is but a continuation of the vapor-tension curve for water. Since the vapor tension of supercooled water is greater than that of ice, water at temperatures below must, according to previous con- clusions ( 43, 3), pass into ice when the two are in contact. How- ever, the vapor tension of water at a temperature slightly above o o T FIG. 25. FIG. 26. will be less than that of ice and we shall have the ice trans- formed into water. It is therefore evident that both above and below the melting-point one of the systems will necessarily disappear. Exactly the same explanation can be offered for the transition point. Below 95.4 the vapor tension of rhombic sulphur is less than that of monoclinic sulphur; above, the vapor tension of the rhombic variety exceeds that of the monoclinic. There is therefore a complete transformation from one system to the other when the temperature is other than 95.4, for the same reason as in the case of the melting-point; moreover, just as ice and water under ordinary pressure can exist side by side indefinitely only at 0, so both modifications of sulphur are coexistent only at 95. 4, since only then is the vapor tension the same for both systems (Fig. 26). Of the various methods for the determination of the transition 108 INORGANIC CHEMISTRY. [68^ point a convenient one is the dilatometric method. It is based on the change of volume (specific gravity) which a body usually undergoes on passing through the transition point. In measuring this a dilatometer is used, an instrument which may be compared to a thermometer of very large dimensions. After rhombic sulphur, for example, has been placed in the dilatometer the latter is filled with a chemically indifferent liquid (kerosene, linseed oil) and put in a large water bath; the temperature is then slowly raised. Below the transition point the volume is seen to slowly and steadily increase with the temperature on account of expansion; as soon as the temperature gets a trifle above 95.4, however, a marked increase of volume is observed, even if the temperature be maintained constant; thereupon expansion again proceeds gradually, as before, if the tem- perature is allowed to rise. The marked change of volume indicates the transition of the rhombic sulphur into the mono- clinic modification. "STABLE," "METASTABLE," AND "LABILE." These terms are coming to be so frequently used in chemistry that they need to be distinctively denned. They are borrowed from mechanics, for which reason it is desirable that they be employed in chemistry in the same sense as in mechanics. In the latter an equilibrium is called labile (apt to slip) when the slightest displacement suffices to transpose the body into a new position of equilibrium. An example is afforded by a cone standing on its apex. It cannot recover from even the slightest disturbance, but gets further and further from the vertical position and finally tumbles over. A labile condition is thus really a limiting case which cannot actually be realized; not even for the cone, though its apex were a mathematical point resting on an absolutely hard surface. All actually occurring equilibria are stable; but there can be different degrees of stability. When a material cone is stood on its apex its equilibrium has very little stability. On the contrary 71.] THE PHASE RULE OF GIBBS. 109 a beam resting on the ground with its largest surface down represents a very stable equilibrium. However, if the beam is stood up on end, its equilibrium becomes less stable. Like the cone resting on its apex, the beam will have a tendency to go over into a more stable condition. In mechanics there is no need of giving such conditions a special name, but in thermodynamics and chemistry they call for special designa- tion, and the term applied to them is metastable. We have an example in undercooled water, something that can exist, but has the tendency to go over into a more stable condition, namely, into ice. Therefore undercooled water is said to be metastable. It follows from the above that expressions such as, " labile compounds " (e.g., for C10 2 ), or " the substance exists in a labile condition," are to be avoided. The word " labile " should be replaced by " metastable." Strictly labile condi- tions are impossible; nevertheless they may possess great theoretical interest, such, for instance, as the case of the continuous transformation of liquid into gas below the critical temperature, which, though it cannot be carried out, has led to very valuable theoretical considerations in the hands of VAN DER WAALS and others, as may be seen in the larger text-books of physics. THE PHASE RULE OF GIBBS. 71. The phase rule treats of the equilibrium in heterogeneous systems, i.e. systems that can be separated mechanically into unlike parts. A saturated salt solution in contact with solid salt is a heterogeneous system, for it consists of solid salt, the solution and vapor; that is, of three parts, mechanically sepa- rable. Each of these parts in . itself is homogeneous, i.e. each part has the same composition throughout. A gas mixture is always homogeneous, as is also a solution. These homogeneous parts, separated by limiting surfaces and of which a heterogeneous system is made up, are called by GIBBS 110 INORGANIC CHEMISTRY. [71- phases. Water and its vapor constitute two phases; ice, water and steam three phases. A heterogeneous system can never have more than one gaseous phase, because all gases are miscible in all proportions* it may, however, consist of different liquid phases, in case it contains immiscible liquids. The number of these liquid phases is seldom more than two; that of the solid phases is un- limited. A further conception, introduced by GIBBS, is that of the components of a system. If the system is composed of only one element, then this element is the only component. Systems made up of one compound have in most cases this compound as the only component. A system consisting of molten and gaseous sulphur, or of water and steam, has but one component. In this case all phases have the same composition. There are systems, however, in which this is not the case; viz. systems that are made up of more than one component. We select as the components those compounds of which the smallest number is necessary to form the different phases. The choice of such compounds may be somewhat arbitrary but their number is always fully denned. Let us consider, for example, the system Glauber's salt- water. This salt has the composition, Na2S04- lOH^O. In order to determine the composition of the phases that are possible here (solid salt, solution, vapor) it is best to choose Na2SO 4 and H2O as components. We might indeed take Glauber's salt itself as one of the components; but then, in case the solid phase was the anhydrous salt, it would be necessary to regard water as a negative part of it, which is undesirable. Sulphuric acid and sodium hydroxide are not components, because they do not occur independently in any phase, neither are they found in any other relation in the phases than as a part of the salt itself. It is a property of the components that they can occur in some of the phases in varying proportions (e.g. in saturated and unsaturated solutions). Let us now take, for example, a saturated solution of salt and water in a vessel that is closed with a movable piston. Under this solution let there be a little solid salt, above it the vapor of 71.] THE PHASE RULE OF GIBBS. HI the solution. The system consists manifestly of two substances and three phases. So long as the temperature remains constant, the vapor of the salt solution possesses a definite tension. If we increase the volume by raising the piston, a definite amount of water will evap- orate; since the solution was saturated, the result will be that a little salt will be deposited; in the end the quantities of vapor, solution and salt will therefore have altered, but the composition of each phase will remain the same. The tension, and hence also the concentration, of the vapor remain unchanged, since the temperature is constant; there is likewise no change in the con- centration of the salt solution. The same is true in case the volume be diminished. It therefore follows that the equilibrium of such a system is independent of the quantities of the various phases. It is dependent only on the temperature chosen; if this is constant, the whole system is defined. Or, if we should select an arbitrary value for the composition, the temperature and pressure would be fully defined. It is therefore evident that the system is completely defined as soon as one of these magnitudes is arbitrarily chosen. The system has only one degree of freedom. Such an equilibrium possesses the following characteristic: At a given constant temperature the vapor pres- sure is definite. Under an even slightly greater or smaller pressure one of the phases will gradually and completely dis- appear, provided the temperature remains constant. On in- creasing the pressure the gaseous phase wholly condenses, so that only solution and salt remain. A decrease of pressure results in the complete evaporation of the solution, vapor and salt only being left. The same is true when the pressure remains constant and the temperature varies. An entirely different behavior is shown by a system made up of an unsaturated salt solution and its vapor. At a constant temperature and a definite position of the piston the vapor tension has a definite value, as in the former case. If, however, the volume of vapor be changed, the tension will correspondingly vary, for, if the volume be increased, for example, more water will evaporate, the solution will become more concentrated and the vapor tension of course lessen. Therefore for every definite 112 INORGANIC CHEMISTRY. [ 71- temperature there are not simply one but infinitely many pres- sures under which this system can be in equilibrium. The result is that the slightest change of volume or pressure does not necessitate the disappearance of one of the phases. Two magnitudes may be chosen arbitrarily before the system is fully denned; it has two degress of freedom. It is evident in this example that the number of degrees of freedom increases by one when the number of phases decreases by one. The phase rule expresses a relation between the numbers of the components S, the phases P, and the degrees of freedom F. It is of the following form: or, in words, The sum of the number of the degrees of freedom and the number of the phases of a system exceeds the number of com- ponents by two. Let us apply the phase rule in the first place to water, a system of one component; the sum of the degrees of freedom and the phases must therefore be three. In the following graphic Representation, Fig. 27, the tempera- Solid Liquid ! Gaseous FiG. 27. 95.4 120 FIG. 28. tures, t, are plotted as abscissie, the pressures, P, as ordinates. Let us first consider liquid water above 0. The number of the phases is two (liquid and vapor) ; the system has therefore only one degree of freedom, or, as we say, it is univariant. To every temperature there corresponds a definite vapor tension. The ordinates of every point in the line OB indicate these vapor ten- 71.] THE PHASE RULE OF GIBBS. 113 sions. If the pressure at a certain temperature were greater than that indicated by the ordinate, the gaseous phase would com- pletely disappear. The line OB therefore represents the boundary between the liquid and gaseous phases for the various temperatures and pressures. Every point in the area COB represents the liquid, every point in A OB the gaseous, phase. Only the points of the line OB indicate the temperatures and corresponding pressures, at which both phases are coexistent. The line OB therefore ends on one side at 0; its other end is at the critical temperature, since at this point vapor and liquid become identical. Let us now allow the temperature to fall below 0. The liquid phase dis- appears and ice takes its place. The system remains unvariant, however, for the number of phases is unchanged. The ordinates of the points on the line OA again give the vapor tensions of ice for different temperatures. For the same reason as above OA is the boundary line between the solid and gaseous phases. Only along this line are the two coexistent. The line OA extends to the absolute zero, since the gaseous phase then disappears. The melting-point of ice depends somewhat on the pressure, being lowered by increasing pressure 0.0075 per atmosphere. Both phases, ice and water, will therefore be coexistent along the line OC, which shows a very considerable rise of pressure for a very slight fall of temperature. In this case also a change of pressure at a constant temperature, or the reverse, involves the complete disappearance of one of the phases. The line OC will end at a point where the liquid and solid phases become identical, i.e. where the whole system turns to a homogeneous amorphous mass. The location of this point has not yet been ascertained. The point (about +0.01), the melting-point of ice at the pressure of the vapor, is, according to the above mode of repre- sentation, the point of intersection of the three lines which sepa- rate the phases and along which two phases are coexistent. It is called a triple point. Only in this point is it possible for the three phases to exist side by side; it is the common point of the areas which represent regions of the three phases. When three phases are coexistent the system has no degree of freedom; it has become non-variant. In the case of sulphur we have one substance and four possible phases: rhombic, monoclinic, liquid, gaseous. Fig. 27 makes 114 INORGANIC CHEMISTRY. [71- plain the relation between these phases. Below 95.4 sulphur is rhombic; the two phases are rhombic sulphur and vapor. The line OA forms the boundary between the two regions. At 95.4 the rhombic phase passes into the monoclinic phase. The ordi- nates of the line OB represent the vapor pressure of monoclinic sulphur at the temperatures 95.4-120. The two crystallizablc phases can exist side by side at the point (the transition point). According to researches by REICHER this transition point depends on the pressure; an increase of pressure of one atmosphere raises it about 0.05. The boundary between the crystallized phases is therefore furnished by a line OC, which shows that a very slight rise of temperature is followed by a very considerable increase of pressure. At we have therefore a triple point, i.e. a point com- mon to both crystallized phases and the gaseous phase. At B, the melting-point of monoclinic sulphur, there is a second triple point, which is wholly analogous to the melting-point of ice. Finally, it should also be noted that the line BC', which separates the liquid and the solid phases, must indicate a rise of melting-point for an increase of pressure, since sulphur melts higher the greater the pressure. The lines OC and EC' are not parallel but intersect, according to TAMMANN'S experiments, at 151 and 1281 atmos- pheres. As the sum of the phases and degrees of freedom is also three with sulphur, the phase rule indicates that all four phases cannot exist in the presence of each other at the same time, not even when the system has become non-variant. At the triple point neither the temperature nor the pressure can be changed without altering the kind of equilibrium. Here the system is non-variant. Along the lines OA, OB and OC it is univariant. When the state of the system is represented by a point within one of the areas it is divariant, consisting then of only one phase. In the succeeding chapters we shall have occasion to concern ourselves with systems of more than one component. 72.] HYDROGEN SULPHIDE. 115 HYDROGEN SULPHIDE, SULPHURETTED HYDROGEN, H 2 S. 72. This gas occurs in nature chiefly in volcanic regions. Cer- tain mineral waters, especially the so-called "sulphur springs/' contain it. It is also found as a putrefactive product of organic bodies. Hydrogen sulphide can be obtained from its elements by synthesis. They unite almost completely when heated together for a long time (about 168 hours) at 310. It can also be obtained by the action of hydrogen on sulphur compounds, as well as by the action of sulphur on compounds of hydrogen; the reduction of silver sulphide, Ag 2 S, with hydrogen at high temperatures illustrates the former case, while the boiling of turpentine oil with sulphur is an example of the latter. None of the above methods is adapted to the preparation of the gas in the laboratory. For this purpose the interaction of a sulphide with a hydrogen compound is employed, iron sulphide and dilute acids being generally used: FeS + 2HC1 = FeClp + H 2 S. In order to have sulphuretted hydrogen always at hand, it being in con- stant demand in analytical work (cf. 73), a very convenient apparatus was devised by KIPP, which can be used for the generation (at ordinary temperatures) of other gases as well. Its construction is shown in the figure (see next page) . The lower globe is joined to the basal portion by a narrow neck, while the upper globe tapers into a long tube, which fits tightly into the lower globe and extends nearly to the bottom of the generator without com- pletely filling the neck. The iron sulphide is put into the middle portion and the dilute acid is poured into the upper portion, the stopcock remain- ing open. As soon as the basal part is filled with the acid the cock is closed and the top part is half filled with more acid. When the cock is opened the liquid sinks in the top part and rises into the middle portion, where it reacts with the iron sulphide to produce hydrogen sulphide, which escapes through the cock. On closing the latter the gas continues to be evolved till it forces the liquid back out of the part containing the iron sulphide. The reaction thus ceases automatically and the generator is ready at any time to supply new quantities of gas on opening the cock, till either acid or sulphide is exhausted. The spent acid can be let out through a stoppered opening near the bottom. On account of the free iron usually present in iron sulphide, the gas 116 INORGANIC CHEMISTRY. [72- prepared in this manner contains some hydrogen. Perfectly pure hydrogen sulphide is obtained by warming antimony sulphide, Sb 2 S 3 , with concentrated hydrochloric acid- FIG. 29. KIPP GENERATOR. Physical Properties. Hydrogen sulphide is a colorless gas of disagreeable odor, when diluted reminding one of rotten eggs. Under a pressure of about 17 atmospheres it becomes liquid at ordinary temperatures; liquid hydrogen sulphide boils at 61.8 and freezes at 85. 1 1. H 2 S gas weighs 1.5392 g. at and 760 mm. pressure. The gas is rather soluble in water, 1 vol. water dissolving 4.37 vols. K 2 S at (" hydrogen sulphide water "). Chemical Properties. --Hydrogen sulphide is combustible and yields on combustion either sulphur dioxide and water or water and sulphur, according to the air supply: In aqueous solution it is slowly oxidized by the oxygen of the air, sulphur being set free; this decomposition is aided by light. In order to preserve hydrogen sulphide water, it must be pre- pared from boiled (air-free) water and put into a dark bottle, filled entirely and closed air-tight. The latter condition is best met by placing the bottle, stopper downwards, in a glass of water. It is poisonous ; as an antidote very dilute chlorine may be inhaled. 73.] HYDROGEN SULPHIDE. 117 Hydrogen sulphide is a powerful reducing-agent. Bromine water and iodine solution are decolorized by it with separation of sulphur ( 45 and 48). Various oxygen compounds are transformed by hydrogen sul- phide into compounds with less oxygen, e.g. chromic acid is reduced in acid solution to a chromic salt ( 295). Fuming nitric acid acts so vigorously that a slight explosion occurs. When hydrogen sulphide is passed over lead dioxide, the gas ignites, while the oxide is reduced. Concentrated sulphuric acid, H 2 SO4, is also reduced; hence it cannot be used for drying the gas. Hydrogen sulphide possesses the character of a weak acid; when it is passed over zinc, copper, tin or lead, the corresponding sulphides are formed and hydrogen is set free. Composition of Hydrogen Sulphide. When a bit of tin is heated in dry hydrogen sulphide in a tube over mercury tin sulphide and hydrogen are formed. After cooling it is seen that the volume of hydrogen is just as great as that of the hydrogen sulphide. The same result is obtained when a platinum wire is heated to redness (by an electric current) in the dry gas, causing the latter to break up into its elements. Since the hydrogen molecule is H 2 , there must be two atoms of hydrogen in the hydrogen sulphide molecule. Now the specific gravity of hydro- gen sulphide has been found to be 1.1912 for air=l, or 17.15 for /v A mi i i ii FIG. 30. DECOMPOSITION OF H,S. O = 16. The gram-molecule there- fore weighs 34.30 g., and, as it contains 2 g. hydrogen, there remains for sulphur 32.3 g. This figure is very close to the atomic weight of sulphur, hence there can only be one atom of sulphur present in the molecule of hydrogen sulphide. We thus conclude that the formula is H 2 S. 73. Use of Hydrogen Sulphide in Analysis. Hydrogen sulphide finds extensive use in qualitative analysis. A large number of metals are precipitated by it from acid solutions as sulphides, viz., gold, platinum, arsenic, antimony, tin, silver, mercury, lead, bis- muth, copper and cadmium, and also certain rare elements. Some of these sulphides have a characteristic color, e.g. the orange- red antimony sulphide, Sb 2 S3, the yellow cadmium sulphide, 118 INORGANIC CHEMISTRY. [ 73 CdS, the brown stannous sulphide, SnS, the yellow stannic sulphide, SnS 2 , and the yellow sulphides of arsenic, As 2 S 3 and As 2 Ss. The rest of the sulphides named are black. Other metals, such as nickel, cobalt, iron, manganese, zinc, chromium, aluminium, etc., are not precipitated by hydrogen sulphide from acid solution . but are precipitated by ammonium sulphide. Still other metals, such as barium, strontium, calcium, magnesium, and the alkalies, are not precipitated from their solutions even by ammonium sul- phide, so that we therefore possess in sulphuretted hydrogen and its ammonium compound a means of separating these elements. An answer to the question, why some elements are precipitated from acid solution by hydrogen sulphide and others are not, is furnished by the ionic theory. Let us take, for example, a dilute solution of copper sulphate, into which hydrogen sulphide is being passed. Copper sulphate is almost entirely ionized, hydrogen sulphide only to a very small degree (d). We therefore have in the solution: Cu"+ S0 4 "+2OT' + S" + (1-)H 2 S, the cations being represented by a point and the anions by a line above and to the right, and the number of these points or lines indicating the ionic valence ( 76). Some of the copper ions and sulphur ions will then unite to form undissociated molecules, CuS, which are only slightly soluble in water and are therefore precipitated. As S-ions thus disappear, the equilibrium between hydrogen sulphide and its ions is dis- turbed; new H 2 S molecules are then split up into ions, so that there are again S-ions present, which can unite with copper, and so on. The action proceeds according to the equation: CuS0 4 + H 2 S = CuS + H 2 SO 4 , Insol. or, if only the ions which take part in it are represented: Cu"+ S" = CuS. This takes place quantitatively if the copper solution is dilute and no considerable amount of any strong free acid was added. However, if these conditions are not fulfilled and, as a result, the 73.] HYDROGEN SULPHIDE. 119 concentration of the hydrogen ions is rather high, the presence of these ions reduces the ionization of H 2 S so much ( 66) that no precipitate can be formed. "The application of the mass- action law to the case is very simple. Copper sulphide, when in contact with water, dissolves to an extremely small extent; in this solution we have the equilibrium: Cu '+ S" * CuS. If the concentrations of the two ions are a and 6, and that of the undissociated -copper sulphide is c } we have the equation k being a constant for a fixed temperature ( 66). The product ab has a definite value for every saturated solu- tion (since c is definite). This value is known as the solubility product of the substance in question. If in any case the product ab is less than this value, none of the substance can separate out, because the solution will then be unsaturated; if, however, the product is greater than the solubility product, the substance will be precipitated. As soon then as the concentration of the S-ions becomes so small (because of the reduction of the ionization of hydrogen sul- phide by the H-ions of the acid) that it makes the value of ab smaller than that of the solubility product for copper sulphide, no precipitate will be formed. If, however, the liquid is diluted, the concentration of the H-ions decreases; then, if hydrogen sulphide is passed in, the concentration of the S-ions increases. The value of the solubility product can in this way be exceeded, in which event copper sulphide will be precipitated. If a small quantity of strong acid be added to a precipitate of copper sulphide suspended in water, only a very small amount of the sulphide will dissolve; to be sure, the H-ions of the strong acid will remove a part of the S-ions, yielding some undissociated hydrogen sulphide, so that in order to establish equilibrium a trace of copper sulphide must go into solution; but soon the point will be reached when so many Cu- and S-ions are again in the solution that the value of the solubility product is reached. After this moment no more copper sulphide goes into solution. Since the value of the solubility product is very low, the solubility of the 120 INORGANIC CHEMISTRY. [ 73- sulphide in dilute strong acids is very slight; this accounts for the practically complete precipitation of the copper sulphide. On the other hand, if the solubility product of a sulphide is greater, as in the case of iron sulphide, the addition of sulphuretted hydrogen to the solution of an iron salt, e.g. ferrous sulphate, FeSO 4 , will cause no precipitate of iron sulphide, and iron sulphide will, unlike the previous case, be dissolved by dilute strong acids. When hydrogen sulphide is led into a solution of ferrous sulphate to the point of saturation, the concentration of the S-ions is, on account of the slight ionization of hydrogen sulphide, not great enough together with that of the Fe-ions to reach the solubility product of iron sulphide, hence no precipitate forms. Moreover, when dilute hydrochloric acid is added to iron sulphide, the H-ions and the S-ions form undissociated H 2 S molecules and the concen- tration of the S-ions therefore becomes too small in comparison with the value of the solubility product to prevent solution ; hence in the presence of enough acid all the iron sulphide goes into solution. It now becomes clear, too, why iron solutions are precipitated by ammonium sulphide. This takes place according to the equa- tion FeS0 4 + (NH 4 ) 2 S = FeS + (NH 4 ) 2 S0 4 . In this case there are no H-ions in the solution to act on the iron sulphide. The reason for the non-precipitation of metals like barium, etc., either by sulphuretted hydrogen or ammonium sulphide lies in the easy solubility of their sulphides. Hydrogen Persulphide. 74. If a solution of sodium sulphide, Na-jS, is digested with sulphur, the sulphur dissolves and the liquid contains compounds called poly- sulphides and having formulae from Na2S 2 up to Na,S 5 , according to the amount of sulphur employed. On pouring such a solution into cold dilute hydrochloric acid an oil separates out which, by distillation under low pressure, yields two compounds, having the formulae H 2 S 2 and H 2 S 3 . The hydrogen disulphide is at ordinary temperatures a yellowish, water- clear liquid with a consistency somewhat like that of water. With alkalies it decomposes violently. Under ordinary pressure the liquid boils with partial stability at 74-75. Its specific gravity is 1.376. Its fumes attack the eyes and mucous membranes vigorously. 75.] COMPOUNDS OF SULPHUR WITH THE HALOGENS. 121 Hydrogen trisulphide at ordinary temperatures is a bright yellow liquid somewhat more mobile than olive oil. Its specific gravity at 15 is 1.496. The odor reminds one of sulphur chloride and camphor. The liquid solidifies at 52. On warming, it turns darker, becomes more viscid and, at about 90, begins an active evolution of hydrogen sulphide. Alkalies produce vigorous decomposition. The sensitiveness of these compounds toward alkalies is so great that they can only be prepared and kept in glass vessels whose inside surfaces have been previously freed from traces of alkali by treating with an acid. Compounds of Sulphur with the Halogens. 75. If chlorine is conducted over molten sulphur, sulphur monochloride, S2G12, is formed. Its formula is based on its vapor density and analysis. It is a yellow liquid of a very disagreeable, pungent odor, which excites one to tears. It boils at 139 and possesses in a high degree the ability to dissolve sulphur as much as 66% at ordinary temperatures. This solution is a thick syrupy liquid. It is used in the vulcanizing of rubber. Sulphur monochloride is slowly decomposed by water: 2S 2 C1 2 + 2H 2 O = SO 2 + 3S + 4HC1. Two other compounds of sulphur and chlorine are known, SC1 2 and SCU. Sulphur dichloride, SC1 2 , is formed, slowly, when sulphur monochloride is mixed with liquid chlorine. The mixture has a yellow color at first but after a few days it turns red. A determina- tion of the vapor density of the red substance and of its lowering of the freezing-point of acetic acid or benzene leads to the formula SC1 2 . It should, however, be borne in mind that a mixture of sulphur monochloride and chlorine, S 2 C1 2 + C1 2 , must give the same molecular weight as the compound SC1 2 . The existence of the SC1 2 compound is proved not only by the above-mentioned change of color of mixtures of sulphur monochloride and chlorine but also by the following observations : (1) The composition of the vapor given off from fresh mixtures of sulphur monochloride and chlorine is entirely different from that of the vapor given off after the mix- ture has turned red. (2) Mixtures of sulphur monochloride and 122 SB INORGANIC CHEMISTRY. [75- chlorine decrease in volume, and this diminution is greatest when the composition corresponds to S 2 C1 2 + C1 2 . (3) On distilling under 4 mm. pressure 80-90% of the liquid could be distilled over almost constant at 24. At ordinary pressure the compound boils at about 59-60 with decomposition. Sulphur tetrachloride, SC1 4 , can be obtained as a fine white powder, apparently not crystalline, when a chlorine-sulphur mixture of the composition S + CU or S + Cle is cooled to 75 (where it solidifies) and then centrifuged. The tetrachloride melts at about 33, giving off chlorine abundantly. The exact tem- perature could not be determined. Here, too, the form of the freezing-point curve indicates unquestionably the formula of the substance, viz., SC1 4 . Cf. 237. With bromine and iodine sulphur gives analogous com- pounds. Fluorine unites with sulphur to form a gas of the formula SFe and of rather surprising properties. It is colorless, odorless and incombustible. At 55 it solidifies with the formation of crystals. Notwithstanding its high percentage of fluorine it is chemically so indifferent that it almost resembles nitrogen in this respect (see p. 164). For instance, it is not decomposed by fused alkalies nor by copper oxide at dull-red heat. It can be heated with hydrogen without yielding hydrogen fluoride. Moreover, sodium can be fused in sulphur hexafluoride without losing its metallic surface, the gas not being attacked by the metal till the boiling- point of the latter is reached. VALENCE. 76. Certain elements have the property whereby their atoms can combine with only one atom of another element. The halogens on the one hand and hydrogen on the other are able to form only compounds of the type HX(X = halogen). This property of the atoms is called univalence. In the case of other elements like oxygen and sulphur each atom can enter into compounds with two univalent atoms (exam- ples: H 2 S, H 2 0). These are therefore called bivalent. The number of univalent atoms that can combine with one atom of a given element serves in an analogous way as a measure of valence in general. An atom of nitrogen, for instance, unites 76.] VALENCE. 123 with three atoms of hydrogen; nitrogen is therefore trivalent; carbon is quadrivalent, etc. /H The valence is ordinarily indicated by lines, as in O\ and /H X H N H, each line representing a valence unit (unit bond). \H The valence of one and the same element may be different according to the nature of the univalent elements with which it combines. Sulphur, for instance, can only unite with two hydrogen atoms, but with univalent chlorine it forms the compound SC1 4 , with fluorine even SF 6 . The valence of sulphur in these cases is therefore four and six. The preparation of sulphur compounds with more than six univalent atoms has not yet been accomplished ; hence its maximum valence is six. The halogens are univalent towards hydrogen, but in relation to each other they display more than one valence, as may be seen from the compounds IC1 3 and IC1 5 ; in the compound C^Oy ( 60) the maximum valence of chlorine can even be assumed to be seven. It has been very generally observed that when the maximum valence of an element is an even or an uneven number, its lower valences are of the same sort; the halogens and sulphur illustrate this. However, these are exceptions to this rule. The valence also depends upon the temperature. We shall soon see that SO 3 dissociates at a high temperature into SO 2 and oxygen; while sulphur is f //\ sexivalent towards oxygen at lower temperatures ( S=O 1, it' is only quadri- \ \ o / / S\ valent towards oxygen above 700 fSr' J . The valence must also depend on the pressure, for the latter exerts a great influence on the dissociation. The basis for the above sort 'of formulae is the idea, borrowed from organic chemistry, that the atoms of a molecule may not assume any conceivable arrangement whatsoever, but that there is a definite order in every molecule. For some extensions of the idea of valence see 317. Valence of Ions. In the solution of an electrolyte the sums of all the positive and all the negative amounts of electricity must be equal, for the solution acts as electrically neutral. In a solu- tion of hydrochloric acid the positive charge of the H-ions must 124 INORGANIC CHEMISTRY. [76- be numerically equal to the negative charge of the Cl-ions and, since the same number of both ions are present, each Cl-ion must carry a charge equal, but opposite in sign, to that of an H-ion. In a sulphuric acid solution, however, the two H-ions together must possess just as much positive electricity as the SO4-ion nega- tive electricity. The SO/' ion is therefore called bivalent in re- spect to the hydrogen ion. It is readily seen how the valence of other ions can be determined in an analogous manner, for it is equal to the numerical value of their electrical charge, that of the hydrogen ion being taken as unity. Compounds of Sulphur with Oxygen. 77. Of those containing only the two elements three are known, viz., S 2 0a, SO 2 , and SO 3 . Especial importance attaches itself, however, only to SO 2 and SOs; the two others have been little studied. Sulphur Sesquioxide, S 2 O 3 . This is obtained when sulphur is treated with its trioxide. It is a blue liquid, which congeals to a malachite-green mass and is soluble in fuming sulphuric acid, giving a blue solution. On being warmed it breaks up into sulphur and the dioxide : 2S 2 3 =3S0 2 + S. Water decomposes it with the formation of sulphur, sulphurous acid and polythionic acids. SULPHUR DIOXIDE, SULPHUROUS ANHYDRIDE, S0 2 . 78. This gas occurs in nature in volcanic gases. It is formed when sulphur burns in the air or in oxygen; the well-known odor of burning sulphur is due to it. A little trioxide is also formed by this combustion. The laboratory method of preparation con- sists in decomposing sulphuric acid with copper. 2H 2 S0 4 + Cu = CuSO 4 + SO 2 + 2H 2 0. For this purpose concentrated sulphuric acid is heated with cop- per turnings, no action taking place at ordinary temperatures. The process can be explained by supposing that at the high tern- 78.] SULPHUR DIOXIDE. 125 perature of the reaction copper is oxidized by sulphuric acid to copper oxide with the evolution of sulphur dioxide: The copper oxide reacts of course with a second molecule of sulphuric acid, producing copper sulphate. The reduction of concentrated sulphuric acid by heating with charcoal is also a convenient method of preparation: 2H 2 SO 4 + C = 2H 2 + 2S0 2 + C0 2 . However, as this equation shows, the gas is obtained mixed with one third of its volume of carbon dioxide, from which it cannot be separated directly. Moreover, sulphur dioxide can be obtained by the action of oxygen on sulphur compounds, thus, e.g. by the roasting of pyrite in a current of air: FeS 2 + 30 2 = SO 2 + FeS0 4 . Pyrite. This reaction is employed on a large scale in the commercial manufacture of sulphuric acid. The action of sulphur on oxygen compounds also yields sul- phurous oxide, e.g. heating copper oxide or manganese dioxide with sulphur: 2CuO + 2S = Cu 2 S + S0 2 ; Mn0 2 + 2S = MnS + S0 2 . Finally, the dioxide is also formed by heating an oxygen com- pound (CuO) with a sulphur compound (CuS) : CuS+2CuO = 3Cu+S0 2 . Physical Properties. At ordinary temperatures and pressures sulphur dioxide is a gas. It has a peculiar taste and odor. It is easily liquefied, the boiling-point being 8. Its evaporation produces a marked depression of temperature, sometimes extend- ing to 50; at 76 it becomes solid. Liquid sulphur dioxide dissolves many salts, in some cases with a characteristic color. It 126 INORGANIC CHEMISTRY. [ 78- is very soluble in water; at 1 vol. H 2 O dissolves 79.79 vols. S0 2 , at 20 39.37 vols. S0 2 . Boiling the solution expels all the gas ( 83). Chemical Properties. Sulphur dioxide is an acid anhydride; its aqueous solution has an acid reaction and behaves in general like that of an acid ( 83). It is easily oxidized by oxidizing- agents to the trioxide. This occurs, for instance, when a mixture of sulphur dioxide and air or oxygen is passed over hot spongy platinum or platinum asbestos. In aqueous solution this oxidation takes place readily at ordinary temperatures. The oxidation of the dioxide can also be brought about by chlorine-water, bromine and iodine : C1 2 +2H 2 + S0 2 =H 2 S0 4 +2HC1; also by chromic acid, which is reduced to chromium sulphate, or by potassium permanganate, which is reduced to a mixture of manganese and potassium sulphates, and therefore loses its color: 2KMnO 4 + 5S0 2 + 2H 2 O = K 2 SO 4 + 2MnSO 4 + 2H 2 S0 4 . Lead peroxide glows faintly in a current of sulphur dioxide and is reduced to lead sulphate from brown to white: Pb0 2 + S0 2 =PbS0 4 . It is to its reducing action that the bleaching effect of sul- phurous oxide on vegetable coloring-matters is due. A red rose, for example, loses color in it. The gas probably reacts with water, setting hydrogen free, which latter effects the reduction and hence the bleaching: S0 2 +2H 2 O =H 2 S0 4 +H 2 . In this case, therefore, bleaching depends on a reduction; as a matter of fact the color returns in many instances, when the bleached article is exposed to the oxidizing action of the air. Silk, wool and straw, i.e. substances that cannot stand the chlorine bleaching, are whitened commercially with sulphurous oxide. It also finds use as an antiseptic. 78.] SULPHUR DIOXIDE. 127 The reduction of iodic acid by sulphur dioxide is sometimes employed as a test for the latter. For this purpose strips of paper are dipped in a solution of potassium iodate and starch, which turns blue in the presence of sulphur dioxide iodine being set free ( 47). If the reaction is carried out in dilute solution, a peculiar phenomenon is observed ; the blue color of starch iodide does not appear directly when the solutions of sulphur dioxide and iodic acid are mixed, but is with- held for a certain number of seconds (definite for every concentration at constant temperature), when it suddenly appears. The following reactions come into play : I. 3SO 2 aq + HIO 3 = 3H 2 SO 4 aq + HI. The hydriodic acid thus formed is at once oxidized by the iodic acid still present : II. 5HI + HI0 3 =3H 2 O + 6I. So long as sulphur dioxide is present, it reduces the iodine in this dilute solution to hydriodic acid: III. 21 + SO 2 aq + 2H 2 O = H 2 SO 4 aq + 2HI. Not until all the dioxide is used up by the reactions I and III does the free iodine suddenly appear according to II. There are some substances which are able to extract oxygen from sulphur dioxide, i.e. the latter can also act as an oxidizing- agent. Ignited magnesium ribbon continues to burn in sulphur dioxide, forming magnesium oxide and sulphur. Hydrogen sul- phide and sulphur dioxide have respectively an oxidizing and a reducing effect on each other, which follows mainly the equation: 2H 2 S + SO 2 =2H 2 O + 3S. Sulphur dioxide is decomposed by electric sparks into sulphur and the trioxide. The action of the electric sparks is to be ascribed solely to the sudden and enormous rise of temperature which they produce and the rapid cool- ing that immediately follows, for the gas particles which have become heated by the sparks are immediately cooled again by surrounding objects. As a result the products formed do not have time to react in the opposite direction. The correctness of this view was demonstrated by ST. CLAIRE DEVILLE with the help of an apparatus which made it pos- 128 INORGANIC CHEMISTRY. [78- sible to cool objects very rapidly from a very high temperature. This apparatus, the cold-hot tube, consists of a rather wide porcelain tube, which is heated to a bright glow in a furnace and which contains a concentric thinner metallic tube, through which cold water is forced so rapidly that the tube maintains a low temperature. When DEVILLE introduced sulphur dioxide into the space between the two tubes, it was seen after some time that the inner tube, which was made of silver-plated copper, had turned black because of the formation of silver sulphide, while at the same time the formation of sulphur trioxide could be detected (by its producing sulphuric acid with water, a precipitate being given by barium chloride). The composition of sulphurous oxide can be determined in the following manner: When sulphur burns in oxygen no change of volume is observed after cooling. Therefore just as many mole- cules of sulphurous oxide have been formed as oxygen molecules consumed. The sulphurous oxide molecule must therefore con- tain two atoms of oxygen. The specific gravity of the gas has been found to be 2.2639 (air = l), or 32.6 (O = 16), so that its molecular weight is 65.2. If we subtract 2X16 from this for two atoms of oxygen, there remains 33.2 for sulphur, the atomic weight of which is 32. We thus see that only one atom of sul- phur is present in the molecule of sulphurous oxide and that the formula of the latter is 862. SULPHUR TRIOXIDE, SULPHURIC ANHYDRIDE, S0 3 . 79. This compound is found in a small amount in the fumes of burning sulphur ( 78). As was stated above, oxygen and sulphur dioxide unite to form the trioxide in the presence of platinized asbestos. On the other hand, the trioxide breaks up into the dioxide and oxygen at an elevated temperature, so that the formation of the trioxide from the dioxide and oxygen is evidently a reversible process, which is expressed thus: If we call the pressure of SO 2 in the equilibrium condition pi, that 79. SULPHUR TRIOXIDE. 129 of O 2 p 2 and that of SO 3 p 3 , it follows from 51 that the equilibrium relation is expressed by where K is the equilibrium constant. The combination of sulphur dioxide and oxygen is easily accomplished (in the presence of platinum) at about 500; that is to say, the above equilibrium is shifted almost wholly to the right at this temperature. If the temperature is raised, the dissociation of trioxide begins and at about 1000 it is com- plete. The union of S0 2 and O 2 also occurs under the influence of ultra- violet rays. These rays are best produced by a quartz-mercury arc lamp. The gases that are to be exposed to the rays must also be contained in quartz vessels, since glass is opaque to ultraviolet rays. Furthermore, an equilibrium 2SO 2 +O 2 <=2SO 3 also establishes itself under the action of these rays; for not only is the union of SO 2 and 2 incomplete, but SO 3 , on the other hand, breaks up under the same experimental conditions, yielding the same equilibrium mixture. This light equilibrium differs, however, in many respects from the " tem- perature equilibrium." In the first place, S0 3 , in the presence of platinum, does not begin to dissociate perceptibly until 300. The light influence, however, is evident even at room temperature. Like the effect of catalyzers, the action of light is retarded by sufficiently careful drying of the gases. There is an optimum moisture content for the con- tact process, but the action of light is effective even when the gases are passed through the illumination vessel in a very moist condition. The light equilibrium is not perceptibly affected by a marked change of tem- perature, but it is very sensitive to varying intensity of illumination. Just exactly as the dissociation increases with rising temperature, so it increases as the illumination grows stronger. The reader can form an idea of the extent of the decomposition from the observation of COEHN and BECKER that, with a mercury lamp consuming 9 amp., the equi- librium established itself when about 35% of the SO 3 was decom- Sulphur trioxide can also be obtained by heating certain sul- phates; in the arts ferric sulphate is thus used: 130 INORGANIC CHEMISTRY. [80- " Fuming sulphuric acid" (oleum) is a solution of sulphur trioxide in sulphuric acid ; the anhydride can be obtained from it by distillation. 80. Physical Properties. Perfectly dry sulphur trioxide melts at 17.7 and boils at 46. It looks much like ice but usually appears in another modification, viz., long asbestos-like needles with a silky lustre. These crystals have no sharp melting-point but sublime on heating. This modification is the stable one, for the other goes over into it spontaneously. This transformation is greatly accelerated by traces of water. The asbestine modifica- tion consists of double molecules (863)2, the glacial form of simple molecules (863). This is shown by the depression of the freezing- point of phosphorus oxychloride. The first is therefore called a polymer of the second. It is also worth noting that the 863 modification is very readily soluble in concentrated sulphuric acid, while the other, (863)2, dissolves with difficulty. Chemical Properties. Sulphur trioxide unites very easily with water to form sulphuric acid: SO 3 +H 2 0=H 2 S0 4 . It therefore fumes vigorously when exposed to moist air. On introducing it into water, combination and great evolution of heat, accompanied by sizzling, results. It reacts energetically with many metallic oxides also, forming sulphates. Baryta, for exam- ple, glows in contact with it. When its vapor is passed through a red-hot tube, it is decomposed into the dioxide and oxygen. Composition. The decomposition just mentioned permits us to establish the composition of sulphuric oxide. The dissociation products, 862 and 62, are formed in the volume ratio 2:1. Now the specific gravity of sulphuric oxide is 2.75 (air = l), from which the molecular weight is calculated to be 79.1. This figure corre- sponds to the formula 863 (32 + 3X16) and it also harmonizes with the above dissociation; for it is clear that 2 vols. 863 must then yield 2 vols. 862 and 1 vol. 2 : 2S0 3 =2SO 2 + 2 . 2 vols. 2 vols. 1 vol. 82.] OXYGEN ACIDS OF SULPHUR. 131 Oxygen Acids of Sulphur. 81. Sulphur forms an unusually large number of acids with oxygen and hydrogen, namely nine. They are as follows: 1. Thiosulphuric acid H 2 S 2 3 . 2. Hyposulphurous acid H 2 S 2 4 . 3. Sulphurous acid H 2 SO 3 . 4. Sulphuric acid H 2 S0 4 . 5. Persulphuric acid H 2 S 2 8 . 6. Dithionic acid H 2 S 2 6 . 7. Trithionic acid H 2 S 3 6 . 8. Tetrathionic acid H 2 S 4 6 . 9. Pentathionic acid H 2 S 5 6 . It is an important fact, however, that of these nine acids only sulphuric acid has really been isolated; all the others are known only hi aqueous solution or hi the form of salts. The two hydro- gen atoms which each of these acids possesses are both replaceable by metals; they are therefore dibasic acids. With such acids it is possible that just one of the hydrogen atoms be replaced by a metal. The salts thus formed are called acid salts. By different methods, e.g. the cryoscopic method, it is found that the aqueous solution of dibasic acids AH 2 contains chiefly the ions H" and HA'; it is only when these solutions are very dilute that the anion HA' splits up further into H* and A". In the case of the M 2 A salts, however, there is an ionization into 2M*+ A"; but in that of the acid salts MHA the ions are chiefly M* and HA'. How far the anion HA' is split up does not depend merely on the concentration, but also to a considerable degree on the strength of the acid, HA' being more ionized in strong than in weak acids of the same concentration. THIOSULPHURIC ACID, H 2 S 2 3 . 82. This acid can only exist in dilute aqueous solution and is even then very unstable, decomposing completely in a short time. The salts are, however, stable and can be prepared in the following ways: 132 INORGANIC CHEMISTRY. [82- 1. By boiling the solution of a sulphite with sulphur; Sodium sulphite. or S0 3 " + 8 = 8203'', only the anion being changed. 2. By the oxidation of sulphides in the air* 2 =2CaS 2 3 . Calcium disulphide. 3. By the action of sulphur dioxide on the solution of a sul- phide: 4Na 2 S + 6SO 2 = 4Na 2 S 2 O 3 + S 2 . The most important salt is the sodium thiosulphate, formerly and even yet often called sodium hyposulphite, or, abbreviated, "hypo." It is very soluble in water; the solution, when used in excess, has the property of dissolving readily the halogen com- pounds of silver, hence its extensive use in photography ( 247). It is easily oxidized by oxidizing-agents, usually to the sulphate. This takes place with potassium permanganate, nitric acid and chlorine, for example. Practical use is also made of this latter property by employing sodium thiosulphate as an antichlor in bleaching, i.e. to remove the last traces of chlorine which cling to the bleached material very obstinately and have an injurious effect. When a dilute acid is added to a dilute solution of sodium thiosulphate, the following decomposition takes place: = 2NaCl + H 2 + SO 2 + S ; or S 2 3 "+ 2H' = HS0 3 '+ H' + S. iipi id. Anion of sulphur- ous acic It may be, however, that the ions first unite partially to form H 2 S 2 3 , which splits up into H 2 O, S and SO2. It is an interesting fact that in this decomposition in a dilute solution the sulphur precipitate is not at once visible, being first noticeable after some seconds, or even minutes, according to the dilution. It was formerly supposed that the thiosulphu'ric acid remained entirely unchanged until the appearance of the sulphur and the decomposition first began at this 84.] HYPOSULPHUROUS, AXD SULPHUROUS ACIDS. 133 moment. This is, ' however, incorrect; for when a dilute solution of thiosulphate is treated with an equivalent amount of dilute acid and the solution again neutralized before the appearance of the sulphur deposit, it is found that the latter appears nevertheless after some time. A certain part of the free thiosulphuric acid must therefore have already decomposed, but the sulphur was in a so very finely divided atate in the liquid that it could not at once be detected, not until it !aad gathered together to form larger particles. Hyposulphurous Acid, 83. As early as the 18th century it was observed that zinc is dis- solved by a solution of sulphur dioxide in water without the evolution of hydrogen. SCHUTZENBERGER was, however, the first to show that a particular acid is formed thereby. A salt of this acid is produced by the action of zinc on a solution of acid sodium sulphite, NaHS0 3 , or by the electrolysis of such a solution, the nascent hydrogen acting as a reducing- agent.. Hyposulphurous acid, as well as its salts, is characterized by a vigorous reducing power. It precipitates the metals from solutions of sublimate (HgCl 2 ), silver nitrate and copper sulphate. Iodine solution is bleached by it with the formation of hydrogen iodide ; indigo is reduced to indigo- white. The solution is also very easily oxidized by free oxygen. It is therefore used to determine the amount of oxygen dissolved in water. For this reason it must be kept in well-stoppered vessels. BERNTHSEN succeeded in preparing the solid sodium salt, which proved to have the composition Na 2 S 2 4 + 2H 2 0, so that the acid itself has the formula H 2 S 2 O 4 . This salt was isolated by preparing a concen- trated solution of it and precipitating it by the addition of a suitable amount of solid common salt. The above formula is also confirmed by a direct synthesis of the sodium salt by MOISSAN, who obtained it by the action of dry sulphur dioxide on sodium: SULPHUROUS ACID, H 2 S0 3 . 84. It is taken for granted that the aqueous solution of sulphur dioxide contains sulphurous acid, H 2 SO 3 , for this solution reacts acid, conducts the electric current, gives salts with bases and evolves hydrogen with some metals, e.g. magnesium. The solution of sulphur dioxide in water does not conform to the law of HENRY ( 9) at ordinary temperatures, which proves that a combina- tion with the solvent has taken place. At higher temperatures, 134 INORGANIC CHEMISTRY. [84- however, the solution obeys this law pretty well. A fact in con- firmation of this is that all the sulphur dioxide can be expelled from the solution by boiling it, the combination being then wholly destroyed. The compound H 2 SO 3 itself has, however, not yet been isolated. The salts have the composition M2SO3 and MHSOs (M being an atom of a univalent metal) . The acid salts are almost all soluble in water, while of the neutral salts only those of the alkalies are soluble. The acid sodium sulphite, NaHSO 3 (sodium bisulphite), is frequently employed in organic chemistry. Sul- phites in solution gradually absorb oxygen from the air, form- ing sulphates. It is a very strange fact that minute quantities of organic substances, e.g. only 0.1% of alcohol and as little as 10~ 5 gram molecule of stannous chloride, greatly hinder this oxida- tion. We have here one of the few examples of a retarding catalytical action. On the other hand, traces of copper sul- phate considerably accelerate the oxidation. SULPHURIC ACID, H 2 S0 4 . 85. Sulphuric acid is the most important acid of sulphur. It can be obtained in various ways; in the first place by direct synthesis from its elements. According to 79 sulphur trioxide can be formed directly from sulphur and oxygen, and this yields sulphuric acid on the addition of water. The acid can be obtained from its salts by distilling them with phosphoric acid. Its formation from the action of oxygen on sulphur compounds is illustrated by the oxidation of an aqueous SCVsolution by the air. On the other hand the action of sulphur on oxygen compounds may also give sulphuric acid; thus it is formed when concentrated nitric acid, HNOs, is boiled with sul- phur; and again, potassium sulphate is formed by heating sulphur with saltpetre (KNOs). 86. For the commercial manufacture of sulphuric acid two processes are now in use, the lead-chamber process and the con- tact process. Enormous amounts of the acid are produced by these two methods. The lead-chamber process is based on the follow- ing reactions: 1. the oxidation of sulphur dioxide by nitric acid in the presence of water; 2. the oxidation by the oxygen in the 86.] SULPHURIC ACID. 135 air of lower oxides of nitrogen formed from the nitric acid in the previous reaction. These are partly reconverted to nitric acid and partly changed to certain stages of oxidation of nitrogen which oxidize sulphur dioxide anew to sulphuric acid. By this last process the lower nitrogen oxides are again formed, but are soon reoxidized by atmospheric oxygen and so on. One might suppose that a certain amount of nitric acid would suffice to con- vert unlimited amounts of sulphur dioxide into sulphuric acid with the aid of the air. In practice this is not true, however; for the nitrogen oxides are to a small extent still farther reduced by sulphurous oxide, so that nitrous oxide or nitrogen are formed, and these are no longer able, under the conditions of the indus- trial process, to combine with oxygen. The chemical processes which lie at the basis of the manu- facture of sulphuric acid will be taken up a little later ( 128). From a technical standpoint the lead-chamber process falls into three separate parts: 1. The preparation of sulphur dioxide; 2. The oxidation of sulphur dioxide; 3. The concentration of the resulting acid. (1) The material for the production of the dioxide is sulphur or pyrite (iron pyrites, FeS2). Sulphur yields a purer acid than pyrite; that prepared from the latter almost always contains arsenic. The roasting of the pyrite is carried on in furnaces, the construction of which varies considerably. In all of them, how- ever, the sulphur dioxide leaves the furnace mixed with a good deal of air. The furnace gases pass through a canal in which the dust particles carried along by the draught are deposited. (2) The oxidation of the sulphurous acid is carried out in ? structure consisting chiefly of three parts, the Glover tower, the lead chambers, and the Gay-Lussac tower. The gases enter the bottom of the Glover Tower, which is made of sheet lead lined with acid-proof brick. It is filled with lump stone, over which is laid a layer of smaller pieces of coke. On top of the tower is a reservoir for collecting the nitroso sulphuric acid (see below) that conies from the Gay-Lussac tower and the lead chambers and is to be concentrated in the Glover tower. It flows down over the stone in the tower from a horizontally revolv- ing tube. From the Glover tower the gases enter the lead cham- 136 INORGANIC CHEMISTRY. [ 86. bers. These are three or four in number and have a total capacity of 4000-5000 cubic meters. Their form is that of a parallelepiped, whose cross-section is nearly a square. Lead has been chosen as the material for the walls of the chambers, because it is the only one of the common metals which is only slightly attacked by sulphuric acid and the substances used in its manufacture. The lead chambers are connected with each other, with the Glover tower and with the Gay-Lussac tower by means of lead pipes. The first two chambers are also furnished with openings for introducing steam. The oxidation of dioxide to trioxide having been accomplished in the lead chambers, the residual gas, principally nitrogen, passes to the Gay-Lussac tower. Usually this is entirely filled with coke. On top of the tower is a reservoir containing 60-62 sulphuric acid (BAUME, see 88), which comes from the Glover v tower. The Gay-Lussac tower serves to collect the nitrous vapors that are still present in the gas as it leaves the lead chambers. These vapors dissolve in the sulphuric acid, forming the nitroso sulphuric acid which is used in the Glover tower. In this way the loss of nitric acid is much reduced. Let us now examine the task that befalls each of these three the Glover tower, the lead chambers and the Gay-Lussac tower. The gases that come from the pyrite furnace consist of a mix- ture of sulphur dioxide and air, a larger proportion of the latter than is required for the oxidation. They have a temperature of about 300 when they enter the Glover tower, A, at the opening, w. The gas current rising in the tower meets an acid mixture flow- ing down from above. The latter consists of the nitroso acid from the Gay-Lussac tower, diluted with the acid (chamber acid) (3) The acid produced in the chambers contains about 67% H2S04 (53 BAUME). In this condition it is employed directly in the manufacture of fertilizers ("superphosphate")- For almost all other purposes it must ^rst be concentrated. Ordinary sulphuric acid of commerce is of about 66 B. (B. = BAUME), i.e. 96-98% H 2 S04. It is prepared from the chamber acid by evap- orating it first in lead pans to about 78% (60 B.) and finally in a platinum vessel, 86.] SULPHURIC ACID. - 137 This crude sulphuric acid of commerce ("oil of vitriol") still contains various impurities and is usually more or less brown in color because of bits of straw (from the packing of the carboys) falling in and charring. It can be purified by diluting it, where- upon the dissolved lead sulphate is precipitated, and then stirring in a little barium sulphide. The latter produces insoluble barium sulphate, and also hydrogen sulphide, which precipitates any arsenic or lead ( 206) still present. The acid is then decanted from the deposit, concentrated, and finally distilled. The contact process . It has already been stated that sulphur dioxide unites with oxygen directly to form the trioxide and that the combination is considerably accelerated by the cata- lytic influence of platinized asbestos. This simple reaction is the basis of the " contact process." In practice, however, air is used instead of pure oxygen. The process falls into four separate parts: 1. The preparation of a mixture of sulphur dioxide and air; 2. The purification of this mixture; 3. The formation of the trioxide; 4. The combination of sulphur trioxide with water to form sulphuric acid. (1) The purification of the gas mixture is much the same as in the lead-chamber process. For reasons which will soon be made clear it is found necessary to conduct the roasting in the presence of a large excess of oxygen. While the equation 2S0 2 + O 2 =2SO 3 demands only 1 vol. O 2 for each 2 vols. S0 2 , the gases are usually- mixed in the ratio of 3 vols. 2 to 2 vols. SO 2 . (2) The platinized asbestos acts efficiently only when the furnace gases are absolutely pure, i.e., when the mixture consists simply of sulphur dioxide and air. The complete purification of these gases has been a problem of exceptional difficulty, but has been accomplished through the perseverance of KNIETSCH of the "Badische Anilin- und Sodafabrik," the great chemical factory at Mannheim, Germany. In the first place the furnace gases must be wholly freed from dust, else the catalyzer would soon become so coated as to lose its activity. In order to determine when the gas is really dust-free it is subjected to the "optical test," i.e., it is passed through a tube closed at both ends with glass, and is 138 INORGANIC CHEMISTRY. [S6- examined with the eye to see whether it is perfectly transparent and free from nebulous masses. Even when this optical test is quite satisfactory the catalyzer suffers a loss in activity if the gas is not entirely free from arsenic compounds; the least traces of the latter have an injurious effect. The presence of arsenic compounds in the furnace gas is due to the occurrence of arsenic in the pyrites ( 86 ; 1) used for roasting. KNIETSCH has finally succeeded in completely eliminating the arsenic compounds by blowing steam into the gas mixture. (3) As already set forth in 79, the equilibrium + O 2 <=2SO 3 is expressed by the equation According to this equation the formation of sulphur trioxide is more complete in the presence of an excess of either sulphur dioxide or oxygen, for as p\ or p 2 increases p 3 must also increase. Since the object in view is to convert the dioxide as completely as possible into the trioxide, it is advantageous to provide a large excess of oxygen. This explains why more than the theoretical amount of oxygen is taken. Compare (1). The equilibrium must also depend on the pressure, for, if this is increased n times, the equation becomes: or np 1 2 p 2 from which it is evident that at a higher pressure (n>l) the for- mation of the trioxide is more nearly complete ( 102, 5). The manufacturer does not find it necessary, however, to employ high pressure, which would involve, moreover, a great complica- tion of the apparatus. If it is desired that the combination of sulphur dioxide and oxy- gen should be as complete as possible, the temperature must be kept at about 400. Since, however, the heat of formation of the trioxide is great, viz., SO 2 +0=S0 3 + 22,600 Cal., 86.] SULPHURIC ACID. 139 the apparatus must be cooled. This is done most practicably by the aid of a fresh portion of the gas mixture, as the next paragraph sets forth. The construction of the apparatus is as follows: The tubes ab (Fig. 31) contain the platinized asbestos 6, supported on little sieves (shown in the middle tubes). The purified furnace gases first pass around the outside of the tubes and are thus warmed to FIG. 31. CONTACT-PROCESS APPARATUS. the desired temperature at the heat expense of the gas system within. When the proper temperature is reached the gases are allowed to enter the tubes, where sulphur trioxide is formed with the evolution of more heat. By increasing or diminishing the rate of flow of the gas current the temperature can be regulated very satisfactorily. When the operation is started the apparatus must first be warmed to 400. (4) The reaction between sulphur trioxide and water is an energetic one. Nevertheless, the manufacture of sulphuric acid 140 INORGANIC CHEMISTRY. [86- from these two compounds involved some difficulty, inasmuch as sulphur trioxide fumes invariably escaped when this substance was introduced into water or dilute sulphuric acid. Only when sul- phuric acid of 97-98% is used as the absorbent and care is taken to keep the acid at this concentration by the simultaneous addition of water does a complete and immediate absorption occur. This is due to two circumstances: first, that traces of water change sulphur trioxide into the asbestine modification ( 80), which is only slowly absorbed by sulphuric acid ; second, that at the concentration of 97-98% H 2 SO 4 the system xS0 3 +yR 2 O has a minimum of vapor tension, which is very low. 87. Physical Properties. The pure compound, hydrogen sul- phate, is an oily liquid at ordinary temperatures, solidifying at a low temperature and melting again at + 10.0. Its specific gravity in the liquid state (15) is 1.8500. Chemical Properties. The concentrated acid obtained by dis- tillation is not the simple compound H^SCU, for it still contains about 1.5% of water. In order to prepare the absolutely pure acid the distilled product must be mixed with the theoretical amount of sulphur trioxide. When pure sulphuric acid is heated, it begins at 30 to give off fumes of sulphur trioxide; this continues until the boiling-point, 317 at 750 mm. Hg. pressure, is reached, when an acid with 1.5% water distils over. On heating the vapor of sulphuric acid above the boiling-point, it begins to break up into water and the anhydride; this dissociation is complete at 450, for the vapor density at that temperature is found to be 25.1, while that of S0 3 + H 2 O is theoretically 24.5. When sulphuric acid is mixed with water, a strong evolution of heat occurs. The mixing must therefore be done with great care, particularly in glass vessels, the acid being poured in a fine stream into the water and the. liquid being steadily stirred. On mixing them in the reverse way, by pouring the water into the sulphuric acid, the intense heat that is produced may cause the glass to crack. However, when the acid is mixed with ice in a certain proportion, a strong cooling follows. The mixing of sulphuric acid and water is attended by a con- traction, i.e. the volume of the dilute acid is smaller than the sum of the volumes of water and acid. It is known that sulphuric acid is able to form hydrates with water ( 237). 87.] SULPHURIC ACID. 141 Sulphuric acid is a strong dibasic acid, but not as strong as hydrochloric acid, for, while the latter is ionized to 95% at a dilution of 0.1 gr. mol. per L, sulphuric acid at the same dilution is only ionized to 55% into 2H' + S0 4 ". At higher concentra- tions HSO 4 ' ions also exist. It acts on many metals, giving off hydrogen. This action is made use of, as stated above, in the preparation of hydrogen; the acid must, however, be dilute, for when it is too strong or warmed, the hydrogen gen- erated partially reduces the sulphuric acid so that the gas given off contains hydrogen sulphide. Sulphur dioxide also is formed when hydrogen is led into hot sulphuric acid. It is upon this action that the reaction of copper with hot concentrated sulphuric acid depends ( 78). Mercury, silver and certain other metals are similar to copper in their behavior. Platinum and gold are not attacked by the acid. Sulphuric acid makes holes in paper, linen, dress goods and the like, when dropped on them. It has a destructive, charring effect on organic substances in general. This is due in many cases to the great tendency of the acid to unite with water, which makes it not only deprive other substances of the water they con- tain, but even withdraw the hydrogen and oxygen from organic compounds to form water. On the other hand, sulphuric acid gives up oxygen to many organic substances, being itself reduced. In order to detect free sulphuric acid in vinegar, for example, the liquid is evaporated on a water-bath with a little sugar. Free sulphuric acid, if present, chars the sugar during the concentration. The most of the salts of sulphuric acid (sulphates) are soluble in water. Barium, strontium, and lead sulphates are insoluble, while calcium sulphate (gypsum) is slightly soluble, but only to a very small degree. The formation of barium sulphate, BaS0 4 , serves as a characteristic test for sulphuric acid, or, as we may better say, for the ion $0 4 ". The sulphates are in general very stable. They can, for instance, be heated to very high temperatures without decomposi- tion. The acid salts lose water on heating, and pass over into* pyrosulphates : 2NaHSO 4 = H 2 O + Na 2 S 2 O 7 . Sodium nvrosulohate. 142 INORGANIC CHEMISTRY. [ 87- If these pyrosulphates are heated still higher, they give off sulphur trioxide and form neutral salts: 88. Uses. Sulphuric acid is of enormous practical value, its uses being most varied. It is employed in the preparation of almost all other mineral acids from their salts. In the manufacture of soda after LE BLANC it is used in astonishingly large amounts and in nearly all other branches of chemical industry it is of some service or other. In the laboratory it is often employed as a drying-agent. A moist substance is dried very thoroughly when placed in a closed apparatus near a dish of the concentrated acid. For this purpose special pieces of apparatus are constructed, called desiccators. The determination of the concentration of sulphuric acid is an operation that is frequently necessary. Ordinarily the specific gravity is made use of, for this can be determined rapidly with an areometer. There are tables so prepared that the proportion of H 2 SO4 or 80s in a dilute acid whose specific gravity and tempera- ture are known can be quickly read. BAUME, a chemist of the latter part of the eighteenth century, constructed an areometer with an arbitrary scale, the zero point of which indicates pure water and the point 10 being reached in a 10% salt-solution. All the divisions are equal. 100% H 2 S04 would then be represented by the line 66.6. In the arts the strength of sulphuric acid is still given as so many " degrees BAUME." Fuming sulphuric acid is the name of a sulphuric acid that contains sulphur trioxide in solution. It is obtained by dissolving the oxide in concentrated sulphuric acid. Fuming sulphuric acid is a thick oily liquid, which fumes vigorously in the air, throwing off the trioxide. Sp. g. = 1.85-1.90. CHLORIDES OF SULPHURIC ACID. 89. When phosphorus pentachloride acts on sulphuric acid a com- pound SO 3 HC1, chlorosulphonic acid, is formed : H 2 S0 4 + PC1 5 = S0 3 HC1 + POC1 3 + HC1. The same compound results from the direct union of sulphur trioxide and hydrochloric acid. It is a colorless liquid, which fumes vigorously on 91.] CHLORIDES OF SULPHURIC ACID. 143 exposure to the air. Sp.g.= 1.766 at 18. Boiling-point, 158. On the addition of water a violent reaction occurs, producing hydrochloric acid and sulphuric acid: SO 3 HC1 +H 2 O= H 2 SO 4 +HC1. 90. A compound, SO 2 C1 2 , sulphuryl chloride, is obtained by the direct union of sulphur dioxide and chlorine, most easily by first saturating camphor with sulphur dioxide (which readily dissolves in it) and then passing chlorine over it. The camphor remains unchanged. Sulphuryl chloride is a colorless liquid, which boils at 69.1, has a penetrating odor, fumes strongly in the air and has a specific gravity of 1.6674 at 20. The addition of a little water converts it into chlorosulphonic acid and hydrochloric acid, much water to sulphuric and hydrochloric acids: SO 2 C1 2 + H 2 O=S0 3 HC1+ HC1. SO,C1 2 +2H 2 O= H 2 SO 4 +2HC1. These decompositions of sulphuryl chloride can be represented in the following way : OH In the place of the two chlorine atoms we have, therefore, two OH (hydroxyl) groups entering. For this reason it is assumed, in close analogy with the methods of organic chemistry, that sulphuric acid con- tains two hydroxyl groups. Sulphuryl fluoride can be obtained by the direct union of sulphur dioxide and fluorine. It has the same remarkable stability as the com- pound SF 8 ( 75). It is a colorless and odorless gas, liquid at 52 and solid at 120. It can be heated with water in a sealed tube to 150 without undergoing decomposition. Alkalies absorb it, though very slowly. Sodium can be fused in it without being attacked. Persulphuric Acid, H 2 S 2 O 8 . 91. The potassium salt, K 2 S 2 O 8 , or, still better, the ammonium salt, (NH 4 ) 2 S 2 8 , of this acid can be obtained by the electrolysis of a cold saturated solution of the corresponding sulphate in sulphuric acid of 1.3 sp. g. In such a solution we may assume we have the ions K* and HSO/; the latter are discharged at the anode and can then unite to form H 2 S 2 O 8 , which forms with the K ' ions present the difficultly soluble potassium salt K 2 S 2 O 8 . This separates out as a white crystalline mass. 144 INORGANIC CHEMISTRY. [91- However, the combination of two HS0 4 groups only takes place when their concentration at the anode is quite high; for if this is not the case there is more opportunity for secondary reactions, such as a union with water to form 2H 2 SO< and 20H, the latter of which is decomposed into H 2 O and O. Such a high concentration at the anode is reached by using a very small electrode. The electric current therefore has a high density at the anode; that is, a large quantity of electricity must pass through a small surface. The effect thereof is that this large quantity discharges a great many HSO/ ions into a small space, or in other words, produces enough HS0 4 groups to make the concentration ver} r high there. As low as 100 it decomposes in the following way: 2K 2 S 2 8 =2K 2 S 2 7 +0 2 . K-pyrosulphate. The barium salt of persulphuric acid is soluble in water, as are also most of the other known salts. The action of 100% hydrogen peroxide on sulphur trioxide or on chlorsulphonic acid yields CARD'S acid; SO 3 +H 2 O 2 = H 2 SO 5 , >OH X)H S0 2 = S0 2 +HC1. \C1+H 2 O 2 \O-OH It crystallizes very prettily and melts at about 45 with slight decom- position. CARD'S acid reacts with another molecule of chlorsulphonic acid according to the equation X)H X)H HO \O-OH XW-1.A J.J.W SO 2 ' +C1-SO 2 OH=SO 2 I \).0-S0 2 forming persulphuric acid, which can be obtained in this way pure and crystallized, with a melting-point of 60 (attended by slight decom- position). A solution of CARD'S acid in sulphuric acid can be prepared in a simple way by mixing H 2 O 2 with an excess of strong sulphuric acid. On the basis of this method of formation BAEYER gave the compound the name sulpho-mono-peracid. It has very strong oxidizing powers. It sets iodine free from potassium iodide, oxidizes sulphur dioxide to trioxide, and ferrous to 92.] POLYTHIONIC ACIDS. 145 ferric salts and also precipitates the higher oxides of silver, copper, manganese, cobalt, and nickel from solutions of salts of these metals. On the other hand, it neither bleaches permanganate solution nor oxidizes solutions of chromic and titanic acids; in these respects it is distinguished from hydrogen peroxide, to which it otherwise shows much similarity. POLYTHIONIC ACIDS. 92. Under this name are grouped four acids of the general formula H 2 S n O6, in which the number of sulphur atoms, n, can be 2, 3, 4 and 5, and this determines the names of the individual acids. Dithionic acid, H 2 S 2 6 . The manganese salt of this acid is obtained when finely powdered manganese dioxide is suspended in water and sul- phurous oxide passed in: 2S0 2 +Mn0 2 =MnS 2 O 6 . From this barium salt the dithionic acid can be liberated by sulphuric acid. The solution can ba concentrated in vacuo over sulphuric acid till its specific gravity reaches 1.347; farther concentration or warming results in a decomposition: Trithionic acid, H 2 S 3 O 6 . Potassium trithionate is formed when a solution of potassium thiosulphate is saturated with sulphur dioxide: 3SO 2 +2K2SA = 2K 2 S 3 O 6 + S. The free acid is unstable ; even at ordinary temperatures it decomposes in a dilute solution into sulphur, sulphurous oxide and sulphuric acid : H 2 S 3 O 6 =H 2 S0 4 +S+S0 2 . Tetrathionic acid. Its salts result from the action of iodine on the solu- tion of a thiosulphate. K 2 S 2 O 3 + 21 -2KI+ K 2 S 4 O 6 . The acid itself can be obtained (also only in dilute solution) by adding sulphuric acid to the barium salt, which is prepared in an analogous manner. In dilute solution it is quite stable; in the concentrated state it breaks up into sulphur, sulphurous oxide and sulphuric acid. Pentathionic acid. On mixing solutions of sulphur dioxide and hydrogen sulphide the principal reaction is a mutual oxidation and reduction of these compounds with the separation of sulphur ( 78). The action is, however, much more complicated, inasmuch as polythionic acids, among them pentathionic acid, are formed in addition at the same time. The mixture of H 2 S.aq and SO 2 .aq is known as " WACKENRODER'S liquid." Well-crystallized salts of pentathionic acid have been prepared. 146 INORGANIC CHEMISTRY. [93. Use of Sodium Thiosulphate in Volumetric Analysis. lodometry. 93. On adding sodium thiosulphate to an iodine solution, the intensely brown liquid loses its color, sodium iodide and sodium tetrathionate, two colorless compounds, being formed: 2Na 2 S 2 O 3 + 21 = Na 2 S 4 O 6 + 2NaI ; or, writing only the ions that take part in the reaction: The disappearance of the color is thus due to the fact that the molecules of iodine are transformed into ions by taking up. two negative charges from 2S 2 (V. Upon this fact a method is based for determining the amount of free iodine in a solution. This is done by allowing a solution of sodium thiosulphate, whose concentration (litre) is known, to flow drop by drop into a definite volume of iodine solution. (For letting out a certain amount of liquid a pipette (Fig. 32) is commonly employed.) The color gradually brightens and finally a point is reached when the liquid is only slightly tinged and the addition of another drop causes the color to entirely dis- appear. This transition can be very accurately detected. The iodine molecules have now entirely disappeared. Since according to the above equation a molecule of thiosulphate is consumed for each atom of iodine, the percentage of iodine in the solution can be calculated from the amount of thiosulphate used. To make the calculation of the result of such a determination (titration) as easy as possible the thiosulphate solution is so stand- ardized that it bears a certain relation to an equivalent of iodine ( = 127 g.), i.e. a certain amount bleaches exactly this much iodine. "Normal solution " is a name applied to a solution containing the equivalent weight ( 23) in grams (gram equivalent) in one liter. Frequently use is also made of a , J, T V or a twice, thrice, etc., normal solution. Normal hydrochloric acid contains 36.5 g. HC1, normal sulphuric acid 49 g. H 2 SC>4 (= i gram molecule), a normal iodine solution 127 g. iodine, per liter. Detailed direc- 93.] VOLUMETRIC ANALYSIS. 147 tions for preparing such solutions can be found in the text-books of analytical chemistry. In order to determine readily the volume of thiosulphate solu- tion that is required in the analysis, use is made of a burette (Fig. 33), a glass tube that is divided into -^ c.c. and closed at the lower end with a glass stop-cock or with a rubber tube and pinch-clamp. In titrating the iodine solution the thiosulphate A FIG. 32. PIPETTE. FIG. 33. BURETTES AND SUPPORT. solution is allowed to flow out slowly and, finally, drop by drop, while the liquid is being stirred. Example. For 50 c.c. of an iodine solution whose strength is to be determined 27.30 c.c. T \ normal thiosulphate solution was necessary before the color completely disappeared. Required the number of grams of iodine contained in 1 liter of this solution. 1000 c.c. T V normal Na 2 S 2 O3 solution (see above) decolorizes T V equivalent of iodine ( = 12.7 g.); 27.3 c.c. therefore decolorizes 12 7 27.3X innn g. iodine. This amount is contained in 50 c.c. of lUUU 148 INORGANIC CHEMISTRY. [93- the iodine solution in question. Hence 1 liter of the latter con- tains 20 X 27.3 X 12.7 X 10- 3 =6.8842 g. iodine. Various other substances which liberate iodine from potassium iodide can be determined by titrating the amount of iodine dis- placed; for example, chlorine and bromine may be thus determined f since they set free the equivalent amount of iodine from potassium iodide solution. SELENIUM. 94. Selenium was discovered by BERZELIUS in 1817. It took its name from (rekrjvrj (the moon) , because it possesses great similarity to the element tellurium (named from tellus=ihe earth) discovered a short time previously. It is rather widely distributed in nature, but it occurs only in small quantities. It is found native, is frequently found in pyrite and also appears in some rare minerals. When this sort of pyrite is employed in sulphuric acid manufacture, the selenium collects in the " chamber-mud " of the lead chambers; from this it is usually obtained. The process is as follows: The selenium deposit is heated with nitric acid, which oxidizes the selenium to selenic acid, H 2 SeO 4 . The solution thus obtained is first boiled with hydrochloric acid, whereby selenious acid, H 2 Se0 3 , is formed with the evolution of chlorine. This latter acid is then reduced by means of sulphurous oxide to selenium, which separates in amorphous red flakes. Selenium displays analogy with sulphur in many respects; for instance, in occurring. in various allotropic conditions. According to SAUNDERS, there is an amorphous red modification, that is soluble in carbon disulphide. From this solution the selenium separates as a second modification, which is the red crystalline selenium, fusing at 170-180. Then there is a metallic form fusing at 217. This modification appears when amorphous selenium is heated to 97, at which point a sudden and marked rise of tempera- ture occurs; or when molten selenium is suddenly cooled to 210 and kept for a time at that temperature. In this metallic state selenium has a metallic lustre, is insoluble in carbon disulphide and conducts electricity. Its conductivity strangely depends very much on the intensity of its illumination, however. The melting-point of selenium is 217, its boiling-point 680. As in the case of sulphur the vapor density decreases with rising 94.] SELENIUM. 149 temperature till about 1400 is reached, when it remains constant. At this temperature it is found to be 81.5 (H=l), corresponding to a molecular weight of 163.0. Now since the atomic weight of selenium, as deduced from the vapor density of its compounds, is 78.9, the above molecular weight agrees very closely with the formula 862. Hydrogen selenide, H 2 Se, can be obtained directly from its elements, as these unite at 400. Analogously to hydrogen sul- phide, it can also be got by the decomposition of iron selenide, FeSe, with hydrochloric acid. At a high temperature hydrogen selenide dissociates into its elements. Its properties are only slightly acidic and it is more poisonous than sulphuretted hydrogen. The heavy metals are precipitated from their solutions as selenides by it. An aqueous hydrogen selenide solution becomes turbid on standing because of the selenium that separates out. Two chlorine compounds, Se 2 Cl2 and SeCU, are known. The latter is much more stable than the corresponding sulphur com- pound, SOU ( 75). Selenium tetrachloride is solid and sublimes without decomposition; dissociation does not begin until 200 is reached. Selenium dioxide, Se0 2 , is the only oxide of selenium known. It results from the burning of selenium in the air. The extremely disagreeable odor which arises is not a property of the dioxide, however, but is probably due to the formation of another oxygen compound of selenium which has not as yet been isolated. Sele- nium dioxide forms long white needles that sublime at 310. Selenium dioxide is an acid anhydride; on dissolving it in water an acid, selenious acid, H 2 SeO3, is formed, which can be isolated (unlike sulphurous acid). This acid crystallizes in large colorless prisms. On being heated it breaks up into water and anhydride. Sulphur dioxide or stannous chloride reduce it to free selenium, which is deposited in red flakes: H 2 Se0 3 + 2SO 2 + H 2 = 2H 2 SO 4 + Se. Sulphuretted hydrogen precipitates from the solution selenium sulphide, SeS, insoluble in ammonium sulphide. When chlorine is passed into the solution of selenious acid or 150 INORGANIC CHEMISTRY. [94- when bromine is added to it, selenic acid, H 2 Se04, is formed. In the pure state this is a crystalline solid, melting at 58. The 95% solution of it is an oily liquid, which has the appearance of sul- phuric acid. The barium salt of the acid, like that of sulphuric acid, is extremely difficultly soluble. On boiling with hydrochloric acid, selenic acid is reduced to selenious acid with the evolution of chlorine. Tellurium. 95. Tellurium is of rare occurrence; it is known in the rative condi- tion and also in combination with bismuth, and with gold or silver (in sylvanite, or graphic tellurium}. It is found chiefly in Transylvania and in the Altai mountains, and also in Boulder Co., Colorado. In the amor- phous condition tellurium is a black powder, but after fusion it is silvery white, of a metallic lustre and a conductor of heat and electricity. The vapor density, as in the cases of selenium and sulphur, decreases with increasing temperature and does not remain constant till about 1400; it then corresponds to a Te 2 molecule. Hydrogen telluride, H 2 Te, results from the action of hydrochloric acid on zinc telluride, ZnTe. The product thus obtained contains more or less hydrogen. It is very poisonous, and dissociates readily. From solu- tions of the heavy metals it precipitates their tellurium compounds (telluride s). Tellurium dioxide, TeO 2 , is formed on burning tellurium in the air. It is very difficultly soluble in water. Tellurous acid, H 2 TeO 3 , is obtained by dissolving tellurium in nitric acid. It dissolves in water with great difficulty and breaks up on warming into TeO 2 and H 2 O. Telluric acid, H 2 Te0 4 , is prepared by fusing the metal or the dioxide with soda and saltpetre and separating the acid from the tellurate formed. The compound H 2 Te0 4 + 2?I 2 O crystallizes out from the aqueous solution; it loses its water of crystallization at 100. The free telluric acid, H 2 TeO 4 , prepared in this way is a white powder, difficultly soluble in cold water. Telluric acid has only feebly acid properties. Selenium and tellurium both combine with potassium cyanide, when they are fused with it, forming compounds corresponding to KCNS, viz., KCNSe and KCNTe. Nevertheless, while potassium t e 1 1 u r o-cyanide is at once decomposed by the oxygen of the air with the separation of tellurium, potassium s e 1 e n i o-cyanide is more stable and does not decompose with the separation of selenium until it is boiled with hydro- 96.] SUMMARY OF THE OXYGEN GROUP. 151 chloric acid. We have here a means of detecting selenium in the pres- ence of tellurium and of separating the two. SUMMARY OF THE OXYGEN GROUP. 96. The elements oxygen, sulphur, selenium and tellurium, like the halogens, form a natural group, particularly in two respects; their compounds correspond to a general type and their physical and chemical properties vary gradually with increasing atomic weight. Their hydrogen compounds have the formula RH 2 , their oxygen compounds and their acids the formulae RO 2 and H 2 RO3, and also ROs and H 2 RC>4. Ozone may be considered with reference to these types as analogous to sulphur dioxide; O-O 2 ozone; 8-62 sulphur dioxide. The following table shows the gradual change, or progression, of the physical properties : O. S. Se. Te. Atomic weight Specific gravity. . . . Melting-point . 16.00 1.124 (at -181) 32.07 1.95-2.07 119.5 79.2 4.2-4.8 217 127.5 6.2 452 Boiling-point . 181 4 450 680 white heat Color light blue yellow red black As the atomic weight increases, the values of the physical con- stants also increase, as the table shows. At the same time the external appearance approaches that of the metals; in tellurium the metallic appearance is quite marked. The instability of the hydrogen compounds increases from oxygen to tellurium; the strength of the oxygen acids diminishes rapidly, sulphuric acid belonging to the strongest, and telluric acid to the very weak, acids. It should also be noted that all of these elements appear in allotropic modifications. 152 INORGANIC CHEMISTRY. [97- THERMOCHEMISTRY. 97. It was stated above ( 20) that a chemical combination or decomposition is accompanied by an evolution or absorption of heat, in other words by a heat change, or caloric effect. In many cases this caloric effect has been carefully measured. The work of BERTHELOT and of THOMSEN along this line has been especially fruitful. That part of chemistry which deals particularly with these caloric effects is called thermochemistry. The caloric effect is always given for molecular amounts of the reacting substances, since in this way only is it possible to compare substances from a chemical standpoint. Hence, when the heat of formation of water is said to be 69.0 calories (kilogram calories), it is implied that this number of calories is evolved by the union of 2 g. hydrogen with 16 g. oxygen: 2H + O = H 2 O + 69.0 Cal. In this equation H and O stand for gram atoms. In expressing a caloric effect it is necessary to indicate the state of matter of the reacting and the resulting substances, in so far as this is not self-evident, because the latent heat of fusion or vaporization must be taken into consideration. The above amount, 69.0 Cal., refers to the formation of water and its conversion to a liquid. It therefore includes the heat of condensation. Since this amounts to 0.536 Cal. per gram, it would in this case (for 18 g.) be 9.6 Cal.; hence the caloric effect of the combustion of hydrogen to steam at 100 is 2H + = H 2 O g as + 58.4 Cal. The caloric effect is also influenced by the state of matter in which the substances react, i.e., whether solid, liquid, or gas, inasmuch as solution is almost always accompanied by a heat change. In the formation of sodium chloride by the mixture of dilute solutions of sodium hydroxide and hydrochloric acid (this being indicated by aq after the formulae of the substances) the caloric effect is: NaOH aq + HCl aq = NaCl aq + H 2 O + 13.7 Cal. 99.] THERMOCHEMISTRY. 153 However, when the salt is prepared by passing hydrochloric acid gas into a dilute solution of the base, the equation is as follows: NaOH aq +HCl g as =NaCl aq + H 2 O + 31.1 Cal. We thus obtain 13.7 Cal. as before, but increased by the heat of solution of gaseous hydrochloric acid in a large amount of water, viz., 17.4 Cal. The heat of formation of chemical compounds must be equal to their heat of decomposition, but have the opposite sign. Were this not the case, heat would be lost or gained when a compound is formed and then decomposed so as to return to the original con- dition, and such a result would be at variance with the Law of the Conservation of Energy. Experience has shown that in the formation of most compounds heat is generated, but that in many cases heat is absorbed. Chem- ical actions of the first sort are called exothermic, those of the second endothermic, reactions. An example of the second sort is the synthesis of chlorine monoxide: 2Cl + = Cl 2 O gas -15.1 Cal. 98. For the determination of the caloric effect various methods are in use. Only those actions are suitable for thermochemical measurements which complete themselves quickly. In measuring the caloric effect in the case of liquids or solutions, as, for example, the heat of neutralization of acids and bases, the heat of solution or of dilution, etc., an ordinary calorimeter is generally used, such as is employed in physics for the method of mixtures, the same precautions being taken in order to secure accurate results. The heat of combustion of a substance is usually measured with the calorimetric bomb of BERTHELOT-MAHLER. This is the usual method with organic compounds. 99. The Law of HESS. The entire caloric effect (the whole amount of energy) produced by the transformation of one chemical system into another is independent of all intermediate stages. This law is a direct consequence of the principle of the con- servation of energy. If HESS'S law did not hold, energy would have to be gained or lost in the transition from one system to another and the subsequent return to the initial condition, which 154 INORGANIC CHEMISTRY. [99- is contradictory to the above principle. A few examples will serve to make this law more clearly understood. (a) A dilute solution of sodium sulphate can be prepared from sodium hydroxide, sulphuric acid and water in various ways. For instance, two gram-molecules of the base can be treated at once with dilute sulphuric acid; or one gram-molecule of the base can be mixed with the acid at first and the second added afterward. Accordingly we get the following caloric effects: (1) 2NaOH +H 2 S0 4 aq -Na 2 SO 4 a q -2H 2 O = 31.4Cal. , C NaOHaq + H 2 SO 4 aq - NaHSO 4 aq - H 2 O = 14 . 75 ( } ( NaOHaq + NaHSO 4 aq - Na 2 S0 4 aq - H 2 = 1 6 . 65 Total... .......................... 31.4Cal. (6) From ammonia, hydrogen chloride and water a dilute solu- tion of ammonium chloride, NH 4 C1, can be prepared, either by letting dry ammonia gas combine with dry hydrogen chloride gas and dissolving the resulting ammonium chloride in water or by first dissolving ammonia and hydrogen chloride in separate por- tions of water and then mixing the solutions. In the first case we have the equations: NH 3 gas + HClgas - NH 4 Clsoiid ............ = 42 . 6 NH 4 Clsoiid + aq-NH 4 Claq .............. = - 4.0 38.6Cal. in the second case: -NH 3 a q ............... = 8.82 HCl + aq-HClaq ................ =17.13 NH 3 aq + HClaq - NH 4 Claq ........ = 12 . 45 38.40Cal. The final effects in the two cases are found to be alike within the limits of experimental error. With the help of HESS'S law the determination of the caloric effect is rendered possible in many reactions which cannot be dealt with directly or are unsuitable for calorimetric measure- ments. In general this is done by making thermochemical meas- 100.] THERMOCHEMISTRY. 155 urements for a series of processes in which the reaction plays a part and finally calculating the caloric effect of the reaction as the single unknown, as will be more fully explained in the examples below. Suppose it were required to find the heat of formation of hydro- gen sulphide. This compound can be formed directly from its elements ( 72), but the reaction is unsuitable for thermochemical study. We will therefore start with the system, H, S, and O, and consider the two ways by which it can form water and sulphur dioxide: (1) hydrogen and sulphur are burned directly to water and sulphur dioxide; (2) (a) hydrogen and sulphur combine and (6) the resulting hydrogen sulphide is burned to water and sulphur -dioxide. Since we started with the same system and in the end reached the same result in each case, the caloric effect must be the same according to HESS'S law, so that, if we measure (1) and (26), we can equate (1) and (2) and solve for (2a), thus: Heat of combustion of 2H-J-heat of combustion of S = heat of formation of H 2 S + heat of combustion of H 2 S. (2H + -H 2 0) + (S + 20 -S0 2 ) = (2H+S -H 2 S) + (H 2 S + 30 -S0 2 -H 2 0). 68.0 + 69.26 = x+ 133.46; .; % z=(S+2H-H 2 S) = 3.8. 100. In using these values of the heat of formation and heat of decomposition it should be noted that they do not represent the amounts of heat liberated by the combination of atoms to form molecules, but that the heat of decomposition of the molecules of the elements (i.e. the amount of heat required to break these molecules up into atoms) is always included. When, for example, chlorine unites with hydrogen to form hydrochloric acid, 22.0 Cal. are given off. That which is measured is the total caloric differ- ence between the initial system H 2 +Cl2 and the 2HC1 formed from it. In the indirect determination of a heat of formation with the help of HESS'S law the calculated caloric effect also includes the heat of decomposition of the molecules of the ele- ments. In the determination of the heat of formation of hvdrogen 156 INORGANIC CHEMISTRY. [100- sulphide, for instance, in the above way the caloric effect of the combustion of this gas is composed of the following parts: 2(2H + S - H 2 S) + 3(2O - O 2 ) = 2SO 2 + 2H 2 O + p Cal. ; that of the combustion of hydrogen of the following: 2(2H - H 2 ) + (20 - O 2 ) = 2H 2 O + q Cal. ; that of the combustion of sulphur of (2S-S 2 )+2(2O-O 2 ) = 2SO 2 +rCal.; (20 2 ), etc., indicating the heat of decomposition of molecules of the elements. The heat of formation of hydrogen sulphide is r+qp. Deduc- ing the value of r+qp from the above equations, we have r+g-p=(2S-S 2 )+2(2H-H 2 )-2(2H-fS-H 2 S), from which it follows that the heats of formation of the sulphur and hydrogen molecules are included in the heat of formation found. CHEMICAL AFFINITY. 101. When a compound is formed, we attribute the phe- nomenon to the affinity which exists between the combining sub- stances. The term "affinity" comes down from an age when it was thought that only those substances could combine with one another which were in a certain agreement with each other (were " in love with each other," as EMPEDOCLES and later also GLAUBER expressed it) . This affinity was originally considered as a force. THOMSEN, for example, defined it as the force which holds the parts of a compound together. Concerning the magnitude of this force our knowledge was for a long time only qualitative. If the substances AB and C interacted to form AC and B } it was said that the affinity of A for C was greater than that of A for B. Comparative study of such reactions led to the arrangement of a series of the elements in decreasing order of affinity; but the absolute, or even relative, magnitude of these affinities was as it were a closed book. Henco it was a great step forward when 101.] THERMOCHEMISTRY. 157 BERTHELOT developed a method of measuring affinity. He con- sidered that the quantity of heat liberated in the formation of a chemical compound was a measure of the affinity satisfied by the action. Thus affinity came to be regarded no longer as a force, but as an amount of work. We know that when water is decom- posed by the current from a dynamo, work must be done in order to drive the dynamo and also to split up the water molecules; and, conversely, when hydrogen and oxygen unite, heat, or in other words energy, is produced. A mixture of hydrogen and oxygen can be compared with a lifted stone ; both possess potential energy. When the stone falls, its potential energy is transformed into kinetic energy. When hydrogen combines with oxygen the potential energy of the system is converted into heat. Since he regarded this heat effect as a measure of the driving force of any chemical reaction, BERTHELOT was led to propose his principe du travail maximum, viz., that of all the chemical processes which can proceed without the application of energy from an outside source that one always occurs which involves the greatest evolu- tion of heat. However, this principle did not prove to be universally applic- able. The very existence of endothermic compounds is at variance with it, for the heat effect of a reaction involving an endothermic compound would be greater if that compound were not formed. Further, the rapidly increasing number of known equilibrium reactions throws doubt on the principle, for, if in an equilibrium A+B<=AB the direct reaction ( ) is exothermic, the opposing reaction (< ) must be endothermic. Yet, even though the principle could not be accepted as a general truth, chemists had to admit that in very many cases it represented the facts, that is, it contained a considerable amount of truth. VAN'T HOFF succeeded in putting things in their proper light. The amount of heat liberated in a chemical reaction represents the total change of the energy of the system, and this is what BERTHELOT regarded as a measure of the affinity. VAN'T HOFF rejected this notion and showed that it is the "free energy" gained in a reaction which must be regarded as a measure of the affinity. By "free energy" we understand the greatest amount of work which the reaction is capable of doing. Now, in order 158 INORGANIC CHEMISTRY. [ 101. to measure the force with which an action tends to proceed, we often make use of an opposing force of known magnitude, which is just great enough to stop the action. If this opposing force is too small, the internal driving force of the system will overpower it and thereby do a certain amount of work, and this amount of work will be the greater, the greater the counter force that is overcome, or in other words, the smaller the difference between this counter force and the driving force of the system. For measuring affinity we can thus make use of the simple mechanical notions which serve for the measurement of forces in general, as, for instance, in an ordinary weighing. We oppose the force to be measured with another of known but variable magnitude and allow the latter to change until equilibrium is established. There is then equality between the known force and the force to be measured. The free energy is in general not equal to the total energy that comes into play in a reaction; but frequently the difference is not great, as, for instance, in reactions between solid com- pounds or in solution. Herein lies the explanation of the excep- tions to BERTHELOT'S principle as well as the reason for its agreement with experiment. The total energy-content of a body consists, according to HELM- HOLTZ, of free and bound energy. The free energy alone is capable of transformation into other forms of work. The bound energy is involved in such changes as those of state. When ice melts a con- siderable amount of heat is absorbed which cannot be transformed into work, but only seems to increase the molecular movements of the water molecules. The bound energy of water is therefore greater than that of ice at the same temperature. Similarly, there are various other processes where the bound energy is changed. It can be proved theoretically that in every action proceeding of its own accord the free energy must decrease. In the case of an exothermic reaction the evolution of heat is due in. part to the decrease of the free energy of the system. Further, the bound energy can at the same time either be partly converted into heat, remain unchanged, or increase less than the decrease of free energy calls for; however, if the decrease of free energy in a reac- tion is less than the increase of bound energy, the whole caloric effect must be negative, which is to say, that the reaction is endo- thermic. 101.] THERMOCHEMISTRY. 159 For the measurement of affinity it is therefore necessary to determine this maximum work or free energy which is involved in chemical reactions. Two means are available, one the deter- mination of the electromotive force that can be created by it, and, secondly, the determination of the equilibrium constant of the reaction in question. We shall learn in the chapter on electrochemistry that reactions can in many cases be conducted so as to produce an electric current. If the reaction is reversible, it can be brought to a stop by sending a current of the same energy through the system in the opposite direction. The energy of an electric current is represented by the product of two factors, the amount of electricity (expressed in coulombs) and the electromotive force (expressed in volts). Now the decomposition of an equiv- alent amount of each compound requires, according to FARADAY'S law, the same amount of electricity, namely, 96,540 coulombs per equivalent weight; whence it follows that the electromotive force must be proportional to the affinity ; in other words, that the electromotive force is a measure of the affinity. Accordingly, the affinity which seeks to bring about a chemical transformation must be opposed by an electromotive force just great enough to prevent the reaction. This electromotive force is then the exact measure of the affinity whose action it prevents. The free energy or maximum amount of work which the reaction produces is accordingly equal to the energy of the electric current produced. The second general method for measuring affinity is applicable in all cases involving a chemical equilibrium. We learn from thermodynamics that the equilibrium constant K and the maxi- mum amount of work A done by the reaction, bear the following relation to each other: A = RTlog e K, when unit concentrations of the reacting substances are in- volved. R is the gas constant ( 35) and T the absolute temperature. K is also dependent on (i.e., a function of) the temperature. The student will find it interesting to learn from the ap- propriate text-books of physical chemistry how these two 160 INORGANIC CHEMISTRY. [ 102- methods are utilized for the ] calculation of affinity in a variety of special cases. THE DISPLACEMENT OF EQUILIBRIUM. 1 02. When two systems are in equilibrium with each other (e.g., 2H2 + O2<= 2H 2 O), the position of this equilibrium is depend- ent on various circumstances. The relationship is expressed by the rule of LE CHATELIER: When any system is in physical or chemical equilibrium, a change in one of its equilibrium factors produces a change in the system, whose effect is opposite to that of the former change. This rule, or theorem, which can be called the principle of the resistance of the reaction to the action, furnishes us with a con- venient means of foretelling in many instances the direction which a reaction will follow. Some examples may be given to illustrate the rule. (1) When a system of water and ice is subjected to increased pressure, the ice melts; that is, that process goes on which involves a contraction, for by this contraction the system diminishes the pressure exerted on it. (2) Monoclinic sulphur, when compressed near the transition point (the temperature of equilibrium for ordinary pressure) , passes over into rhombic sulphur, since this process involves a lessening of volume, and in the end also a diminution of pressure, as in the previous case. (3) When a solution is diluted the osmotic pressure decreases according to BOYLE'S law; in the case of a solution of an electro- lyte dilution will be followed by further dissociation, since this increases the osmotic pressure. (4) When a liquid is heated, more vapor is formed; since the vaporization absorbs heat its effect is opposite to that of the heating. (5) In partially dissociated N 2 4 an increase of pressure drives back the dissociation, while diminution of pressure increases the dissociation. The former change carries with it a pressure decrease, the latter a pressure increase. 103. VAN'T HOFF'S principle of mobile equilibrium is a special case of LE CHATELIER'S rule, but was derived from thermodynamics 104.] THERMOCHEMISTRY. 161 independently. It says: An equilibrium between two different states of matter (systems) displaces itself under constant pressure by a J r- of temperature to that one of the two systems whose formation rise evolves T _ T r heai. A few examples will serve to make this clear absorbs (1) Rhombic sulphur becomes monoclinic when heated above the transition point, since heat is absorbed by this transition. Below this temperature the inverse transition takes place. (Ordinary pressure is assumed in each case.) The reaction works in opposition to the temperature change produced from without. (2) A salt whose heat of solution is negative (saltpetre) dis- solves to a greater degree if the temperature rises. If its heat of solution is positive, a rise of temperature causes a separation from solution. ( 235). This principle leads to a further very remarkable deduction. Since an elevation of the temperature requires the displacement of the equilibrium in the direction of that system which is formed with absorption of heat, endothermic reactions must predominate at high temperatures. On the other hand, exothermic reactions must be generally associated with low temperatures. From the mathe- matical formulation of the principle it follows that at the absolute zero all reactions must be exothermic. At the prevailing room tem- perature, which is not so very far approximately 300 above the absolute zero, most reactions are still exothermic, although endothermic reactions do occur (formation of C1 2 O, etc.). At the temperature of the electric arc (2000-2500-) exothermic compounds are mostly incapable of existence, endothermic com- pounds being obtained. We saw above that ozone and hydrogen peroxide, both endothermic, are formed at a very high tempera- ture ( 36 and 38); and we shall later see that the endothermic compounds nitric oxide ( 120 and 127) and acetylene ( 180) can be synthesized in the heat of the electric arc. PASSIVE RESISTANCES.. 104. A reaction can only proceed of itself in case it yields free energy. It could no more do otherwise than a stone could fly up into the air. Still we observe that certain reactions which 162 INORGANIC CHEMISTRY. [ 104- would undoubtedly liberate free energy do not occur. Accord- ingly, we have to assume that circumstances can arise to prevent the occurrence of a strictly possible reaction. Such hindering conditions can be comprehensively termed passive resistances. Their effect is noticed, for instance, in the retarding of reaction velocity, especially at low temperatures. We saw an example of this in 12 in a mixture of hydrogen and oxygen. Moreover PICTET has shown that sodium, which reacts rather vigorously with alcohol at ordinary temperatures, floats on it quietly at 80 without any apparent reaction; even concentrated hydro- chloric acid and marble do not react upon each other, or at least only very slowly, when they are cooled to a low tempera- ture. In general, the reaction velocity lessens as the temperature falls. It is to this circumstance in many cases that we have to attribute the non-occurrence of reactions which are thermodynam- ically possible. It has been found that the variation of the reaction velocity with the temperature may in general be expressed thus: when the temperature increases arithmetically, the velocity increases geo- metrically. Experience has also shown that a temperature rise of ten degrees in the neighborhood of room temperature generally involves about a doubling or trebling of the reaction velocity. It is not difficult to see what an exceedingly important role the passive resistances above referred to play in nature. Were it not for them, the phenomenon of combustion and the oxidation of metals, etc., could take place at ordinary temperatures; every- thing combustible would then burn and there could be no animal and vegetable life on the earth. NITROGEN. 105. This element occurs free in the air, which contains about 80% nitrogen and 20% oxygen. In combination, it is found in the salts of nitric acid, e.g. saltpetre, and also in the albuminoids, which form an important constitutent of animal and vegetable organisms. Nitrogen can be easily isolated from the air by removing the oxygen. This is accomplished in various ways. Phosphorus, when burned in the air, absorbs the oxygen to form phosphorus 106.] NITROGEN. 163 pentoxide, and the residual gas, aside from slight admixtures ( 110), is nitrogen. Again, air can be passed over heated copper in a finely divided condition, whereupon copper oxide is formed and nitrogen left. In this process the oxygen of the air soon converts all the copper into copper oxide, so that of course only a limited amount of nitrogen can thus be obtained with the aid of a given amount of copper. However, it the air is first passed through ammonia water, the process can be carried on continuously, since the hydrogen of the ammonia, NH 3 , constantly reduces the oxidized copper. Copper can also absorb the oxygen of the air at ordinary tem- peratures, if it is treated with a solution of ammonia and am- monium carbonate. Moist phosphorus combines with oxygen even at ordinary temperatures, so that a volume of air which remains in contact with pieces of phosphorus for some minutes loses its oxygen. An alkaline solution of pyrogallol also has the ability to absorb oxygen at ordinary temperatures. These reactions are made use of in gas analysis. 1 06. Pure nitrogen is obtained by the direct decomposition of certain of its compounds, especially by heating ammonium nitrite. NH 4 NO 2 = N 2 + 2H 2 O. This is usually accomplished by boiling a solution of equal parts by weight of potassium nitrite, KN0 2 , sal ammoniac, NH 4 C1, and potassium dichromate, K 2 Cr 2 7 , in 3 parts of water. The NH 4 C1 and KNO 2 react to form KC1 and NH 4 N0 2 . By heating ammonium chromate, (NH 4 ) 2 Cr0 4 (a mixture of ammonium chloride and potassium dichromate is more convenient), nitrogen is also set free: K 2 Cr 2 7 + 2NH 4 C1 = N 2 + Cr 2 3 + 2KC1 + 4H 2 O. An example of the formation of nitrogen by the indirect decom- position of its compounds is the reduction of nitrogen oxides by hot copper: 2NO-f2Cu=N 2 +2CuO. 164 INORGANIC CHEMISTRY. [106- Physical Properties. Nitrogen is a colorless and tasteless gas. Its specific gravity based on air is 0.9682, its density compared with hydrogen is therefore 13. X 93. 1 1. N weighs 1.2521 g. at and 760 mm. It is one of the most difficult gases to condense, its critical temperature being 146. Its boiling-point is 194. At 214 it becomes solid. It is only slightly soluble in water, even less so than oxygen. Chemical Properties. Nitrogen is chemically very indifferent; it unites with no element at ordinary temperatures and at higher temperatures with only a few. Boron, silicon, titanium, barium, strontium, calcium, magnesium, chromium and also certain rare elements combine directly with nitrogen at red heat, forming nitrides. The direct union of nitrogen with a large number of elements, notably metals, is best accomplished by a special experimental arrangement, as follows: In a liquid mixture of 90% argon and 10% nitrogen an electric arc is produced between a silver anode and the element concerned. The low temperature of the bath prevents the decomposition of the nitrides formed in the arc. Nitrogen unites with oxygen under the influence of induc- tion sparks directly (reddish brown N0 2 being formed); with hydrogen it combines in a similar way. When a mixture of hydrogen and nitrogen, together with a few drops of concen- trated hydrochloric acid, is introduced into a tube over mercury and induction sparks are sent through, clouds of ammonium chloride, NH 4 C1, are produced, the nitrogen and hydrogen having united to form ammonia, NH 3 . These last two reactions and the fact that nitrogen is not able to support combustion serve for the identification of nitrogen gas. The molecule of nitrogen consists of two atoms, this having been demonstrated in the same way as for oxygen and other gaseous elements. 107.] THE ATMOSPHERE. 165 THE ATMOSPHERE. 107. The air was regarded as an element up to the end of the eighteenth century. It finally developed from the investigations of PRIESTLEY and LAVOISIER that it is not a simple body. The correct explanation of the phenomena of combustion led to this conclusion. Before LAVOISIER'S time the explanation of the phenomena of com= bustion was just the reverse of the present one. It was then thought that all combustible or oxidizable substances had a common constituent, phlogiston. According to this theory, which was presented by STAHL (1660-1734), the combustion of a body is due to the escape of phlogiston. If this occurs in a violent manner we have the phenomenon of fire. The more inflammable a substance is, the more phlogiston it was supposed to contain. Sulphur, phosphorus, carbon and hydrogen therefore ranked as being very rich in phlogiston. As to the real nature of this phlogiston opinions were decidedly different. At various times experiments were performed with the hope of isolating the substance. For a while it was thought with CAVENDISH that hydrogen was pure phlogiston. The prevailing ideas were as follows: Substances that possess much phlogiston can transfer it to those which have none or very little. The metals, for example, are substances that contain a certain amount of phlogiston, which they give off on being heated in the air; by this process they are changed to calxes (now called oxides) , which contain no phlogis- ton. When one of these calxes is heated with carbon or hydrogen, it absorbs phlogiston from them and is changed back again to the metal. The fact that sulphur, phosphorus or any other inflammable substance soon ceases to burn when it is enclosed in an air-tight space was explained by the supposition that the air has then become so saturated with phlogis- ton that the latter can no longer escape from the burning body. We see from the above that this theory led men to view many phenom- ena from a common standpoint and undoubtedly contributed in no small degree to the advancement of chemistry. So long as the phenomena of burning were regarded in that light, there was no occasion to doubt the elemental nature of air. They believed that bodies lose something when they are burned, while we now know that on the contrary something is taken up from the air. The great mistake of the phlogiston theory was, that it did not regard the increase in weight of the burned body; as soon as LAVOISIER and others drew attention to this most important fact, the phlogiston theory could no longer be upheld. On the first of August, 1774, PRIESTLEY had discovered oxygen, which 166 INORGANIC CHEMISTRY. [ 107- he himself regarded as air devoid of phlogiston (" dephlogisticated air ") ; * LAVOISIER, however, recognized this substance as the essential principle of all burning and oxidation. It now required only a step to reach the conception that air is not an element, but contains another gas in addition to oxygen, and that this gas does not support combustion. The experi- ment by which LAVOISIER demonstrated this has been already described ( 9). By measuring the amount of nitrogen which remained after the absorption of the oxygen by hot mercury he was able to determine fairly accurately the composition of air. 1 08. Constituents of the Atmosphere. Besides oxygen and nitrogen air contains argon and the other elements described in 110, hydrogen, and also variable amounts of water vapor, carbon dioxide (very nearly 0.04% on the average), ammonia, ozone, and perhaps hydrogen peroxide (the last three in extremely small quantities). Incidentally sulphur dioxide and other gases are found in the air (e.g., in the vicinity of volcanoes). The lower strata of air always contain floating dust particles, microbes, etc. Analysis of Air. The proportional amounts of oxygen and nitrogen in carefully dried air, free from carbon dioxide, etc., have been repeatedly determined with all due precaution. According to the method of DUMAS and BOUSSINGAULT this can be done as follows : The tube, ab (Fig. 34) , containing copper turnings is connected with the globe, V t all air having been removed from both. The end of the tube marked 6 is attached to the various pieces of apparatus Cj B and A, which are to remove the carbon dioxide and water vapor from the inflowing air. The globe, V, is first carefully weighed without air. Thereupon the tube is heated by means of a furnace and a slow current of air is allowed to pass through it to the globe by partially opening the stop-cocks u and r, the oxygen being meanwhile asborbed by the hot copper. By subsequently weighing the globe the amount of nitrogen which it contains can be determined and by weighing the tube before and after we can * From the letters and laboratory notes of SCHEELE, published by Baron NORDENSKIOLD (Stockholm, 1892), it is evident that oxygen was known to SCHEELE sooner than to PRIESTLEY; he called it " Feuerluft." However, this discovery did not seem to lead him any nearer than PRIESTLEY to a correct understanding of the phenomena of burning. 109-] THE ATMOSPHERE. 167 find the amount of oxygen. ' In this way the ratio of oxygen to nitrogen in air can be ascertained. FIG. 34. ANALYSIS OF Am. Another method is the eudiometric method. A known volume of air is mixed with a sufficient known volume of pure (electro- lytic) hydrogen. On allowing an electric spark to pass through, the hydrogen and oxygen unite to form water, which is deposited on the sides of the vessel. Inasmuch as 2 vols. hydrogen combine with 1 vol. oxygen, one-third of the volume that disappeared must have been oxygen. 109. These and other methods of investigation have shown that the composition of the air is nearly constant. In all parts of the earth, as well as at the highest altitudes which balloons have reached, it consists of 20.81% oxygen and 79.19% nitrogen by volume; and 23.01% " "" 76.99% " " weight. The observed variations from this ratio amount to hardly 0.1%. Moreover, the composition does not appear to change with time; our present analyses agree with those of DTJMAS and BOUSSINGAULT made in 1841. This result seems surprising at first thought, because oxygen and nitrogen are constantly being removed from the air and again returned to it and it does not necessarily follow, indeed it is rather an improbability, that the losses and gains will exactly balance. 168 INORGANIC CHEMISTRY. [ 109 The oxygen passes through the following cycle: Free oxygen is consumed in all sorts of oxidations of which the mineralization of organic matter is the most important. By the term " minerali- zation " is meant the oxidation of the residues of plants and animals by the oxygen of the air with the aid of bacilli. The carbon of these residues is oxidized to carbon dioxide; the nitro- gen, phosphorus, sulphur and other elements return to the " mineral" state, as nitrates, sulphates, etc. Along with this process there are the other oxygen-consuming processes of the respiration of animals and plants and the burning of fuels, carbon dioxide being formed in all cases. This carbon dioxide is em- ployed by the plants in their process of assimilation, the oxygen in it being again given back to the air. It will therefore depend on the relative magnitude of this process as to whether just as much oxygen gets back into the air as was previously taken up in the formation of carbon dioxide. The oxygen which serves for other oxidations does not necessarily return to the air. Different investigators have attempted to estimate the amount of carbon which annually enters into the cycle of organic life. DUBOIS calculated that every year the plants assimilate 118.5 million million kilograms C0 2 , which is almost TO of the total carbon dioxide in the atmosphere. The amount of C0 2 given off by the entire animal world is estimated at 2.5 million million kilograms, which brings us to the startling result that only about 2% of the existing plant material is engaged in the cycle with the animal life. All the rest of the carbon dioxide required by the plants comes from the process of mineralization. The amount of carbon dioxide produced by the burning of coal, etc., is estimated by CREDNER at 1.3 million million kilograms. Nitrogen passes through a cycle too. Most of the nitrogen that occurs in the form of organic compounds in animal and vege- table tissues remains in the combined state after the death of the organism, either as ammonia or as nitric acid or in other nitro- genous products. During the process of decay the combined nitrogen is partially liberated; in the burning of plant and animal remains all of it is set free. On the other hand, certain plants, the Leguminosce, are able by symbiosis with bacteria to absorb free nitrogen from the air directly. There are also bacteria which, acting alone, can assimilate nitrogen. Moreover, in storms some 109-] THE ATMOSPHERE. . 169 nitrogen combines with oxygen, and again, silent electric discharges, such as must frequently pass between earth and clouds, cause the nitrogen to enter into combination. Here the question again arises whether as much comes back to the air as goes out. From what has been said it is sufficiently clear that it would fos a mere coincidence if exactly as much oxygen should happen to be withdrawn as is given back. Approximate compensation p: : oably takes place, but, even if it should not, the atmosphere is PC vast that its composition would be only slightly affected in the course of centuries. The following calculation will convince one of the soundness of this argument: The normal atmospheric pressure is 760 mm. mercury; this is due to the weight of the air and the moisture in it. Granted that the pressure of the latter averages 10 mm., we have 750 mm. left for the pressure of the air itself; i.e. the weight of the air is equal to that of a layer of mercury 750 mm. thick extending over the entire surface of the earth. This weight can be calculated thus: The volume of the space between two concentric spheres is 4;r.R 2 r, if R is the radius of the inner sphere, and r the thickness of that space. The radius of the earth (R) is, on the average, 6,370,284m.; r is 0.75m.; therefore, taking into consideration the specific gravity of mercury (13.59), we have for the desired weight of mercury or air 5.2 XlO 18 kilograms. Since 1 m.* air at and 760 mm. pressure weighs 1.2932 kg., the above weight corresponds to a volume of air of 4 XlO 18 m. 3 (at and 760 mm.) or f XlO 18 =8 X 10 17 m. 3 of oxygen. In comparison with this the amount of oxygen which is withdrawn from the air in breathing, burning, etc., is very small, as may be seen from the figures on the preceding page for the quantities used by animals and plants. Since, on the other hand, the assimilative process of the plants yields a considerable amount in addition, ths variations in the proportion of oxygen in the air must obviously be imperceptible with our present analytical methods. The air is a mixture. It cannot be a compound of nitro- gen and oxygen for the following reasons: (1) the ratio of nitrogen to oxygen is different than it would be for a compound of the two elements, for in the latter case it would have to correspond to the ratio of the atomic weights or a multiple of the same; (2) by mixing nitrogen and oxygen in x the ratio in which they exist in air a synthetical air is obtained which is in every respect like that around us. (This excludes the possibility of air containing a per- ceptible amount of a compound of the two elements in addition to 170 INORGANIC CHEMISTRY. [ 109- free nitrogen and free oxygen.) (3) The ratio of the solubilities of the oxygen and the nitrogan of the air in liquids is the same as that calculated from the solubilities of the pure gases oxygen and nitrogen, after taking into account their partial pressures. This could not be the case if the air contained a compound of oxygen and nitrogen; (4) when liquid air boils the first part of the distillate is chiefly nitrogen. The liquefaction of air is now carried on in commerce. The methods used by LINDE and by HAMPSON are based on the same principle, namely, cooling the air by expansion. Further details may be found in text-books on physics. Liquid air is very mobile and has a bluish tint. It is usually somewhat cloudy because of suspended particles of ice (congealed atmospheric moisture) and solid carbon dioxide. These may be removed by filtration through filter-paper. It boils at about 190. It is now extensively used in producing, and demonstrat- ing the effects of, very low temperatures. When carbon dioxide, for example, is led into a flask containing liquid air, it falls in the solid form like snow-flakes. In spite of its low temperature liquid air can be poured upon the hand without danger; it does not even feel cold (on account of the LEYDENFROST phenomenon). Liquid air is much richer in oxygen than the gaseous air of the atmosphere, containing about 50%. If a glowing splinter is dipped into the liquid, the wood begins to burn very vigorously 9 producing a violent reaction. It can be preserved for a rather long time in vacuum flasks. By fractional distillation of liquid air practically pure oxygen and nitrogen can be obtained. According to ERDMANN pure nitrogen is obtained in the cooling down of liquid air, whereupon nitrogen crystallizes out. ARGON, HELIUM AND COMPANION ELEMENTS. no. Argon. Despite the fact that air had been already analyzed times without number, it was first discovered in the course of investigations by RAYLEIGH and RAMSAY in 1894 that there are other elements in the air than nitrogen and oxygen. One of these, named argon by its discoverers, is even found to the extent of 0.9% by volume, or 1.2% by weight. It was on account of its extraordinary resemblance to nitrogen that it 110.] ARGON, HELIUM AND COMPANION ELEMENTS. 171 was so long overlooked. The first indication of its presence was the observation that the specific gravity of the nitrogen isolated from the air is somewhat higher than that of the nitrogen pre- pared from ammonium nitrite and other compounds. 1 liter of nitrogen from air weighed 1.2572 g., while the same amount from chemical compounds weighed 1.2521 g. ; in both cases at and 760 mm. There must therefore be another gas heavier than nitrogen, mixed in with the nitrogen of the air. One of the simplest methods for obtaining argon from the air is to heat air with a mixture, of 1 g. magnesium, 0.25 g. sodium and 5 gr. freshly ignited lime. On account of the high temperature free calcium is formed : Mg+CaO = MgO + Ca, and it is in such a finely divided condition that it absorbs oxygen greedily and also nitrogen, so that only argon is left. Argon can also be isolated with the help of calcium carbide. When calcium carbide (better, mixed with 10% calcium chloride) is brought in contact with air at about 800, it absorbs both oxygen and nitrogen: 2CaC 2 + O 2 = 2CaO + 4C ; CaC 2 + N 2 = CaCN 2 + G. This is a suitable method for preparing argon in large quan- tities. After argon had been once discovered it was found elsewhere than in the atmosphere ; some mineral waters contain it in solution, certain rare minerals yield it when heated, etc. Argon is a colorless, odorless gas, having a vapor density of 19.957. It has been condensed to a colorless liquid, that boils at 186.9, by cooling with boiling oxygen and compressing to about 50.6 atmospheres; it solidifies at 189.6. It is somewhat more soluble in water than is nitrogen (0.05780 parts in 1 vol. at and 760 mm. pressure). As to its chemical nature, it is interest- ing that no one has yet succeeded in preparing a compound of argon. It is certain that what is now called argon is neither a mixture nor a compound, but an element. The boiling-point and the 1 melting-point are constant, and the vapor pressure of argon like- 172 INORGANIC CHEMISTRY. [110- wise remains constant during the liquefaction, so long as any gas is present. Moreover, when a certain volume of argon is three- fourths dissolved in water, the undissolved gas shows exactly the same spectrum as the dissolved. All of the above are charac- teristics of a homogeneous substance. The extraordinary stability of the gas in the presence of all sorts of reagents is a strong argument against its being a compound. in. After the discovery of argon RAMSAY and TRAVERS detected four other rare gases in the atmosphere, though their quantity is very small. These are helium, neon, krypton, and xenon. In a spectroscopic inves- tigation ( 265) NORMAN LOCKYER had detected in the atmospheres of the sun and many fixed stars considerable quantities of a gas unknown on the earth ; he named it helium. In 1895 RAMSAY and TRAVERS succeeded, however, in obtaining it in small amounts on heating the rare mineral cleveite. Afterward it was also met with as a companion of argon in certain other, chiefly uraniferous, minerals as well as in mineral springs, for instance, those of Bath; and at last it was also discovered in the air. At ordinary temperatures helium is a colorless gas. It is of all gases the most difficult to condense; yet KAMERLINGH ONNES recently achieved the task. Helium boils at ^4 absolute temperature (269 C.). By quickly evaporating it a temperature of ^ 1 .5 absolute was reached, which is the lowest thus far attained. The critical pressure is ^ 3 atmos- pheres. Dens, at 4.29 abs. =0.122. In water helium is less soluble than argon. For its relation to radium, see 267. Helium and neon (0.00086 wt. % of air) are found in the most volatile part of liquid air. DEWAR proved that helium and neon can be isolated directly from the air by bringing the air in contact with ignited charcoal at 185. The charcoal has the curious property of condensing in its pores all the other gases of the air, and a gaseous residue is here obtained which shows clearly the spectral lines of He and Ne. While helium and neon were found in the most volatile part of the air, krypton and xenon were obtained, on the contrary, from the residue, after a large quantity of liquid air had been allowed to evaporate slowly. Their separation was rendered possible by the fact that krypton still has a rather large vapor tension at the temperature of liquid air, while the vapor tension of xenon is then imperceptible. Both these elements occur only in extremely small amounts in the atmosphere. Krypton makes up 0.028%, xenon 0.005% (by weight) of the air. In the following table some of the data of these elements are given. The elements form a natural group. 111.] ARGON, HELIUM, AND COMPANION ELEMENTS. 173 Helium. Neon. Argon. Krypton. Xenon. Density (O = 16) Atomic weight 1.98 3 99 10.1 20 2 19.94 QQ Qt> 41.45 82 9 65.1 1 *}O 9 Boiling-point at 760 mm. . . 4 abs. 86.9 abs. 121. 9 abs. 163.9abs. These gases have three properties in common which are worthy of mention here. In the first place they display characteristic spectral lines in PLUCKER tubes ( 263), whereby it has been possible to recognize them and to judge of their purity. In the second place, no one of these elements has been found to enter into combination with other elements ; they may therefore be considered nullivalent. In the third place, their molecule consists of only one atom. This fact could not be discovered in the ordinary way, described in 33 and 34, because of the entire absence of compounds for investigation. It has, however, been possible to ascer- tain it from the molecular heat of the gases. This is the amount of heat that must be imparted to a gram molecule of a gas in order to raise its temperature one degree. This quantity of heat differs, according as the gas is under constant pressure or under constant volume. It is greater in the first case because under constant pressure the gas expands on heating and so does work, which evidently involves an expenditure of heat. We saw in 34 that for one gram molecule of a gas the equation PV = 2T is applicable, the 2T expressing in' calories the external work done when a gas under constant pressure P increases its volume by V, or when a gas being generated under the pressure P comes to occupy a volume V. If the temperature is raised one degree we have PF = 2(7 7 + 1); for each gram molecule of gas extra work is therefore done equivalent to 2 calories. The molecular heat at constant pressure is thus 2 cal. more than that at constant volume. From the kinetic theory of gases it can be deduced that the molecular heat of a monatomic gas at constant pressure is 5 cal. At constant volume it must be 2 cal. less, or 3 cal. The ratio of these quantities of heat is therefore 5:3 = 1.66. When the molecules of the gas consist of more than one atom, more heat is absorbed for the same rise of tem- perature, because heat is then used not only for the movement of the molecules, but also for that of the atoms in the molecule. The ratio then becomes 5+m:3+m, if m is the additional heat. The resulting ratio is thus less than 1.66. By determining this ratio (which can be found from the velocity of propagation of sound in the gas by a well-known physical formula) we can ascertain whether the gases are monatomic or poly- atomic, For the gases of this group the ratio was found to be 1.66, proving that their molecules contain only one atom. 174 INORGANIC CHEMISTRY. [112. Compounds of Nitrogen and Hydrogen. ii2. Until recent years only one compound of hydrogen and nitrogen has been known, viz., ammonia, NH 3 . At present, how- ever, we know five: the others being hydrazine, N 2 H 4 , hydrazoic acid, N 3 H, and the compounds of the latter with ammonia and with hydrazine (NH 3 -N 3 H and N 2 H 4 -N 3 H). Of these five compounds, however, ammonia is by far the most important. AMMONIA. The material now used for obtaining ammonia is the " ammonia liquor" of the gas-factories and coke ovens. The gases that are given off in the dry distillation of coal are passed through water, which dissolves the ammonia. In order to obtain a pure ammonia, the ammonia liquor is heated with milk of lime and the expelled ammonia is led into concentrated sulphuric acid. In this way crystallized ammonium sulphate is obtained. It is purified by recrystallization and again distilled with lime to recover the free ammonia. Ammonia can be prepared synthetically by the following methods. The direct synthesis from the elements was given above ( 107). There are also examples of its formation by the direct decomposition of its compounds. Thus we obtain it by heating the ammonia compounds of certain salts, as #CaCl 2 ?/XH 3 and zAgCl ?/NH 3 . A number of organic compounds yield nitrogen in the form of ammonia on heating. Moreover, ammonia results from the action of hydrogen on certain nitrogen compounds, as, for example, when nitric acid, HNO 3 , comes in contact with nascent hydrogen (generated from zinc or iron filings and dilute sulphuric acid), or when a mixture of nitric oxide, NO, with hydrogen is passed over platinum black: 2NO + 5H 2 = 2NH 3 + 2H 2 0. The formation of ammonia by the action of free nitrogen on liydrogen compounds has not been brought about, but the gas can be produced by the interaction of a hydrogen compound with a nitrogen compound. An illustration of this is the decomposition of magnesium nitride by water: Mg 3 N 2 + 3H 2 = 2NH 3 + 3MgO. 112..] AMMONIA. 175 The putrefaction of organic matter (faeces, urine, etc.) evolves ammonia. By the action of electric . sparks on moist air am- monium nitrate is produced. These last two methods of for- mation are responsible for the slight traces of ammonia in the air. For the formation of ammonia from calcium cyanamide, see OEG. CHEM., 266. Physical Properties. Ammonia at ordinary temperatures is a gas with a characteristic odor, that excites one to tears. Its specific gravity is 8.5 (O= 16) or 0.589 (air = 1) ; 1 1. NH 3 at and 760 mm. pressure weighs 0.76193 g. It can be easily liquefied; it boils at 33.7 and becomes solid at 75; it then forms white translucent crystals. It is extremely soluble in water; at and normal pressure 1 vol. H 2 O dissolves 1148 vols., or 0.875 parts by weight, of NH 3 . The specific gravity of the solution of ammonia in water grows smaller as the concentration increases. The evaporation of liquid ammonia involves a considerable depres- sion of temperature. This is the principle of most of the ice- machines now in use. Chemical Properties. The characteristic property of this corn- compound is that it combines with acids directly to form salts: NH 3 + HC1 = NH 4 C1. NH 3 + HNO 3 = NH 4 N0 3 . Ammonium Ammonium chloride. nitrate. 2NH 3 +H 2 SO 4 = (NH 4 ) 2 S0 4 , Ammonium sulphate. In these salts (which are almost all readily soluble in water) the atomic group NH 4 plays the part of a metal; they correspond in every respect to the compounds KC1, KNO 3 , K 2 SO 4 , etc. The group, or radical, NH 4 has been given a particular name; it is called ammonium. More than one attempt has been made to isolate this ammonium, but always in vain. However, when sodium amalgam comes in contact with a concentrated ammo- nium chloride solution, the mercury swells to a soft spongy mass that rapidly decomposes at ordinary temperatures into ammonia and hydrogen and is in all probability, therefore, ammonium amal- gam. If sodium amalgam is allowed to react with ammonium iodide dissolved in liquid ammonia at 39, a hard metallic mass 176 INORGANIC CHEMISTRY. [ 112- is obtained, which swells with rising temperature because of decom- position into mercury, hydrogen (1 vol.) and ammonia (2 vols.): 2NH 4 =2NH 3 + H 2 . The aqueous solution of ammonia reacts strongly basic; so do the moist fumes of ammonia. We must therefore assume that this solution contains a compound NH 4 OH, ammonium hydroxide, and hence also the ions NH 4 and OH in analogy with other soluble bases, e.g. potassium hydroxide, KOH. As a matter of fact, however, the solution of ammonia conducts the electric current much more poorly than a solution of sodium hydroxide of equiva- lent concentration ( 234). Ammonium hydroxide has not yet been isolated. When the solution of it is evaporated, NH 4 OH splits up into NH 3 and H 2 0. Concordant herewith is the well- known fact that ammonia can be entirely expelled from its aqueous solution by boiling. Ammonia does not burn in the air but does in oxygen; in addi- tion to water and nitrogen traces of ammonium nitrite, NH 4 N0 2 , and nitrogen dioxide, N0 2 , are also formed. A mixture of ammonia and oxygen explodes violently when it is ignited. The oxygen conveyed by soil bacteria may also cause the oxidation of ammonia, producing nitric acid. Chlorine takes fire when passed into ammonia, forming nitrogen, N 2 , and hydrochloric acid; the latter then unites with the remaining ammonia to form sal am- moniac, NH 4 C1. The hydrogen of ammonia is replaceable by metals. Magnesium, e.g. burns in ammonia, forming magnesium nitride, Mg 3 N 2 . When ammonia is conducted over hot potassium or sodium, potassium amide, NH 2 K, or sodium amide, NH 2 Na, is formed. These and analogous metal compounds are decomposed by water, yielding ammonia again and also metal oxide or hydroxide. At high temperatures (produced by induction sparks) ammonia splits up almost completely into its elements, the volume being doubled: 2 vols. 1 vol. 3 vols. On the other hand, nitrogen and hydrogen can unite to form ammonia under the influence of induction sparks ( 106). Equilibrium 114.] HYDRAZINE, OR DIAMIDE. 177 is reached when 3% of ammonia is formed in the gas mixture, N 2 + 3H 2 , i.e., 6% of the theoretical yield. This is the reason why ammonia cannot be split up by electric sparks to more than 97%: 2NH 3 ^N 2 +3H 2 . 3 per cent 97 per cent Nevertheless N 2 +3H 2 can be completely converted into 2NH 3 by induction sparks if the gas mixture is brought in contact with an acid, for by this means ammonia is constantly withdrawn from the gaseous system N 2 + 3H 2 <=* 2NH 3 ; the remaining gas mixture will therefore form new NH 3 in order to restore the equilibrium, and so on, until all the nitrogen and hydrogen have combined. Attempts to manufacture ammonia commercially by direct synthesis have at last been successful. Some metal, such as uranium, iron, manga- nese, or molybdenum, is utilized as a catalyzer and the process is carried on at a pressure of 200 atmospheres and a temperature above 500. The yield is approximately proportional to the pressure. Certain foreign sub- stances are found to act as promoters, others as poisons, of the catalysis. 113. Composition of Ammonia. If an aqueous ammonia solu- tion (to which has been added a little sodium chloride to aid conduction) is subjected to electrolysis, nitrogen, and hydrogen are generated in the volume ratio of 1:3; from this it follows that the molecule must contain 3 H-atoms to every 1 N-atom, i.e., the empirical formula is NH 3 . Since the specific gravity of ammonia gas is 8.5 (0=16), the molecular weight is 17, which corresponds to the above formula. HYDRAZINE, OR DIAMIDE, N 2 H 4 . 114. This compound is now manufactured by the process of RASCHIG. He showed that the reaction of sodium hypochlorite and ammonia yields chloramine: NH 3 -i-NaOCl = NH,C1 + NaOH. At a low temperature in a vacuum this chloramine distils in the form of pale yellow, oily drops having the odor of nitrogen chloride. It decom- poses slowly in dilute solutions, faster in concentrated solutions, yielding nitrogen, ammonia, and hydrochloric acid : 3NH 2 C1=N 2 +NH 3 178 INORGANIC CHEMISTRY. [114- If a large excess of ammonia is present, it acts upon the chloramine with the formation of hydrazine hydrochloride : NH 2 C1+NH 3 ==NH 2 -NH 2 -HC1. The addition of certain substances, e.g., a small amount of albumen, increases the yield, giving as much as 80% of the theoretical yield. By fractional distillation of the aqueous solution the hydrate N 2 H 4 -H 2 is obtained, which boils constant at 118.5; it is a liquid at ordinary temperatures and freezes below 40. LOBBY DE BRUYN showed that the molecule of water can be removed by treatment with barium oxide and that the free hydrazine can be obtained in the pure state by distillation under reduced pressure. This substance is liquid at ordinary temperatures, congeals at 1.4 and boils* under ordinary pressure at 113.5. Sp.g. = 1.014 at 15. It unites with water to form the above hydrate with the evolution of heat. Both the free hydrazine and its aqueous solution have a strong reducing action. The former gradually oxidizes in the air, reacts vigorously with the halogens, etc. The aqueous solution precipitates the metals from solutions of salts of copper, mercury, silver, etc., at ordinary tem- peratures. Hydrazine, like ammonia, unites with acids directly to form salts; it can take up either one or two molecules of a monobasic acid, N 2 H 4 HC1 and N 2 H 4 -2HC1 being both known. The aqueous solution of hydra- zine is strongly basic. Its salts are easily soluble in water, excepting the sulphate, N 2 H 4 -H 2 SO 4 , which is rather difficultly so. HYDRAZOIC ACID, N 3 H. 115. This interesting compound, like the preceding one, was first discovered by CURTIUS in the decomposition of an organic compound. It can now be prepared in a good yield by treating hydrazine hydrate with nitrous acid. This is best done by boiling an alcoholic solution of hydrazine hydrate with amyl nitrite and sodium alcoholate, which gives the sodium salt of the hydrazoic acid. An aqueous solution of the free acid is best obtained by distilling lead hydrazoate, Pb(N 3 ) 2 , with dilute sulphuric acid. By fractional distillation of this solution the pure acid can be obtained. Pure hydrazoic acid is a liquid with a penetrating, unbearable odor; it boils at 37 and is extremely explosive, even in aqueous solution. It is a strange fact that hydrazoic acid displays more or less analogy with the hydrogen acids of the halogens; it forms, like them, difficultly soluble salts of silver, mercury (ous) and lead. These are, however, 116.] COMPOUNDS OF NITROGEN WITH THE HALOGENS. 179 soluble in strong mineral acids. They are also very explosive, hence extremely dangerous, the sodium salt being the least so. An aqueous 1% solution of the acid is only 0.008 ionized; it is thus a rather weak acid; it gives off hydrogen in contact with many metals, e.g. Zn, Fe, Cd, and Mg. It is characteristic of the metal hydrazoates (or "azides") that they crystallize anhydrous and yield the pure metal when heated. Compounds of Nitrogen with the Halogens. 116. When chlorine gas is allowed to act on a concentrated solution of ammonium chloride, most conveniently by inverting a flask full of chlorine over the warm (30-40) solution, oily drops are formed, which are best collected in a leaden saucer placed under the mouth of the flask. These drops contain some hydrogen as well as nitrogen and chlorine. By treating with chlorine once more pure nitrogen trichloride, NCls, is obtained as a yellowish oil with a disagreeable pungent odor and a specific gravity of 1.65. This is one of the most dangerous of substances, because it explodes in a most violent manner, not only on contact with certain organic substances (e.g. turpentine), but very often spontaneously. It. dissolves in carbon disulphide, benzene and other solvents, form- ing yellow solutions. These solutions are relatively harmless; they decompose in the sunlight. Concentrated hydrochloric acid decomposes nitrogen trichloride according to the equation: NC1 3 +4HC1=NH 4 C1+3C1 2 ; aqueous .ammonia also breaks it up in a similar way: NC1 3 + 4NH 3 = 3NH 4 C1 + N 2 . Nitrogen trichloride is strongly endo thermic : N+3C1-NC1 3 = -41.9 Gal. When a solution of sodium azide, NaN 3 , is mixed with a solution of sodium hypochlorite in the relation of molecule for molecule and the mixture is acidified, the liquid assumes a yellow color and gives off a colorless gas with an odor like that of hypochlorous acid and having the composition N 3 C1, showing it to be chlorazide. On being passed 180 INORGANIC CHEMISTRY. [ 116- into caustic soda it forms sodium azide and sodium hypochlorite in equivalent amounts : N,C1 +2NaOH = NaN 3 +NaOCl +H 2 O. The chlorazide is likewise extremely explosive. 117. Nitrogen Iodide. If a solution of iodine in potassium iodide is mixed with ammonia solution, a precipitate is usually obtained of the composition NI 2 H; if the conditions are slightly- altered another compound, N 2 I 3 H 3 (i.e. NH 3 + NI 3 ), is deposited which breaks up on continued treatment with water into ammonia and nitrogen tri-iodide. These compounds are likewise very explosive. Another method is to digest pulverized iodine with ammonia water. The product so obtained is still more explosive, often exploding even when damp or when it is being washed with water or by the action of hydrochloric acid. In the presence of ammonia solution it is stable. Nitrogen iodide is decomposed by dilute hydrochloric acid, forming ammonia and chlorine iodide: Nitrogen iodide is also decidedly endothermic. Hydroxylamine, NH 2 OH. 118. Hydroxylamine is a reduction product of many oxygen compounds of nitrogen intermediate to the formation of ammonia; e.g. it is formed when tin acts on dilute nitric acid. Here the nascent hydrogen effects the reduction: HN0 3 + 3H 2 = NH 3 O + 2H 2 0. It is manufactured by the electrolytic reduction of nitric acid dis- solved in sulphuric acid. The free hydroxylamine is best prepared by heating the phosphate. It is a crystallized solid, melting at 30 and boiling under 60 mm. pressure at 70. When heated in the air it explodes with a yellow flame. 119.1 NITROUS OXIDE. 181 Hydroxylamine is easily soluble in water; its solution reacts strongly alkaline. It forms salts in the same way as ammonia, i.e. by direct addition of the acid: NH 2 OH-HC1, NH 2 OH-HN0 3 , etc. These salts are rather stable; the hydrochloride, however, must be preserved over lime, else it slowly decomposes, for the following reason. The salt is split up to a very small degree into hydrochloric acid and hydroxylamine. Now free hydrochloric acid accelerates catalytically the decomposition of the salt. When, however, the hydrochloric acid is absorbed by the lime, the decomposition becomes so slow that it is imperceptible. The free hydroxylamine and its aqueous solution are somewhat unstable, especially in the presence of alkalies; it decomposes easily into ammonia, water and nitrogen. A further characteristic of hydroxylamine is its great reducing power; it precipitates reddish-yellow cuprous oxide from an alkaline copper solution at ordinary temperatures, even when strongly diluted; mercuric chloride, HgCl2, is reduced to calomel, Hg2Cl2; silver nitrate to silver, etc. The following reaction is also peculiar : A solution of ferrous sulphate is precipitated with an excess of sodium hydroxide and warmed; if hydroxylamine (or one of its salts) is now added to the green ferrous hydroxide, red ferric hydroxide is formed very quickly, the hydrox- ylamine being reduced in this alkaline solution to ammonia. On acidify- ing, an acid solution of a ferric salt is obtained; if this is treated with a hydroxylamine salt, it is suddenly decolorized because of reduction to ferrous salt, the hydroxylamine being now in the oxidized condition in the acid solution. Compounds of Nitrogen with Oxygen. Those included under this title are: nitrous oxide, N 2 O; nitric oxide, NO; nitrogen trioxide, or nitrous anhydride, N 2 3 ; nitrogen dioxide, NO 2 , or tetroxide, N 2 04, and nitrogen pentoxide, or nitric anhydride, N 2 Os. NITROUS OXIDE, N 2 O. 119. This compound cannot be obtained directly from its ele- ments; the ordinary method of preparation consists in heating ammonium nitrate to about 250: 182 INORGANIC CHEMISTRY. [ 119- This method is analogous to that of preparing nitrogen from ammonium nitrite ( 105). If the nitrate is heated above 250, the gaseous product partially decomposes. Physical Properties. Nitrous oxide is a colorless and odorless gas ; which when liquefied boils at 87 and solidifies at 102. The evaporation of the liquid produces a great depression in the temperature, which may even reach 140 under reduced pressure. Its specific gravity is 1.52 (based on air), or 21.89 for 0=16. 1 1. N 2 at and 760 mm. pressure weighs 1.9657 g. It is rather soluble in water (1 vol. H 2 dissolves 1.305 vol. N 2 O at 0); hence it must be collected over hot water. In alcohol it is still more soluble. Chemical Properties. Nitrous oxide supports combustion. Phosphorus, carbon and a glowing splinter burn in it as in oxygen. A mixture of nitrous oxide and hydrogen explodes like detonating- gas when it is ignited, only not quite so loud. These properties might lead one to confuse it with oxygen on a superficial examina- tion. However, it is very easily distinguished from the latter by the fact that it gives no red fumes when mixed with nitric oxide ( 120) and always leaves residual gas (nitrogen) after a combustion. A faintly burning piece of sulphur is moreover extinguished by nitrous oxide. Nitrous oxide is endothermic: 2N+O N 2 0= 17.7 Cal. BERTHELOT has made the general observation that endothermic substances can suffer an explosive decomposition; in this case this may be brought about by touching off the gas with fulminating mercury. It is easy to explain BERTHELOT'S observation. When an endothermic substance decomposes, heat is evolved. Now, we saw in 13 and 104 that chemical reactions are accelerated in a very high degree by rise of temperature. Suppose that a sudden decomposition is caused at a certain point in a mass of an endo- thermic compound. The heat given off raises the temperature of the surrounding molecules and they too split up suddenly, evolving still more heat, and so on. The whole mass will thus reach a condition of sudden decomposition, that is, it will explode. To bring this about it is only necessary that the first impulse be vigorous enough for the sudden decomposition of so many mole- cules that the heat evolved is sufficient to raise the surrounding ones to the temperature of decomposition. 120.] NITRIC OXIDE. 183 Composition. Under the protracted action of induction sparks the gas splits up into a mixture of nitrogen and oxygen, the volume of which is half again as great as that of the nitrous oxide. When potassium and sodium are burned in the gas, potassium and sodium oxides respectively are formed, together with nitrogen; the gas volume after cooling is unchanged. Both of these observations point to the same formula, N 2 O, and this is confirmed by the fact that the relative density of the gas, which should theoretically be was found to be 21.89. NITRIC OXIDE, NO. 1 20. This gas is only obtained by the reduction of nitric or nitrous acid. The ordinary method of preparation is by allowing copper to act on nitric acid or else by covering copper (in the form of thin sheets) with a saturated solution of saltpetre and adding concentrated sulphuric acid drop by drop ( 127): 3Cu + 8HN0 3 = 3Cu(NO 3 ) 2 + 4H 2 + 2ND. In this reaction the hydrogen, which would be expected to be given off, reduces another portion of the acid. In order to prepare nitric oxide by the reduction of nitric acid or a nitrate a boiling-hot solution of ferrous chloride, EeCl 2 , in hydrochloric acid is found very satisfactory; the ferrous chloride is converted into the ferric chloride, FeCl 3 , by the reaction: HN0 3 + 3FeCl 2 + 3HC1 = 3FeCl 3 + 2H 2 + NO. Perfectly pure nitric oxide is obtained by treating a mixture of yellow prussiate of potash and potassium nitrite with acetic acid: 2K 4 Fe(CX) 6 + 2KNO a +4C 2 H 4 O 2 =K 6 Fe 2 (CN) 12 +4KC 2 H 3 O 2 +2H 2 O +2NCX Yellow prus- Pot. ni- Acet. acid. Red prus- Pot. acetate, siate. trite Physical Properties. Nitric oxide is a colorless gas, whose specific gravity has been found to be 1.039 (air=l). It can be condensed to a blue liquid, which boils under ordinary pressure at 184 INORGANIC CHEMISTRY. [ 120- 153.6. The critical temperature is 93.5 ; the critical pressure 71.2 atm. It is not very soluble in water, but dissolves easily in a solution of ferrous sulphate, FeSCU; strange to say, this solution is quite dark brown in color, although the ferrous salt solution is pale green and nitric oxide colorless. The compound which is formed here has not been isolated, but it has been shown to consist of FeS0 4 and NO in equimolecular proportions. Chemical Properties. It is characteristic of this gas, above all other properties, that it combines with oxygen immediately, forming nitrogen dioxide, a reddish-brown gas. On heating it with hydrogen no explosion occurs; the mixture burns with a white flame, forming water and nitrogen. If burning phosphorus is introduced into the gas, it continues to burn; a lighted candle is, however, extinguished ; sulphur and charcoal do not burn in it either. A mixture of nitric oxide and carbon disulphide burns with an intensely luminous blue flame, that is very rich in chemically effective rays. Nitric oxide is a strongly endothermic compound; it can be made to explode by fulminating mercury ( 119). According to 103 NO must be formed at a high temperature. NERNST proved that the reaction N 2 + 02<=^2NO accords strictly with the law of mass-action ( 49); from whichever side one starts, the results are in agreement with those calculated, assuming both reactions to be bimolecular. See further 127. The formation and the decomposition of NO are much slower than in the case of ozone. Accordingly a short heating of air, followed by a rapid cooling, produces ozone, while a slower heating and cooling yield NO, ozone being broken up during the extended period of cooling. The following experiment illustrates this: When moist air is directed with a velocity of less than 7 m. per sec. against an incandescent NERNST filament ( 291) NO is formed; a more rapid current gives ozone. Composition. When sodium is heated in contact with a measured amount of nitric oxide, sodium oxide and nitrogen are formed ; the latter takes up exactly half the volume of the original gas. The specific gravity of nitric oxide is 15 (H = l), hence its molecular weight is 30. According to the above decomposition the gas con- tains one atom of nitrogen (14 parts by weight). There remain for the oxygen, therefore, 16 parts by weight, i.e. just one atom. Hence the formula is NO. 122.] NITROGEN DIOXIDE AND TETROXIDE. 185 Since nitrogen is trivalent or quinquivalent (the latter in ammonium salts, e.g. NH 4 C1) and oxygen is bivalent, it must be assumed that there is a free valence bond in NO, i.e. -N=O. The same applies to N0 2 r Free bonds like these are very rare. Nitrous Anhydride, N 2 0g. 121. Upon allowing nitric acid of 1.3 specific gravity to react with arsenic trioxide and drying very carefully the gas that comes off and finally condensing this gas, a liquid is obtained of the composition N 2 3 . At ordinary temperatures the liquid is green but below 2 it is deep indigo-blue. When cooled by liquid air it solidifies in dark blue crystals . 2 x 14. +3 v 16 The compound N 2 3 should have the vapor density - ^ t - = 38. In a perfectly dry state its vapor density was found to vary in a series of experiments between 38.1 and 62.2, so that the compound appears in that condition to be partly polymerized. The least trace of moisture, however, causes a dissociation into NO 2 and NO. For instance, it was found to be sufficient merely to leave a small bulb full of it with a capillary tube open a few seconds in the air; upon resealing the tube the vapor density was found to have fallen to 28.2. This very striking property of water, whereby even the slightest trace of it brings about dissociations which are not observed in the perfectly dry state, will be met with in several examples in later chapters. The phenomenon was discovered by BRERETON BAKER. NITROGEN DIOXIDE AND TETROXIDE, N0 2 AND N^0 4 . 122. Nitrogen dioxide is formed from nitric oxide plus oxygen, or more convenient!^ by heating well-dried lead nitrate: When so prepared it is a very deep-brown gas. On leading it into a strongly cooled vessel it condenses to a bright-yellow liquid, which solidifies at 20 to colorless crystals, that melt at 12. The color becomes darker on warming and at +26 the liquid begins to boil, changing back again into the brown gas. The vapor density of this gas at 26 is found to be 38.0, while that cal- culated for N 2 O 4 is 45.9 and that for NO 2 22.9 (H = l). ^ Since the value found is between the two, "it may be assumed that at this temperature the vapor consists partly of N 2 O 4 molecules and partly of NO 2 molecules. A simple calculation indicates the percentage 186 INORGANIC CHEMISTRY. [ 122- of the former to be 34.4%. As the temperature rises, the vapor density steadily decreases till about 150 is reached, when it becomes constant at 22.9. There is evidently complete dissociation of molecules in this case, 1 vol. 2 vols. and, inasmuch as the color of the gas grows darker, we must sup- pose that NO 2 is dark brown, while N 2 C>4 is colorless, which is true of the latter in the solid state. This supposition is supported by the fact that not only can the degree of dissociation be estimated from the intensity of the color, but that it can even be measured quantitatively in this way. According to 51 the equilibrium between the two gases is expressed by the equation P-x=kx 2 , where P is the total pressure of the gas mixture and x that of the dioxide, k being a constant. From this equation it follows that the dissociation (at a constant temperature) depends on the pres- sure (51), which has been shown to be the case. This also fol- lows from the theorem of LE CHATELIER ( 102). On bringing nitrogen tetroxide in contact with water or, better, with alkalies, nitrous and nitric acids are formed; we may therefore consider it as a mixed anhydride of these two acids : Nitric acid. Nitrous acid. Both N0 2 and N 2 4 possess strong oxidizing power; many substances burn in their vapor; they precipitate iodine from solu- ble iodides. The composition of nitrogen dioxide follows from its synthesis- equation, 2NCM-02, and from the vapor density. Nitrogen Pentoxide, N 2 5 . 123. This compound can be obtained by the action of chlorine on silver nitrate or by distilling fuming nitric acid with phosphorus pent- oxide. It is a colorless crystalline solid. It melts at 30, and at 45-50 breaks up, giving off brown fumes. If the heating takes place rather 125.] HYPONITROUS ACID. 187 rapidly the decomposition is explosive in nature; sometimes a spon- taneous explosion takes place, hence it can not be kept long. As nitrogen pentoxide is strongly endothermic, its spontaneous explosion must be explained in the same way as is indicated in 119. Only we must conclude in this case that the decomposition at ordinary temperatures is vigorous enough to sufficiently heat the neighboring molecules. It unites with water, forming nitric acid with the evolution of much heat. As might be expected, it has strongly oxidizing properties. Phos- phorus and potassium, for instance, burn with great brilliance in the slightly warmed anhydride. The composition of nitrogen pentoxide is ascertained by heating with powdered copper; the amount of nitrogen evolved corresponds to the formula N 2 O 5 Oxygen Acids of Nitrogen. 124. Four acids of nitrogen are known: hyponitrous acid, H 2 N 2 O 2 ; nitrohi/droxi(laminic acid, H 2 N 2 03; nitrous acid, HNO 2 ; nitric acid, HNO 3 . The nitrous acid is known only in dilute aqueous solution; mtrohydroxylaminie acid is known only in its salts; but the others are known in the pure state. Only certain ones of the above nitrogen oxides can be regarded as acid anhydrides. The pentoxide is undoubtedly one and the tetroxide may be considered as a mixed anhydride of nitric and nitrous acids ( 122). Nitrogen trioxide gives a solution of nitrous acid when mixed with water at a low temperature; however, this solution undergoes a decomposition slowly at ordinary, more rapidly at higher, temperatures ; nitric acid and nitric oxide being formed: 3HN0 2 =HN0 3 +2NO + H 2 O. The acid corresponding to nitric oxide, NO, is nitrohydroxyl- aminic acid. However no one has yet been able to obtain this acid from nitric oxide and water. The same is true for nitrous oxide, to which hyponitrous acid corresponds. Hyponitrous Acid, H 2 N 2 2 . 125. This acid is formed when nitrogen trioxide is introduced into a methyl-alcoholic solution of hydroxylamine. The free acid does not liberate iodine from potassium iodide at once; the reaction is delayed 188 INORGANIC CHEMISTRY. [ 125- for a time, probably on account of a decomposition, by which nitrous acid is formed. Hyponitrous acid belongs to the class of weak acids; its aqueous solution is a poor conductor. Both neutral and acid salts of this acid are known. Nitrohydroxylaminic Acid, H 2 N 2 8 . This acid does not exist in the free state, being known only in salts. Its sodium salt is obtained by mixing an alcoholic solution containing sodium alcoholate and hydroxylamine with ethyl nitrate: C 2 H 5 ONO 2 + NH 2 OH = C 2 H 5 OH + H 2 N 2 O 3 . Ethyl Nitrate The alcoholate is added in order to convert the free acid directly into its sodium salt. If the attempt is made to liberate it by adding a stronger acid, it is immediately decomposed according to the equation: NaNA + 2HC1 = 2NaCl + 2NO + H 2 0. The sodium salt, heated in aqueous solution, gives sodium nitrate -and nitrous oxide. When the sodium salt is heated dry until it begins to melt, it is decomposed into nitrite and hyponitrite: 2Na 2 N 2 3 = 2NaNO 2 + Na 2 N 2 2 . NITROUS ACID, HN0 2 . 126. It was remarked above that this acid is only known in dilute solution at ordinary or low temperatures; its salts are, however, stable. In order to prepare them we usually employ potassium or sodium nitrate, which gives off oxygen when heated and is converted into nitrite. This decomposition takes place more readily if lead is added during the heating as a reducing agent: 2KN0 3 = 2KN0 2 + 2 . 127.J NITRIC ACID. ] 8 g Its salts are all easily soluble in water, "with the exception of silver nitrite, AgN0 2 , which is rather difficultly soluble at ordinary temperatures ; it is obtained as a yellow crystalline precipitate, when not too dilute solutions of silver nitrate are mixed with a nitrite. The addition of strong sulphuric acid to a nitrite at once pro- duces red fumes; in this way a nitrite can be distinguished from a nitrate, for the latter does not produce them. It may be assumed that in this reaction free nitrous acid is primarily formed; this is, however, broken up directly into water and nitrogen trioxide, the latter of which at once splits up again into NO 2 + NO; thereupon the nitric oxide unites immediately with the surrounding oxygen to form dioxide. The red fumes thus consist solely of nitrogen dioxide, N0 2 . On treating a very dilute nitrite solution with the equivalent amount of sulphuric acid a dilute solution of free nitrous acid is obtained. This solution can act either oxidizing or reducing. As examples of the former action we have the liberation of iodine from a solution of potassium iodide, the oxidation of sulphurous acid in dilute solution to sulphuric acid, the oxidation of ferrous sul- phate, FeS0 4 , to ferric sulphate, Fe 2 (SO4)3, and the conversion of the. yellow to the red prussiate of potash. In all of these cases lower oxides of nitrogen, chiefly nitric oxide, are formed. An example of its reducing action (in which nitrous acid is oxidized to nitric acid) is the bleaching of potassium permanganate, KMn0 4 , in sul- phuric acid solution: 2KMnO 4 + 5HNO 2 + 3H 2 S0 4 = K 2 SO 4 + 2MnS0 4 + 5HNO 3 + 3H 2 0. This last reaction offers a means of determining quantitatively (volumetrically, see 93) the strength of a dilute solution of nitrous acid. NITRIC ACID, HN0 3 . 127. This is the best known acid of nitrogen. It is manufac- tured on a large scale, since its uses are many and varied; in the organic dyestuff industry, for example, large quantities are employed. The commercial process of manufacture depends on the decom- position of Chili saltpetre, NaN0 3 , by strong sulphuric acid: NaN0 3 + H 2 S0 4 = NaHS0 4 + HN0 3 One of the simplest methods of carrying it out is as follows: 190 INORGANIC CHEMISTRY. 127- In the cast-iron .retort (C, Fig. 35), saltpetre and sulphuric acid (chamber-acid) are mixed in proportions corresponding to the above equation, a slight excess of sulphuric acid, however, being added, because this makes the residue easier to remove from the retort, The retort is connected with a row of earthenware bottles (EE f ) containing a little water. These receive the distilled acid. The last bottle connects with a coke tower through which water is FIG. 35. MANUFACTURE OF NITRIC ACID. trickling down to dissolve the uncondensed acid vapor. By this process a liquid of a specific gravity of 1.35 and containing 60% acid is obtained. If the saltpetre is previously dried and concen- trated sulphuric acid is used, a nitric acid of sp. g. 1.52 and almost 100% pure can be obtained. In some cases two molecules of saltpetre are used to one of sulphuric acid. If heat is moderately applied, the reaction pro- ceeds according to the above equation, but on heating to a higher temperature the acid sodium sulphate that is formed acts on the second molecule of nitrate, also forming nitric acid: NaN0 3 4- NaHS0 4 = Na 2 S0 4 + HN0 3 . A large part of the nitric acid, however, dissociates at the same time as follows: 2HN0 3 =2N0 2 -fH 2 + 0. The N0 2 -fumes dissolve in the distillate. The liquid thus obtained is red and its specific gravity is 1.52-1.54; it fumes strongly in the air and is known as "red fuming nitric acid." 127.] NITRIC ACID. 191 For some years the distillation of saltpetre with sulphuric acid has been carried on ; n. a vacuum. The yield of acid in such a case approaches closely to the theoretical and the product obtained is entirely free from nitrous fumes. An entirely distinct method for the industrial preparation of nitric acid was Lay:nted a few years ago by BIRKELAND and EYDE. They make use of the nitrogen and oxygen of the atmos- phere. The problem of making nitric acid from this rather in- exhaustible source has been studied for many years, but these men are the first to handle it with success on a commercial scale. The solution of the problem became a really pressing matter, because the principal material Tor the preparation of nitrogen compounds, Chili saltpetre, bids fair to be exhausted in thirty- five years, and saltpetre has not only a large significance in the industrial world but a still larger one in agriculture as a nitro- genous fertilizer.^ The method of BIRKELAND and EYDE is based on the long established fact that oxides of nitrogen are formed in an electric arc burning in the air. The reason why previous investiga- tions did not succeed lies in the fact that an ordinary electric arc has too small a volume, and therefore, cannot let a sufficient quantity of air pass. This dif- culty is now obviated by mount- ing the arc between the poles PP of a very powerful electric magnet EE (Fig. 36). The arc is produced between two hol- low bars of copper, which are kept cool by circulating water in them. When an alternate current is used for producing the arc, the latter spreads out FIG. 36. DIAGRAM OF BIRKELAND AND in the shape of a flat disc that EYDE NITRIC ACID APPARATUS. reaches a diameter of 2 m. in the industrial form of the apparatus; the tension employed is 5000 volts. This flame disc is inclosed in a box through 192 INORGANIC CHEMISTRY. [ 127. which a rapid current of air is forced, and the contact with the flame is sufficient to form somewhat more than 1% of NO. Instead of broadening out the electric arc to a sun-shaped disc by the action of powerful magnets, SCHONHERR (Badische Anilin- und Sodafabrik) forms an arc in the inside of an iron pipe through which air is passed. Under these circumstances the arc is developed in a peculiar manner. When the current is turned on, the arc forms at the first instant in the lower part of the metal pipe, between the pipe itself, which serves as an electrode, and a second electrode, which is separated by only a few millimeters from the lower end of the pipe. Forthwith, however, the arc is carried along upward in the pipe by the current of air, which is given a tangential motion as it is passed into the pipe, so that the arc comes to occupy the portion of air along the axis of the pipe and does not touch the wall of the pipe (or the efflux end of the pipe or a specially devised separate electrode) until a considerable distance from the lower electrode is reached. Thus there is established in the axis of the pipe a continuous and quietly burning column of light of very powerful actinic effect. In this long-drawn-out arc the passing air is partially transformed into nitric oxide. This is quickly chilled by contact with the wall of the pipe, which is- externally exposed to the atmosphere, and so prevented from redecomposition. The gaseous product is half again, if not twice, as rich in nitric oxide (yield about 2%) as by the BIRKE- LAND-EYDE process. The NO must be looked upon as the primary product, which subsequently unites with oxygen to form NO 2 , the latter being carried to water-absorption towers much like the GAY-LUSSAC towers in sulphuric acid plants. The N0 2 cannot be the primary product, for it dissociates at about 600 into NO and 2 . With the water the NO 2 forms nitric acid and nitrous acid : N 2 4 + H 2 = HNO 3 + HN0 2 . The latter yields NO 2 and NO, however, when the liquid becomes more concentrated: 2HNO 2 = H 2 + NO 2 + NO. 127.] NITRIC ACID. 193 NO is once more converted into NC>2 and the N02 again gives nitric acid; eventually all is converted into that acid. Instead of marketing sodium nitrate, to duplicate the Chili saltpetre, calcium nitrate is produced by saturating the nitric acid with lime and the resulting calcium nitrate is used for fertilizing and other purposes. Nitrites are also manufactured directly by leading N2O 4 into caustic : N 2 O 4 + 2KOH = KN0 3 + KN0 2 + H 2 O. The nitrate and nitrite are separable by fractional crystallization. Physical Properties. Absolute nitric acid, i.e. the compound HNO 3 in the pure state, is prepared by distilling the nearly pure acid of commerce (sp. g. 1.5) with concentrated sulphuric acid in vacuo. The liquid distillate has a specific gravity of 1.559 at and becomes solid at 40; it boils under ordinary pressure at 86, but with partial decomposition. Chemical Properties. Nitric acid, especially when pure, is a rather unstable compound; at ordinary temperatures it is decom- posed by sunlight to a slight extent, turning yellow on account of the small amount of nitrogen dioxide formed. At an elevated temperature the acid also breaks up, decomposition into nitrogen dioxide, water, and oxygen being complete at 260. When strong nitric acid is subjected to repeated distillation under atmospheric pressure, its boiling-point gradually rises, while the acid becomes proportionately weaker, until finally a 68% acid is obtained, which boils constant at 120.5. The same mixture is obtained when one starts with dilute acid and distils it. In both cases the boiling- point of the original liquid is lower than that of the product; it rises during the boiling to a maximum at 120.5. We have here, therefore the case of a liquid mixture with a maximum boiling-point, which is discussed in ORG. CHEM., 22. The mixture of hydrogen chloride and water also has a maximum boiling-point (110). Nitric acid is very extensively ionized in aqueous solution; it is one of the strongest acids known. When it comes in contact with metals, the salts of nitric acid (nitrates) are formed, but without any evolution of hydrogen, since part of the acid present is reduced by the nascent hydrogen. The nitrates are all easily soluble in water. The action of nitric 194 INORGANIC CHEMISTRY. [ 127- acid on the metals is not the same in all cases. It does not attack gold or platinum. Silver, mercury, and copper are only imper- ceptibly dissolved at ordinary temperatures, but on warming they dissolve with the evolution of nitric oxide. This and the other NO-compounds are powerful catalyzers in the dissolving of the above-named metals, for nitric acid which is perfecty free from them does not dissolve these metals, while the reaction immediately begins as soon as a little of these substances is added: It may be supposed that on warming nitric acid traces of NO-compounds are formed, which together with the elevation of the temperature accelerate the reaction. Iron, zinc, and magnesium reduce nitric acid to nitrous oxide and even to ammonia. Under the action of iron filings and dilute sulphuric acid the reduction of nitric acid to ammonia in dilute solution is quantitative. There are also various denitrifying bacteria known, Bacillus pyocyaneus being the best studied of them. Nitric acid frequently acts as a powerful oxidizing agent, especially at an elevated temperature If sulphur is boiled with it, the sulphur is converted to sulphuric acid, similarly phosphorus to phosphoric acid. A glowing piece of charcoal dropped upon the concentrated acid continues to burn with a bright glow. In all these cases the highest oxidation stages are formed. Nitric acid is used particularly in the organic branches of chemical industry. The composition of nitric acid can be deduced from that of its anhydride. A weighed amount of the latter is introduced into water; nitric acid is formed, which is neutralized with baryta water. By evaporation it is possible to determine how many parts by weight of barium oxide, BaO, combine with the anhydride. It is found that 153.37 parts (=!BaO) combine with 108.02 parts (=1N 2 O 5 ) of the anhydride; the formula of barium nitrate thus becomes Ba(NO 3 ) 2 , hence that of nitric acid itself imjist be HN0 3 . Pernitric Acid, HN0 4 . Pernitric acid is formed in very dilute aqueous solution by the oxidation of nitrous acid with hydrogen peroxide : 2H 2 O 2 + HN0 2 = HN0 4 + 2H 2 O. Nitric acid does not yield it when treated in the same way; on the contrary, the pernitric acid breaks up even in a cold dilute aqueous 128.] DERIVATIVES OF THE NITROGEN ACIDS. 195 solution inside of about an hour completely into nitric acid and hydrogen peroxide : HNO 4 + H 2 - HNO 3 + H 2 2 . Pernitric acid has the very characteristic property of liberating bromine from potassium bromide solutions, something that neither hydrogen peroxide nor nitrous acid nor nitric acid does. Derivatives of the Nitrogen Acids. 128. In discussing the manufacture of sulphuric acid ( 86) we already referred to the chamber crystals, HSO 5 N. They are formed in the lead chambers in case not enough steam is supplied. The following equation expresses the action that takes place: 2S0 2 + N 2 O 4 + + H 2 O = 2SO 5 NH. The ordinary method of preparing this substance is by conduct- ing carefully dried sulphurous oxide into cooled fuming nitric acid: S0 2 + HNO 3 =SO 5 NH. The crystalline mass obtained is spread out on porous earthenware to allow the adhering liquid to be absorbed. The chamber crystals have the appearance of a coarse crystal- line, colorless mass; they melt at 73. They are at once decom- posed by water into sulphuric and nitrous acids: SO 5 NH + H 2 O = H 2 SO 4 + HN0 2 . For this reason the compound is considered as the mixed anhy- dride of sulphuric and nitrous acids. According to 90 the struc- OTT ture S0 2 xidation of phosphine by various substances, for instance, nitrogen dioxide, by which ordinary hydrogen phosphide can be made spon- taneously inflammable. The mixture of PH 3 and P 2 H 4 can be separated by passing it through a well-cooled tube; the latter substance con- denses to- a colorless liquid, which boils at 57-58 (under 735 mm.) and has a specific gravity of 1.01. It is easily decomposable and can- not be preserved, because it rapidly changes to the gaseous and the solid hydrogen phosphides. The same decomposition is also effected by hydrochloric acid. It must be condensed in the dark, as sunlight aids decomposition. The empirical composition is indicated by the formula PH 2 ; but since phosphorus is trivalent, we take P 2 H 4 , i.e. H 2 P-PH 2 , as the formula of the molecule. Liquid hydrogen phosphide thus becomes analogous to hydrazine. Solid Hydrogen Phosphide, P 12 H 6 . 139. This substance is formed by the decomposition of the preceding one, especially easily when phosphine charged with P 2 H 4 vapor, as it is evolved from the decomposition of Ca 3 P 2 with water, is led over granu- lated calciumchloride. The solid hydrogen phosphide separates out as a bright yellow powder of the empirical formula P 2 H, whose molecular weight has been determined crysocopically to be P 12 H 6 (using yellow phosphorus as a solvent). On being heated to about 200 it breaks up into phos- phine and a new solid hydrogen phosphide of the empirical formula P 9 H 2 . This latter compound is also formed when P 12 H 6 is treated with liquid ammonia. When heated in the air P 12 H 8 catches fire at 160. It is insoluble in water. 140.] PHOSPHORUS TRICHLORIDE. 209 Halogen Compounds of Phosphorus. Phosphorus unites with all four of the halogens to form com- pounds of the types PXs and PX 5 ; the most important are the chlorides. PHOSPHORUS TRICHLORIDE, PC1 3 . 140. This compound can only be obtained by direct combina- tion of the elements. In preparing it a rapid current of dry chlorine is led over phosphorus in a retort. The phosphorus burns with a pale yellow fiame and a mixture of trichloride and penta- chloride distils over into the receiver, which is kept cold. A little phosphorus is added to the distillate in order to convert the pen- tachloride to trichloride, and the liquid is redistilled. An easier method is to introduce phosphorus into a flask with some phos- phorus trichloride and lead chlorine into the mixture. Physical Properties. Phosphorus trichloride is a colorless liquid of a very pungent odor; it boils at 76 and remains liquid as low as -115. Sp. g. = 1.6129 at 0. Chemical Properties. Water decomposes it very rapidly with the formation of hydrochloric and phosphorous acids: PC1 3 + 3H 2 = H a PO 3 + 3HC1. It is because of this decomposition that it fumes in moist air. Continued treatment with chlorine converts it into the penta- chloride. 210 INORGANIC CHEMISTRY. [ 141- PHOSPHORUS PENTACHLORIDE, PC1 6 . 141. This substance is prepared by passing chlorine over phos* phorus trichloride. Fine light-yellow crystals at once appear and the entire mass finally becomes solid, indicating that all is con- verted into penta chloride. This compound fumes strongly in moist air, being immediately decomposed by watei with the forma- tion of hydrochloric and phosphoric acids. When heated it sub- limes without melting. In the transition to the gaseous state it breaks up at a rather low temperature into the trichloride and chlorine; this dissociation is complete at 300, for at that point the vapor density is just half of that calculated for the penta- chloride. The vapor, which at moderately low temperatures is almost colorless, takes on the yellow color of chlorine for the above reason, as the temperature rises. The dissociation products, phosphorus trichloride and chlorine can be separated by diffusion. Phosphorus pentachloride evaporates in an atmosphere of the tri- chloride with almost no dissociation (cf. 51). By the addition of a little water it is converted into phosphorus oxy chloride: PC1 5 +H 2 = POC1 3 + 2HC1. With more water phosphoric and hydrochloric acids are produced. Phosphorus pentachloride is used in organic chemistry to replace hydroxyl groups with chlorine. In inorganic chemistry, it can also be employed for the same purpose; thus sulphuric acid reacts with it in the following manner ( 89) : PHOSPHORUS OXYCHLORIDE, POC1 3 . 142. The best method of preparing this compound is by the oxidation of the trichloride with sodium chlorate: 3PC1 3 + NaClO 8 = 3POC1 3 + NaCl. In order to moderate the great vigor of this reaction sodium chlorate 144.] COMPOUNDS OF PHOSPHOROUS. 211 is placed under phosphorus oxy chloride and the trichloride is then added slowly by means of a dropping-funnel. Phosphorus oxychloride is a colorless, mobile liquid, that boils at 107.2 and, when solid, melts at -1.5. Sp. g. =1.7118 at 0. In the presence of water, with which it is not miscible, it slowly changes to phosphoric and hydrochloric acids: POC1 3 + 3H 2 = H 3 P0 4 + 3HC1. THE COMPOUNDS OF PHOSPHORUS WITH THE OTHER HALOGENS. 143. These are very analogous to the chlorine derivatives. They are likewise prepared by direct synthesis from the elements. Inasmuch as the reaction is very vigorous, it has to be moderated by dissolving the phosphorus and the halogen separately in carbon disulphide, slowly adding one to the other and then distilling off the solvent. The fluorides have special methods of preparation. All these compounds are broken up by water like the corresponding chlorides, the fluorides, however, quite slowly. The composition of these compounds can be ascertained in the following way: on being decomposed by water they yield phos- phoric or phosphorous acid and a halogen acid, so that the quan- tities of phosphorus and halogen present can be found by deter- mining the amounts of these acids. Moreover, the molecular weight can be obtained by measuring the vapor density, though it must be borne in mind, however, that compounds of the type PX 5 are usually dissociated in the gaseous state. Oxygen Compounds of Phosphorus. 144. Three compounds of this class are known: phosphorus trioxide, P20s; phosphorus tetroxide, P204,' and phosphorus pent' oxide, or phosphoric anhydride, P2O 5 . Only the last is of any great importance. Phosphorus Trioxide, P 2 3 . This compound is produced when phosphorus burns in a slow cur- rent of dry air in a tube. The principal product is phosphorus pent- oxide, which can be collected by a wad of glass fibers. The phosphorus trioxide passes through as a vapor and is condensed in a well-cooled 212 INORGANIC CHEMISTRY. [144- tube. It is a white waxy substance when thus formed, but it can also be obtained in crystals; the latter melt at 22.5 and boil at 173.1 (in a nitrogen atmosphere). The vapor density has been found to be 109.7, while that calculated for P 4 6 is 110. On being heated to 440 it is decomposed into red phosphorus and phosphorus tetroxide. It turns yellow in the light, which explains the fact that *Pfosphorus pentoxide sometimes takes on a yellow color. It dissolves slowly in cold water forming phosphorous acid; with hot water it produces red phosphorus, self-inflammable hydrogen phosphide and phosphoric acid in a vigor- ous reaction. When heated to 50-60 in the air it takes fire and burns to the pentoxide. Phosphorus Tetroxide, P 2 4 , is obtained from the P 2 3 compound, as was stated above. It forms colorless glistening crystals, that break up in water into phosphorous and phosphoric acids. In this respect its conduct is analogous to that of nitrogen tetroxide, which yields nitrous and nitric acids with water. PHOSPHORUS PENTOXIDE, P 2 5 . This compound is the product of the combustion of phosphorus in oxygen or an excess of dry air. It forms a white, voluminous, snow-like mass, that takes up water rapidly to produce phosphoric acid. It is the most powerful desiccating-agent known. MORLEY ascertained that it dries the air down to 1 mg. water vapor in 40,000 1. air. It exists in two modifications, both of which are formed simultaneously in the above process. The one is crystal- line, subliming at 250; the other amorphous and not volatile below red heat; the vapor condenses crystalline. When heated above 250 the crystalline modification passes over into the amorphous form. Heating with charcoal reduces it to phosphorus. The vapor density of phosphoric anhydride at bright redness was found to correspond to the formula (P 2 O 5 ) 2 . Acids of Phosphorus. 145. Only two of the above described oxides of phosphorus, viz. P 2 O 3 and P 2 5 , form corresponding acids; these oxides can unite with different amounts of water to form acids. From P 2 Os we have: 145.] ACIDS OF PHOSPHOROUS. 213 P 2 0s + H 2 O = 2HP03, metaphosphoric acid, H4P 2 O7, pyro phosphoric acid, and = 2H 3 PO4, orthophosphoric acid. From the other^xide two acids can be derived: metaphos- phorous acid, HPO^phd phosphorous acid, H 3 PO 3 . Besides these there are two acids of phosphorus, whose anhydrides are unknown. viz. hypophosphorous acid, H 3 PO 2; and hypophosphoric acid, H 4 P 2 O 6 . The relation between ortho-, meta-, and pyrophosphoric acids can be shown in another way, which leads us to make some general obser- vations. It was remarked in 141 that phosphorus pentachloride is transformed by water into phosphoric and hydrochloric acids. The action of water on the pentachloride may be regarded as .consisting first of a substitution of all five chlorine atoms by hydroxyl: P|C1 5 + 5H|OH =5HC1 + P(OH) 5 . This compound, which would strictly be regarded as orthophos- phoric acid, is unknown; a molecule of water is at once split off, form- ing the ordinary phosphoric acid, H 3 PO 4 , which we are accustomed to call orthophosphoric acid. In a similar way the metaphosphoric acid can be derived from the acid P(OH) 5 by the splitting off of two molecules of water: C)H IQH OH IOH P O H - OP^OH; P OJH -> O 2 P-OH; OH ^OH |OH Metaph 9 sphoric O H Orthophos- O H phoric acid. while the pyrophosphoric acid can be regarded as 2P(OH) 5 -3H 2 0: 0|H H]<3 IOH H_0| .OIL V OH P O H H O P -> OP^-OH OP OH OH HO \o/ IOH HIO J - ! Pyrophosphoric acid. Orthophosphoric acid can also be derived from phosphorus oxy- chloride : OP(OH) 3 . 214 INORGANIC CHEMISTRY. [ 145- This way of looking at them makes plain not only the connection between the different acids, but also their structural formulae. The same method can be applied to many other cases. As an example we may select the per-iodic acids. In 62 only one was mentioned. There are salts, however, of various per-iodic acjdg, e.g. MIO 4 , M 3 IO 5 , M 5 I0 6 , etc. These can be derived from a hypothetical acid I(OH) 7 in which iodine is joined to as many hydroxyls as correspond to its maximum valence. M 5 IO 6 would come from I(OH) 7 1H 2 O; M 3 IO 5 from I(OH) 7 -2H 2 O; and MIO d from I(OH) 7 -3H 2 O. ORTHOPHOSPHORIC ACID, H 3 P0 4 . 146. This acid can be obtained by direct synthesis from its elements; phosphorus burns to the pentoxide and the latter yields the acid on dissolving in water. Its formation by the action of nitric acid on phosphorus was mentioned in 134. It can also be obtained by the oxidation of compounds containing phosphorus and hydrogen; phosphine and the lower acids of phosphorus are oxidized to phosphoric acid. Ordinarily this acid is prepared by the oxidation of phosphorus with nitric acid or by liberating it from its salts, particularly the calcium salt, Ca 3 (PO4)2- The latter is stirred into the theoretical amount of dilute sulphuric acid, forming calcium sulphate, which is only slightly soluble in water, and phosphoric acid, which goes into solution. On evaporating this solution the acid remains. At ordinary temperatures orthophosphoric acid is a crystalline solid. It melts at 38.6, is odorless and extremely soluble in water, forming a strongly acid solution. It has the character of a strong acid; however, it is consider- ably less ionized than hydrochloric acid; a solution of 1 mole phosphoric acid in 10 1. water contains about one-fourth as many hydrogen ions as hydrochloric acid of the same molecular con- centration. It is ionized chiefly into H' and H^PO/. It generates hydrogen with metals, all three hydrogen atoms being replaceable by metallic atoms; it is therefore tribasic. Three classes of salts are possible and known to exist; these are the primary, secondary and tertiary salts. Of the alkali salts all three kinds are soluble; of the alkaline earth salts only the primary, the tertiary and secondary being insoluble. The other phosphates are insoluble in water but are dissolved by mineral acids. 146.] ORTHOPHOSPHORIC ACID. 215 This latter property is due to the fact that phosphoric acid is a weaker acid than the strong mineral acids, hydrochloric, nitric and sulphuric. On treating an insoluble phosphate with one of these acids, e.g. 'hydrochloric, undissociated molecules of phos- phoric acid are formed in the liquid; the more hydrochloric acid, the more the association, since the hydrochloric acid reduces the ionization of phosphoric acid. H 2 PO 4 ' and H' ions thus disappear and, in case enough hydrochloric acid is added, the concentration of the H 2 P0 4 ' ions remaining will not be great enough together with that of the metal ions present to reach the value of the solu- bility product; hence all the phosphate must dissolve ( 73). For the same reason, as a general rule, salts that are insoluble in water will only dissolve in acids that are stronger than the acid of the salt. The only exception to this is the case when the value of the solubility product of the insoluble salt is very small, examples. of which we have seen in certain sulphides ( 73). When heated to 213 orthophosphoric acid gives off water, forming mainly the pyro-acid but also a little meta-acid through- out the reaction. The pyro-acid on the other hand is converted by further heating into the meta-acid. With silver nitrate orthophosphates give a yellow precipi- tate of silver phosphate, Ag 3 PO 4 , soluble in nitric acid and ammo- nia. In the case of a primary or secondary phosphate, the pre- cipitation is not complete, since nitric acid is liberated in the reaction: Na 2 HP0 4 +3AgN0 3 = AggPO* +2NaNO 3 +HN0 3 , or, expressed in ions: HP0 4 " + 3Ag- < Ag 3 P0 4 +H'. If, however, an excess of sodium acetate is added, the precipi- tation is practically complete. The reason for this is obvious. By the addition of acetate the acetic anions C2H 3 02' are forced to combine with the H' ions, for acetic acid is only very slightly ionized and its ionization is, more- over, considerably lessened by the excess of sodium acetate. The result is that in the equilibrium HPO 4 " + 3Ag' <= Ag 3 P0 4 +H* the H' ions are removed. The inverse reaction < is then no longer 216 INORGANIC CHEMISTRY. [ 146- possible, and the direct reaction > must therefore become com- plete, or in other words, all the phosphoric acid is precipitated as silver phosphate. It was stated above that the alkali salts of phosphoric acid are soluble in water. These aqueous solutions differ markedly in reaction. The solution of a primary salt, KH 2 P0 4 , is acid, that of a secondary salt feebly alkaline, and that of a tertiary salt strongly alkaline. The cause of this variation must be more fully explained. The acid reaction of a salt such as KH 2 PO4 must be attributed to the fact that its anion, H 2 PO 4 ' (analogous to the anion HSO 4 '), is capable of splitting up into the ions H' and HPO 4 ", the former producing the acid reaction. The feebly alkaline reaction of a salt like K 2 HPO 4 is accounted for by hydrolysis ( 66). Such a salt is extensively ionized in dilute solution into 2K' and HPO 4 ". However, while H 3 P04 is rather highly dissociated (into H* and H 2 PO 4 '), H 2 PO 4 ' is but slightly ionized into H* and HPO 4 ". In this case H 2 P0 4 ' behaves as a weak acid. Hence, if there is a large proportion of HPO 4 " ions in a solution, they will tend to unite with H' ions, because the system H* + HPO 4 ":^H 2 PO 4 ' is only in equilibrium when the right-hand side preponderates. The necessary H' ions are supplied by the water, which is split up to a very slight extent into H' and OH'. But when the H* ions unite with HP0 4 " ions we have a surplus of OH' ions in the solution and the latter takes on an alkaline reaction. Entirely analogous is the explanation of the strongly alkaline reaction of the tertiary phosphates, such as K 3 PO 4 . Their aqueous solutions contain the ions PO 4 '", which have a still stronger tendency to unite with H' ions than the HPO 4 " ions. The P0 4 '" ion, there- fore, causes the presence of an even larger proportion of OH' ions, not compensated by H' ions, so that the result is a strongly alkaline reaction. Phosphoric acid is precipitated from an ammoniacal solution by a magnesium salt as white crystalline ammonium magnesium phosphate, NH 4 MgPO 4 +6H 2 0. Another very characteristic test for phosphoric acid is that in nitric acid solution a finely crystal- line, yellow precipitate is produced by ammonium molybdate, especially on warming. This precipitate has approximately the composition 14Mo0 3 +(NH 4 ) 3 P0 4 +4H 2 0, i.e. it is an ammo- 148.] PYROPHOSPHORIC ACID. 217 mum phospho-molybdate. Precipitation in acid solution is of great advantage here ; since most of the phosphates are soluble only in acids. PYROPHOSPHORIC ACID, H 4 P 2 7 . 147. One method of producing this acid was given in the pre- ceding paragraph. In preparing it, it is more practicable, however, to heat the secondary sodium phosphate (the ordinary sodium phosphate of commerce), because in this case only one molecule of water can be driven off from two molecules of the salt: 2Na 2 HPO 4 = H 2 O + Na 4 P 2 O 7 . After being heated the sodium pyrophosphate is dissolved in water and lead acetate is added to precipitate lead pyrophosphate, which is then decomposed with hydrogen sulphide. Pyrophosphoric acid can be obtained from its solution as a colorless vitreous mass by .evaporation in a vacuum at a low tem- perature. When dissolved in water of ordinary temperature, the acid remains unchanged for quite a while ; on warming this solu- tion, especially after the addition of a little mineral acid, it is converted in a few hours into ortho-acid ( 145). All four hydrogen atoms are replaceable by metals; we should therefore expect to find four classes of salts. In reality only two are known, M 4 P 2 O7 and M 2 H 2 P 2 O7. The neutral, as well as the acid, salts of the alkalies are soluble in water; the neutral salts of other bases are insoluble, the acid salts chiefly soluble. Pyrophosphoric acid is distinguished from the ortho-acid by the fact that solutions of its salts give a white precipitate, Ag 4 P 2 O7, with silver nitrate, and from the meta-acid by not coagulating albumen and giving no precipitate with barium chloride. METAPHOSPHORIC ACID, HP0 3 . 148. This acid is obtained by heating the ortho- or the pyro- acid till no more water passes off, or by heating ammonium phos- phate (NH 4 ) 2 HP0 4 . Moreover, on dissolving phosphorus pent- oxide in cold water, the product is at first chiefly meta-acid. At ordinary temperatures metaphosphoric acid is a vitreous solid (hence the name glacial phosphoric acid), which can be melted and easily drawn out into threads. On being heated strongly 218 INORGANIC CHEMISTRY. [ 148- it volatilizes without breaking up into water and pentoxide. When boiled in aqueous solution it goes over into orthophosphoric acid. It is very deliquescent; use is made of this property occasionally. Metaphosphoric acid is monobasic, corresponding to the formula HPOs. Its alkali salts only are soluble in water. In solution the meta-acid can be distinguished from the ortho- and the pyro- acids by its ability to coagulate albumen and give white precipi- tates with chlorides of barium or calcium. The vapor of this substance at bright-red heat consists chiefly of H 2 P2Oe molecules (di-metaphosphoric acid), which are apparently liable to undergo partial dissociation and even to lose a small quantity of water. There are salts of various acids known, which must be regarded as polymers of metaphosphoric acid, e.g. K 2 P 2 6 , potassium di-metaphos- phate; there exist also tri-, tetra-, and hexa-metaphosphates, i.e. salts of the acids H 3 P 3 O 9 , H 4 P 4 O 12 , and H 6 P 6 18 . Hypophosphoric Acid, H 4 P 2 O 6 . 149. When sticks of phosphorus are suspended in a solution of sodium acetate in such a way that only 0.5 cm. is exposed above the level of the liquid and the temperature is kept between 6 and 8, the phosphorus oxidizes slowly and the difficultly soluble acid sodium salt of hypophosphoric acid, Na 2 H 2 P 2 O 6 +6H 2 O soon begins to crystallize out. It can be purified by crystallization from a dilute solution of acetic acid. If this salt is dissolved in water and barium chloride added, a precipitate of barium hypophos- phate is formed, from which an aqueous solution of the free acid can be obtained by means of dilute sulphuric acid. This can be evaporated at 30 to a sirupy consistency without decomposition and, when left in a vacuum, yields crystals of the acid. At an elevated temperature and in the presence of a mineral acid phosphorous and phosphoric acids are formed. This behavior justifies the consideration of hypophos- phoric acid as a mixed anhydride of the two last-named acids: OH HO OH OH OPOH HOP -> OPOH POH |OH__H]0 \ o / Phosphoric Phosphorous Hypophosphoric acid. acid. acid. 151.] PHOSPHOROUS ACID. 219 However, it has not yet been possible to prepare the hypophopphoric acid by melting together the other two acids. From the determination of the molecular weight of the methyl ester it seems probable that the formula of the acid is H 2 PO 3 and not H 4 P 2 6 Metaphosphorous Acid, HP0 2 . 150. This compound was discovered by VAN DER STADT during the slow oxidation of phosphine under reduced pressure ( 137) : PH 3 + O 2 =H 2 + HPO 2 . The sides of the vessel become covered with feather-like crystals of HP0 2 . These melt at a much higher temperature than the crystals of phos- phorous acid and are converted into the latter by the action of water vapor. PHOSPHOROUS ACID, H 3 P0 3 . 151. In 149 it was mentioned that this acid is formed by the slow oxidation of phosphorus in moist air. It is more easily pre- pared by decomposing phosphorus trichloride with water: PC1 3 + 3H 2 = H 3 P0 3 + 3HC1. The hydrochloric acid can be expelled by evaporating at 180 and the phosphorous acid crystallizes out on cooling. The melting-point of phosphorous acid is 70.1. It is a very hygroscopic substance Heating decomposes it into phosphoric acid and phosphine. It has a strong reducing action, being itself oxidized to phosphoric acid. The oxygen of the air acts on it very slowly. It precipitates the metals from solutions of gold chloride, mercuric chloride, silver nitrate, etc. A characteristic reaction is the reduction of sulphur dioxide to sulphur, which takes place at ordinary temperatures, when solutions of the two substances are mixed. In spite of its three hydrogen atoms, phosphorous acid acts as a dibasic acid. As we have already observed, the ionization of poly basic acids sometimes affects only one H* ion at first, the others being split off with increasing difficulty. According to OSTWALD it may be supposed that ionization beyond 2H' and HPO 3 " is in this case so difficult that the acid seems to be only dibasic. The phosphites are not oxidized by the air, but they yield to the action of 220 INORGANIC CHEMISTRY. [152- oxidizing-agents; e.g. they liberate the precious metals from their salts, as does also the acid itself. Heating breaks them up into hydrogen, pyrophosphates and phosphide. The double phosphites give precipitates with baryta- or lime-water. Hypophosphorous Acid, H 3 P0 2 . 152. Salts of this acid are produced by heating phosphorus with caustic soda, lime-water or baryta-water ( 136): 3Ba(OH) 2 + 8P + 6H 2 O = 3Ba(H 2 PO 2 ) 2 + 2 It can be set free from these salts by sulphuric acid; the aqueous solution is concentrated at 80-90 and then cooled strongly, whereupon the acid crystallizes out. Melting-point, 26.5. On being heated at 130-140 the acid splits up into phosphorous acid and phosphine; at a somewhat higher temperature the latter acid yields phosphine and phosphoric acid. The equations are: 3H 3 PO 2 = 2H 3 PO 3 + PH 3 ; 3H 3 PO 3 = 2H 3 PO 4 + PH 3 . Hypophosphorous acid is a very strong reducing-agent. Gold, silver and mercury are precipitated from solutions of their salts by the free acid as well as its salts. Sulphur dioxide is reduced to sulphur at ordinary tem- peratures. In these reactions the acid itself is, converted into phosphoric acid. It is distinguished from phosphorous acid by its behavior towards copper sulphate solution; when it is warmed with the latter, a red precipi- tate of copper hydride, Cu 2 H 2 , is formed. Hypophosphorous acid is mono- basic. Compounds of Phosphorus and Sulphur. 153. Various compounds of this sort are known; all of them are obtained by warming the two elements together. As the reaction is very vigorous with yellow phosphorus, the red form is usually employed. The compound P 2 S 5 , which is of service in organic chemistry, is a yellow crystalline substance, melting at 274-276 and boiling at 518. On being warmed with water it yields phosphoric acid and sulphuretted hydrogen. P 2 S 5 unites with 3 molecules of K 2 S to form a sulphophosphate, K 3 PS 4 , i.e. a phosphate whose oxygen is replaced by sulphur. Several compounds containing a halogen in addition to phosphorus and sulphur are known, e.g. PSC1 3 . This phosphorus sulphochloride can be pre- pared by treating phosphorus pentachloride with hydrogen sulphide, a method 155.] COMPOUNDS OF PHOSPHORUS AND NITROGEN. 221 analogous to that of forming the oxychloride from the pentachloride and water. A more convenient method is by the action of the pentachloride on the pentasulphide, which carries out the analogy, to oxy-compounds still farther ( 142) : 3PC1 5 +P 2 S 5 =5PSC1 3 . It is a colorless liquid, boiling at 125. Water decomposes it into phosphoric acid, hydrochloric acid and hydrogen sulphide. Compounds containing Phosphorus and Nitrogen. 154. The compounds of this class are also numerous. Among them are amidophosphoric acid, ^Pi > an d diamidophosphoric acid, OTT OP />^jj -, . As their names indicate, these compounds behave like acids. If dry ammonia is conducted over phosphorus pentachloride, a white mass is obtained which consists supposedly of ammonium chloride, NH 4 C1, and a compound PC1 3 (NH 2 ) 2 . With water it forms phosphamide, PO(NH)(NH 2 ), a white insoluble powder. On being boiled with water secondary ammonium phosphate is formed: PO(NH)(NH 2 ) The name phospham is given to a compound P 3 H 3 N 6 , which is formed from the product of the action of ammonia on phosphorus pentachloride, when it is heated in the absence of air till no more ammonium chloride fumes appear. It is insoluble in water. When fused with potassium hydroxide, it breaks up as follows: P 3 H 3 N 6 + 9KOH + 3H 2 = 3K 3 P0 4 + 6NH 3 . By the interaction of P 2 S 5 and NH 3 it is easy to obtain a com- pound P 3 N 5 , phosphorus nitride. ARSENIC. 155. Arsenic occurs in nature in the free state native. More frequently it is found in combination with sulphur (realgar, As2S2, and orpiment, As2S c ) and with metals (arsenopyrite, or mispickel, Fe AsS, and cobaltiie, Co AsS) ; also with oxygen as As2O 3 (arsenolite) . The extraction of the element from these minerals is simple 222 INORGANIC CHEMISTRY. [ 155- Arsenopyrite yields arsenic on mere heating, the latter subliming. Arsenolite is reduced with carbon: 2As 2 3 +6C = As + 6CO. Physical Properties and Allotropic Conditions. The condition in which arsenic usually occurs is the crystalline. It then has a steel-gray color and a specific gravity of 5.727 at 14 and is a good conductor of electricity. It sublimes under ordinary pressure without melting; under increased pressure, however, it melts at 500. By sublimation in a current of Hydrogen a second crystal- lized form can be obtained together with a black modification, which according to RETGERS is also crystallized. An amorphous modification results from the decomposition of hydrogen arsenide by heat, the arsenic appearing as a dark brown deposit on the sides of the glass. Finally there is a yellow modification which is formed when arsenic vapor is condensed in a dark room by liquid air. This yellow arsenic is very sensitive to light; even at the tem- perature of liquid air ( 180) at which it is stable in the dark it is converted into the black modification by the light of a Welsbach burner. It is a remarkable fact that a solution of the yellow modification in chlorine is much more stable toward light and heat than is the pure substance. Such solutions are obtained in concentrations up to 7%; when they are cooled yellow arsenic crystallizes out. The relation between yellow and black arsenic is very analogous to that between yellow and red phosphorus, except that in the case of arsenic the yellow form is much less stable. At an elevated temperature (360) all the modifications pass over into the ordinary crystalline form. Vapor Density. The lemon-yellow vapor of arsenic has a density of 10.2 (air=l) at about 860, which makes the molecular weight 293.8. At 1600-1700 the vapor density is less by half, being 5.40. Since the atomic weight of arsenic is 75, its molecule therefore contains four atoms at about 860 and two at 1600-1700. Chemical Properties. Arsenic is not affected by dry air at ordinary temperatures; in moist air it becomes covered with a coating of oxide. At 180 it burns with a bluish flame to the oxide As 4 Oe, giving off a peculiar garlic-like odor. At an elevated 156.] HYDROGEN ARSENIDE. 223 temperature it combines with many elements directly; it unites with chlorine without the aid of heat ; producing scintillations. HYDROGEN ARSENIDE. ARSINE, AsH 3 . 156. Direct synthesis from the elements is not possible with this compound. It is formed when almost any arsenic compound comes in contact with nascent hydrogen (zinc + sulphuric acid). When thus prepared it contains considerable hydrogen, however. Pure arsine is obtained by treating zinc arsenide or sodium arsenide with dilute sulphuric acid : As 2 Zn 3 + 3H 2 S0 4 = 2AsH 3 + 3ZnSO 4 . Physical Properties. Hydrogen arsenide is a gas; it liquefies at 40, but does not solidify as low as 110. Sp. g. = 38.9 (H= 1). It must be handled with great care, as it is very poison- ous. Fortunately its presence can be easily detected by its peculiar, disagreeable odor. Chemical Properties. Arsine can be decomposed into its ele- ments by heat. If the gas is passed through a hot glass tube, arsenic is deposited on the sides in the form of a metallic mirror. Induction sparks also decompose it. By the latter means it can be shown that the resulting volume of hydrogen is 1J times as large as that of the gas itself, in accord with the formula AsH 3 . It is an endothermic compound, As + 3H-AsH 3 = -36.7 Cal., and has been made to explode by fulminating mercury ( 419). Hydrogen arsenide burns with a pale flame, yielding water and arsenious oxide, As2O 3 , if sufficient air is present; if such is not the case, or if the flame is cooled, arsenic is deposited. On heating potassium or sodium in the gas, an arsenide, AsK 3 or AsNa 3 , is formed. Hydrogen arsenide precipitates the yellow compound AsAg 3 3AgNO 3 from a very concentrated solution of silver nitrate : AsH 3 + 6AgN0 3 = As Ag 3 3 AgN0 3 . This is decomposed by the addition of water into arsenious acid, nitric acid and metallic silver, the latter being deposited. 224 INORGANIC CHEMISTRY. [ 156- This reaction is called GUTZEIT'S test. It is usually carried out in the following way:. A drop of 50% AgNO 3 solution is placed on a piece of filter paper and the moist spot is held over a test-tube containing some zinc, dilute sulphuric acid and the substance to be tested for arsenic. A plug of cotton is inserted near the top to protect the paper from being spattered by the effervescing solution. If arsenic is present, the spot becomes yellow, and turns black when moistened with water. Composition of Arsine. If arsine is passed over hot copper oxide, water and copper arsenide are formed. The ratio of hydrogen to arsenic in arsine is determined from this reaction. For 1 part (by weight) of hydrogen 24.97 parts of arsenic are obtained. The molecular weight of the compound, as found from the specific gravity (see above), is 77.9; since the atomic weight of arsenic is 75, the formula of arsine must be AsH3. Detection of Arsenic. 157. The majority of arsenic compounds are very poisonous. Several of them are of practical use and hence are on the market, e.g. white arsenic, ^gA (rat-poison); orpiment, As 2 S 3 ; Schweinfurt green, or copper arsemle. Poisonings with these substances happen occasion- ally. Some arsenic compounds, because of their pretty green color, are still used, though much less than formerly, in dyeing tapestries, portieres, and the like. Rooms in which these are hung usually con- tain particles of arsenical matter, which are injurious to the health. Further, a certain species of mould, penicilliwn hrevicaule, which is sometimes found in such tapestries, has the power of generating volatile and very poisonous arsenic compounds. The chemist is therefore quite frequently called upon to analyze a given sample (of dyed materials or the like, or the contents of a stomach) for arsenic. For this pur- pose a method has been devised which enables him to detect with cer- tainty extremely small amounts of arsenic. It involves the following operations: The organic substance in question is at first disintegrated as well as possible, usually by digestion with hydrochloric acid on the water bath, a little potassium chlorate being added from time to time. Thus the arsenic compound is oxidized to arsenic acid. When the chlorine has been expelled by warming and the liquid has been filtered, hydrogen sulphide is passed in for some time at a temperature of about 80 to precipitate the arsenic as sulphide. The sulphide is then dis- solved in nitric acid (in case the presence of antimony is suspected it must first be removed); this solution is evaporated to dryness to get rid of the excess of acid, the dry residue is dissolved in water, and this 157.] DETECTION OF ARSENIC. 225 liquid is then tested in the MARSH apparatus, a simple form of which is shown in Fig. 38. This consists of a small flask, in which hydrogen is generated from zinc and sulphuric acid; the liquid to be investigated is poured down the thistle-tube; if arsenic is present, arsine is formed. The mixture of hydrogen and arsine is dried by calcium chloride in the wide tube and then enters a tube of hard glass, which is narrowed at several places and drawn to a point at the further end. As the gas leaves the ta- pering end, which is bent upward, it is lighted. Thereupon the tube is heated with a flame on the near side of a narrowed place. The arsine is broken up and arsenic is deposited as a bright metallic mirror in the narrowed part. From the extent and thickness of the deposit one can estimate the number of milligrams of arsenic present. If the hydro- gen arsenide is not heated, it passes on to the flame and is burned. A FIG. 38. MARSH APPARATUS. cold porcelain dish held in the flame is soon coated with a deposit of arsenic, which is readily soluble in sodium hypochlorite solution (sodium arsenate being formed) . This solubility enables us to distinguish arsenic from antimony. Arsenic is very widely distributed, although in small amounts; hence we always have to reckon with the possiblity of traces of it being present in the reagents and glass utensils of the laboratory. In order to test this a "blank experiment" is performed, i.e. all the operations are carried out with duplicate amounts of the required chemicals but without the addition of the substance to be analyzed. Not until the materials used are proved to be free from arsenic is it permissible to use them in an actual test. Whether or not textile fabrics and the like have been dyed with Schweinfurt green (copper arsenite) can be determined easily by the GUTZEIT test. Another method is to use the above-mentioned peni- cillium brevicaule. This is cultivated on bread which is soaked with 226 INORGANIC CHEMISTRY. L 157- the liquid to be tested for arsenic. The least trace of the latter reveals itself by a characteristic garlic-like odor, caused by the evolution of arsenical gases. Compounds of Arsenic with the Halogens. 158. Three arsenic-halogen compounds of the type AsX 5 are known; viz., the pentachloride, AsCl 5 , the penta-iodide, AsI 5 , and the pentafluoride, AsF 5 . Aside from these only compounds of the type AsXs are known. Arsenic trichloride, AsCls, can be obtained by direct synthesis or by the action of hydrochloric acid on white arsenic. The latter way is analogous to the formation of metal chlorides from the oxide and hydrochloric acid. This compound is a colorless oily liquid having a specific gravity of 2.205 (dS). It freezes at 18 and boils at 130.2. It is extremely poisonous. When exposed to the air it throws off dense white fumes. With a little water it forms an oxychloride, As(OH) 2 Cl; with much water hydrochloric acid and arsenious oxide. In this latter system a rise of temperature results in partial re-formation of the trichloride, which is volatile with the water vapor. The following equilibrium seems to exist: As 2 O 3 + 6HCl <= 2AsCl 3 + 3H 2 O. Oxygen Compounds of Arsenic. Two such compounds are known: As 2 O3, arsenious oxide, and arsenic oxide. ARSENIOUS OXIDE, As 2 3 . 159. Arsenious anhydride (commonly called "arsenic" or " white arsenic") is found in nature. It is formed by the com- bustion of arsenic in air or oxygen and by the oxidation of arsenic with dilute nitric acid. It is manufactured commercially by roasting arsenical ores; the oxide volatilizes and is condensed in brick-walled chambers, where it collects as a white powder ("arsenic meal") .It is refined by sublimation from iron cylinders,, Physical Properties. Arsenious oxide is an odorless solid, that does not melt under ordinary pressure, but sublimes. Under higher pressure it is possible to melt it. At 800 its vapor density 159:] ARSENIOUS OXIDE. 227 is 198 (O = 16), which makes the molecular formula AS^G. Above this temperature dissociation begins and at 1800 the vapor density corresponds to the formula As 2 O3. By the ebullioscopic method (elevation of the boiling-point) the molecular formula has been found to be As^e at 205 also (in boiling nitrobenzene). Various Modifications. Arsenious oxide is known in a vitreous form as well as in crystals of the regular and monoclinic systems. The vitreous modification is produced when the compound is sub- limed or heated to the sublimation-point. Sp. g. =3.738. After stand- ing for some time at ordinary temperatures, this form becomes white like porcelain because of conversion into isometric crystals. The latter form is better obtained by dissolving the vitreous modification in water or hydrochloric acid and letting it crystallize out. During the crystal- lization the strange phenomenon of bright luminescence is observed, which is caused by the breaking of the crystals. This phenomenon, which is also noticed in other crystallizations, is called tribolumines- cence. The transformation of the amorphous into the regular variety is accompanied by the evolution of heat (5.330 Cal.). The monoclinic form is obtained by conducting the crystallization above 200 instead of at ordinary temperatures. If the lower half of a sealed glass tube containing arsenious oxide be heated above 400, it will be found after cooling that the lower heated part contains vitreous, the middle mono- clinic, and the upper octahedral, arsenious oxide. Since the transformation of amorphous into crystallized arsenious oxide takes place even at ordinary temperatures (rapidly at 100) and with the evolution of heat, the octahedral form is to be regarded as the stable one at ordinary temperatures; the glassy form is only able to exist at these temperatures, because the velocity of transformation is then very small. According to the above, if octahedral arsenious oxide is gradually warmed, we have first a transformation into monoclinic and then another into amorphous arsenious oxide. The transition tem- peratures have not yet been determined. Chemical Properties. Arsenious oxide is easily reduced to arsenic ; for example, by heating with charcoal or nascent hydrogen. It is also easily oxidized to arsenic oxide and is therefore useful as a reducing-agent. This oxidation can be brought about by chlorine, bromine (bromine- wa ter) , iodine solution, potassium perman- ganate, strong nitric acid, etc. It is slightly soluble in water; the solution has a salty metallic taste and a weak acid reaction. In acids it dissolves much more easily, because it acts towards them as a basic oxide. It was stated above ( 158) that a solution of 228 INORGANIC CHEMISTRY. [ 153- the oxide in hydrochloric acid gives off arsenious chloride. White arsenic is a rank poison; freshly precipitated ferric hydroxide serves as an antidote. ARSENIC OXIDE, 160. This compound cannot be prepared like the correspond- ing phosphorus compound by burning arsenic in the air, for the oxidation goes no farther than to arsenious oxide. The higher oxide can only be prepared by heating arsenic acid in the air : 2H 3 AsO 4 -3H 2 O = As 2 5 . This arsenic anhydride is a white glassy substance, that dis- solves in water slowly, going over into arsenic acid. By heating with carbon it is easily reduced to arsenic. At an elevated tem- perature it breaks up into oxygen and arsenious oxide. Its molec- ular weight is not known; the formula As20s is simply empirical. Oxyacids of Arsenic. Two of these are known: arsenious acid, H 3 As03 (only in aqueous solution and salts) and arsenic acid, H 3 As04. ARSENIOUS ACID, H 3 As0 3 . 161. This acid exists in the aqueous solution of the anhydride. It still remains to be discovered, however, which hydrate, H 3 AsO3, HAsC>2 or some other, is present. On evaporation the anhydride and not the acid separates out. This acid forms three classes of salts, according as one, two, or three of its hydrogen atoms are replaced "by metals; it is therefore tribasic. Certain salts are known which are derived from a meta-arsenious acid, HAsO2- The salts of the alkalies are soluble in water; those of the other metals are not, but dissolve easily in acids, however. A neutral arsenite- solution gives a yellow precipitate of silver arsenite, Ag 3 As0 3 , with silver nitrate. The solution of the free acid is easily oxidized to arsenic acid by iodine solution: H 3 AsO 3 + I 2 +H 2 = H 3 As0 4 +2HL Such a solution can therefore also be employed for the titration of iodine ( 93). 163.] ARSENIC ACID. 229 ARSENIC ACID, H 3 As0 4 . 162. This acid is most easily obtained by the oxidation of a solution of arsenious acid by warming it with nitric acid. On concentrating the solution the compound 2H 3 AsO4+H 2 separates out (below 15) ; this substance gives off its water of crystallization at 100 and yields orthoarsenic acid, H 3 AsO4, which crystallizes in fine needles. When heated further it gives off water (at 180) and goes over into pyroarsenic acid, H^A^Oy, which separates in the form of hard glistening crystals. On being heated still higher the latter compound gives up another molecule of water, the final product being white crystalline meta-arsenic acid, HAsO 3 . This conduct is completely analogous to that of phosphoric acid; however, metaphosphoric acid cannot be converted into the anhy- dride by heat as can arsenic acid ( 160). The pyro- and meta- arsenic acids are stable only in the solid state; when treated with water they are converted into the ortho acid, the transformation being much quicker than with the corresponding phosphorous acids. Orthoarsenic acid is easily soluble in water. Its salts, the arsenates, exist in three classes; of the tertiary only those of the alkalies are soluble in water. The reactions of arsenic acid are very similar to those of phosphoric acid ( 146) ; in this case also a mixture of ammonia, ammonium chloride and magnesium sul- phate (magnesia mixture) precipitates a white crystalline am- monium magnesium salt, Mg(NH 4 )AsO4+6H 2 0. Ammonium molybdate produces a yellow finely crystalline precipitate, whose composition and appearance correspond to those of the phos- phorus compound. The precipitates formed with silver nitrate are, however, unlike in color: Ag 3 P04 is yellow, Ag 3 AsO* reddish brown. Sulphur Compounds of Arsenic. 163. Three are known: arsenic disulphide (realgar) , arsenic trisulphide (orpiment), As 2 S 3 ; arsenic pentasulphide, Arsenic disulphide, As 2 S 2 , occurs in nature as realgar (155). It forms beautiful ruby-red crystals of a specific gravity of 3.5. It is used as a pigment. It is manufactured artificially by fusing sulphur and arsenic together; the resulting products vary in composition, however. 230 INORGANIC CHEMISTRY. [ 163- ARSENIC TRISULPHIDE, Arsenic is precipitated from the acid solution of arsenious oxide by sulphuretted hydrogen as sulphide; in this respect too it behaves as a heavy metal. In the above reaction arsenic tri- sulphide is deposited as an amorphous yellow powder. A pure solution of arsenious acid gives no precipitate with sulphuretted hydrogen, but simply a yellow liquid ( 196). Arsenic trisulphide occurs in nature as orpiment ( 155), having a laminated crystal- line structure; it owes its name to its beautiful golden lustre. By fusing artificial arsenic trisulphide a product is obtained which is very similar to the natural orpiment, but has a lower specific gravity (2.7 instead of 3.4). Commercially the trisulphide is prepared by fusing white arsenic with sulphur; the product still contains the oxide, however, and is therefore poisonous. Arsenic trisulphide is insoluble in water and in acids. ARSENIC PENTASULPHIDE, After sulphuretted hydrogen has been led into a warm acidu- lated solution of arsenic acid for some time, arsenic is precipi- tated as an amorphous yellow powder of the composition As 2 S 5 . The latter is also obtained by fusing arsenic trisulphide with the required amount of sulphur. In the absence of air it can be sublimed without decomposition. It is insoluble in water and in acids. SULPHO-SALTS OF ARSENIC. 164. The trisulphide and the pentasulphide of arsenic dissolve easily in alkali sulphides, forming salts of sulpho-acids: As 2 S 3 + 3K 2 S = 2K 3 AsS 3 ; Pot. sulph- arsenite. As 2 S 5 +3K 2 S=2K 3 AsS 4 . Pot. sulph- arsenate. The formation of these sulpho-salts can be regarded as analo- gous to that of an oxy-salt from a basic oxide and an acid anhv- dride, e.g. : BaO + SO 3 =BaSO 4 . ANTIMONY. . 231 The trisulphide and the pentasulphide are therefore to be con- sidered as sulpho-anhydrides of those sulpho-acids. The sulpharsenates can also be obtained from arsenic tri- sulphide with the aid of an alkali polysulphide: As 2 S 3 + K 2 S 3 = 2KAsS 3 . Pot. sulpho- meta-arsenate. This reaction can be explained by supposing that the arsenic trisulphide is converted into the pentasulphide by the excess of sulphur, just as the trioxide is oxidized to the pentoxide. They are also produced by treating an arsenate with hydro- gen sulphide: = K 3 AsS 4 + 4H 2 0. The sulpharsenates and sulpharsenites of the alkalies dis- solve easily in water and can be obtained in the crystalline form from the solution; those of the other metals are insoluble. The free sulpho-acids are unknown. On the addition of an acid to the solution of a sulpho-salt, the liberated sulpho-acid breaks up into hydrogen sulphide and arsenic tri- or pentasulphide. ANTIMONY. 165. Antimony occurs in nature in stibnite, Sb 2 S 3 , as well as in many less common minerals. Stibnite was known to the ancients. In Japan it is found in magnificent large crystals. Antimony was frequently employed by the alchemists. BASILIUS VALENTINUS in the latter part of the fifteenth century described its extraction from stibnite in a monograph entitled " The tri- umphal car of Antimonium." The element is at present obtained from stibnite by two processes. In one the mineral is roasted, being thus transformed into antimonious oxide. This oxide is then reduced with charcoal to metallic antimony: I. 2Sb 2 S 3 +90 2 =2Sb 2 O 3 +6S0 2 ; II. 2Sb 2 O 3 +3C =4Sb + 3CO 2 . The other method is to fuse the mineral with iron: 232 INORGANIC CHEMISTRY. [ 165- The crude antimony thus obtained usually still contains arsenic, lead, sulphur, etc. It can be refined by fusing with a little salt- petre, the impurities being oxidized. Physical Properties. Antimony is silvery-white and has a high metallic lustre and a laminate-crystalline structure (rhombo- hedral) ; as a result of the latter it is very brittle and can be easily pulverized. Sp. g. = 6.71-6.86. Melting-point, 629 boiling- point, 1440. MENSCHING and V. MEYER succeeded in determin- ing the vapor density at 1437, i.e. slightly below the boiling-point, and found that the molecule, unlike that of phosphorus or arsenic, consists of less than four atoms. Like arsenic antimony has a black and a yellow modification; the latter is obtained by passing air into liquid stibine, cooled to 90. It is even less stable than the yellow arsenic. Chemical Properties. At ordinary temperatures the element is not affected by the air; when heated, it burns with a bluish- white flame to the trioxide. It combines with the halogens directly, producing scintillations ( 27). It is dissolved by hydrochloric acid, although very slowly, with the evolution of hydrogen, thus asserting its metallic character. Aqua regia dissolves it readily. Uses. Antimony is a constituent of various alloys. The most important of these is type-metal, from which printer's type, is made. Its approximate compostion is lead (50%), antimony (25%) and tin (25%). HYDROGEN ANTIMONIDE, STIBINE, SbH 3 . 166. Stibine is formed when nascent hydrogen acts on a solu- ble antimony compound. It is best prepared by treating an alloy of one part of antimony and two parts of magnesium with dilute hydrochloric acid. The product consists principally of hydrogen, but contains 10-14% SbH 3 . If this gas mixture is passed through a U-tube and the whole is plunged in liquid air, stibine con- denses to a white solid mass, that soon melts after the tube is removed from the liquid air. It vaporizes to a relatively stable gas. The least trace of oxygen, however, causes some antimony to be deposited. If an electric spark is passed through stibine gas it explodes,, antimony is set free and the volume of hydrogen liberated is found to be 1J times that of the stibine, which is in accord with the formula SbH 3 . It is also decomposed rapidly by heating the containing vessel above 150. 167.] HALOGEN COMPOUNDS OF ANTIMONY. 233 Stibine has a characteristic musty odor, quite unlike that of phosphine or arsine. When the mixture of hydrogen and stibine evolved from the alloy of antimony is heated, as in the MARSH experiment ( 157), it produces a metallic mirror and, when ignited, the flame gives a spot on cold porcelain similar to that of arsenic, but differing from the latter in its darker color, insolubility in hypochlorite solution and less volatility when heated in a current of hydrogen. Stibine precipitates a black powder from silver solution, consisting of a mixture of silver and silver antimonide, The decomposition of stibine has been carefully investigated by STOCK. He arrived at the conclusion that the decomposition velocity in clean glass vessels at room temperature proceeds at first with extreme slowness but increases more and more as more antimony separates out. Furthermore, mirrors of antimony produced by heating stibine and mirrors of black antimony made by subliming antimony in a vacuum and condensing the vapor at the temperature of liquid air, and mirrors of sublimed metallic antimony, all had different effects. The effective- ness of the mirrors varied not only with the size, but in large measure also with the form of the antimony surface. It was found that the stibine dissociation in the layer adsorbed by antimony was proportional to the mass of the layer; under this assump- tion the progress of the dissociation could be calculated theoretically, and was found to agree well with the experimental results. Since the amounts of adsorbed gas depend on the surface tension and, therefore, on the form of the adsorbing surfaces, the explanation of the influence of the different antimony mirrors is obvious. The fact that the walls of the vessel influence the velocity of a reaction has also been established in many other instances. Halogen Compounds of Antimony. 167. Two compounds of this element with chlorine are known: SbCl 3 and SbCl 5 . Antimony trichloride, SbCls, is obtained by treating antimony sulphide or oxide with concentrated hydrochloric acid. It forms a colorless laminar-crystalline mass, which is so soft that it was formerly known as "antimony butter" (butyrwn antimonii). Its melting-point is 73.5 and its boiling-point 223.5; its vapor density 7.8 (air = l) makes the formula SbCls. It dissolves in water containing hydrochloric acid. Water 234 INORGANIC CHEMISTRY. [167- decomposes it, forming difficultly soluble oxy chlorides. The composition of the precipitate depends on the amount and the temperature of the water used in the decomposition. There is evidence of the existence of the compounds SbOCl and Sb 4 5 Cl 2 (=2SbOCl, Sb 2 O 3 ), both of which crystallize. The pre- cipitated oxychlorides on being repeatedly boiled with water eventu- ally lose all their chlorine and go over into the trioxide, Sb 2 O 3 . Powder of Algaroth, once used in medicine, is obtained by the decomposition of antimony trichloride with water and has nearly the same formula as the second of the above-mentioned oxy- chlorides. Antimony pentachloride, SbCl 5; is prepared by heating anti- mony in a current of chlorine or treating fused trichloride with chlorine. It is a yellow, fuming, ill-smelling liquid, which crys- tallizes at 6. When heated it dissociates into the trichloride and chlorine. It unites with water, forming SbCl 5 -H 2 O and SbCl54H 2 0. Hot water decomposes it into hydrochloric and pyroantimonic acids. Oxygen Compounds of Antimony. 168. Three are known: antimony trioxide, Sb 2 Os, antimony tetr oxide, Sb 2 O4, and antimony pentqxide, Sb 2 O 5 . Antimony trioxide occurs as a mineral, senarmontite. It can be obtained by burning antimony in the air, as well as by the oxidation of antimony with dilute nitric acid. It is dimorphic, occurring in both regular and rhombic crystals. It is'x^a light yellow crystalline powder, almost insoluble in water. It volatilizes at 1560; the vapor density at this tem- perature corresponds to the formula Sb^e- It is insoluble in sulphuric and nitric acids but easily soluble in hydrochloric and tartaric acids and in alkalies. On being heated in the air it turns to the tetroxide. The corresponding hydroxide is Sb(OH) 3 . This hydrate sepa- rates out when tartar emetic (see below) is decomposed with dilute sulphuric acid. It gives up one molecule of water readily and passes over into the hydroxide SbO -OH, meta-antimonious acid. The latter is more easily obtained by treating a solution of the trichloride with soda solution: 169.] ANTIMONY PENTOXIDE AND ANTIMONIC ACID. 235 2SbCl 3 + 3Na 2 CO 3 + H 2 O = 2SbO OH + GNaCl + 3CO 2 . It appears as a white precipitate, which is converted into an- timonic oxide by boiling with water. This meta-antimonious acid is dissolved by alkalies, forming salts of the acid. One of them which has been obtained crystallized is the sodium meta- antimonite, NaSbO 2 +3H 2 O. The latter is difficultly soluble in water, and decomposes on concentration of its solution. On the other hand, antimony hydroxide displays basic proper- ties by uniting with acids to form salts. There are salts known of Sb(OH) 3 , as well as of SbO -OH. Examples of the former kind* are the crystallized antimony sulphate, Sb 2 (SO 4 ) 3 , and the nitrate, Sb(NO 3 ) 3 . In analogy with other trivalent metals double salts are known, e.g. KSb(SO 4 ) 2 . As to the salts derived from SbO-OH, we may look upon the group SbO as taking the place of a uni- valent metal. Thus SbO-OH may be compared with KOH. For this reason the group (SbO) has been given the name antimonyl; one of its salts is antimonyl sulphate, (SbO) 2 S04. The most familiar antimonyl compound is tartar emetic, potassium antimonyl tartrate, (SbO) ^ 4 ^- 4 ^ 6 ~^~ i^ 2 O, which is employed in medicine. See ORG. CHEM. 192. ANTIMONY PENTOXIDE AND ANTIMONIC ACID. Antimonic acid, H 3 Sb04, is obtained by warming antimony with concentrated nitric acid and also by decomposing the penta- chloride with water. It is a white powder, almost insoluble in water and nitric acid; nevertheless, when moist, it turns litmus paper red. On heating saltpetre with powdered antimony the potassium salt of meta-antimonic acid, KSb0 3 , is formed in an explosive reaction. When this is boiled with water it dissolves, producing monopotassium orthoantimoniate, KH 2 SbO4; on fusing with potash potassium pyroantimoniate, K4Sb 2 Oy, is formed, which dissolves in water, giving 2KOH and K 2 H 2 Sb 2 O 7 +6H 2 0. In the case of antimony, as in that of phosphorus, we meet with three kinds of acids belonging to the highest stage of oxidation: their formulae correspond to those of the analogous phosphorus compounds. 236 INORGANIC CHEMISTRY. Antimony pentoxide, Sb 2 0s (molecular weight unknown), can be obtained by heating antimonic acid at 300. It is a yellow amorphous powder, soluble in hydrochloric acid. If heated strongly it gives up part of its oxygen and goes over into anti- mony tetroxide, Sb 2 C>4, a white powder that turns yellow on heating but resumes its original color on cooling. This tetroxide can be regarded as antimonyl meta-antimoniate, SbOa-SbO. Sulphur Compounds of Antimony. 169. Antimony trisulphide, Sb 2 S 3 , is found in nature ( 165). It can be made by leading hydrogen sulphide into a hydrochloric acid solution of the trichloride, from which it is deposited as an amorphous red powder. It can be melteci; on cooling it crystal- lizes and takes on the appearance of stibnite. Antimony pentasulphide, Sb 2 S 5 , is precipitated when hydrogen sulphide is passed into the acidified solution of antimonic acid. It is more easily obtained by the decomposition of sodium sulph- antimoniate with dilute sulphuric acid. It forms an amorphous orange-red powder, which splits up into sulphur and the trisulphide on being strongly heated. It is insoluble in dilute acids; boiling- iiot concentrated hydrochloric acid dissolves it, forming antimony trichloride, hydrogen sulphide and sulphur. In aqueous solutions of alkalies and their sulphides it dissolves easily with the formation of sulphantimoniates, M3SbS4 The best known of these is sodium sulphantimoniate, Na 3 SbS 4 +9H 2 (" SCHLIPPE'S salt"). It can be obtained by boiling antimony trisulphide with sulphur and caustic soda solution. It crystallizes in large colorless tetrahedrons, is easily soluble in water (1 part by weight in 2.9 parts water at 15) and reacts alkaline. It is decomposed by acids, depositing pentasulphide; even carbonic acid causes this, hence the crystals become covered with a yellowish-red coating of pentasulphide after having stood some time in the air. The free sulphantimonic acid is not known. BISMUTH. 170. This elejnent belongs undoubtedly among the metals, so far as its physical character is concerned; its chemical properties also class it with them in almost every respect, inasmuch as its oxides are mainly basic in their behavior. 171.] COMPOUNDS OF BISMUTH. 237 It is found chiefly in the native state; but a sulphide, Bi 2 Ss, bismuth glance and a telluride, tetradymite, also occur in nature. Bismuth is obtained from the latter by roasting to the oxide Bi 2 3 and reducing with charcoal. The native metal is usually very pure. If refining is necessary, the fused metal is allowed to flow over a hot, somewhat inclined iron plate, so that the impurities are oxidized. The amount of bismuth found in nature is not very great. Physical Properties. Bismuth is externally very similar to antimony; it is crystallized and very brittle and has a metallic lustre, but differs from antimony in having a reddish-white color. Sp. g. = 9.823. It melts at 286.3 and boils at 1420. it can be distilled in a current of hydrogen. Chemical Properties. At ordinary temperatures bismuth is unaffected by the air. On being heated it turns to the trioxide. It combines with the halogens directly. It is not attacked by hydrochloric or sulphuric acid at ordinary temperatures, but nitric acid dissolves it readily to form the nitrate. On being heated with sulphuric acid, it gives off sulphur dioxide and forms the sul- phate. No hydrogen compound of bismuth is known. Bismuth is employed in the manufacture of easily fusible alloys such as are used in making casts of woodcuts, stereotypes, etc. The most common of these alloys are NEWTON'S metal (8 bismuth, 5 lead, 3 tin; melting-point 94.5), ROSE'S metal (2 bismuth, 1 lead, 1 tin; melting-point 93 75) and WOOD'S metal (4 bismuth, 2 lead, 1 tin, 1 cadmium; melting-point 60.5). Halogen Compounds. 171. Compounds of the type BiX 3 only are known. Bismuth chloride, BiCl 3 , is formed by direct synthesis from the elements, but more easily by dissolving bismuth in aqua regia. It is white and crystallized. Its melting-point is between 225 and 230 and its boiling-point at 435. Its vapor density, 11.35 (air=l), gives it the formula BiCl 3 . On being dissolved in a little water it forms a sirupy liquid; an excess of water gives bismuth oxychloride, BiOCl, and hydrochloric acid. This oxychloride is a white powder, in- soluble in water but soluble in acids. 238 INORGANIC CHEMISTRY. [ 172, Oxygen Compounds. 172. Three oxides are known: BiO, Bi 2 O 3 , and Bi02. Bismuthous oxide, BiO, is obtained by adding an alkaline stannous chloride solution to a solution of bismuth chloride. It is deposited as a dark-brown precipitate of BiO. When heated in the air it smolders like tinder. It is doubtful whether this precipitate is a homogeneous substance or a mixture of Bi 2 3 with finely divided bismuth. Bismuth trioxide, Bi 2 O 3 , is the most familiar oxide of this ele- ment. It has strictly basic properties. In order to prepare it we can heat the nitrate or carbonate or we can precipitate the hydroxide from the solution of a bismuth salt by means of a base and heat the precipitate. If a boiling solution of a bismuth salt is treated with caustic potash, the trioxide separates out in glistening needles of microscopic dimensions. Like the corresponding oxides of arsenic and antimony, it is dimorphic. Bismuth dioxide, BiO 2 , has been little studied; it is a reddish-yellow powder. Bismuth pentoxide, Bi 2 5 , appears as a reddish-brown powder, which is very unstable and evolves oxygen on heating, as it also does when warmed with sulphuric acid. Hydrochloric acid does not convert it into the corresponding pentachloride, BiCl 5 , but produces the trichloride BiCl 3 and free chlorine. Hydroxides and Salts. 173. Bismuth hydroxide, Bi(OH) 3 , is obtained by precipitating a bismuth salt with an alkali. It is an amorphous white powder, insoluble in potassium hydroxide or ammonia. At 100 it goes over into the compound BiO -OH with the loss of a molecule of water. Both of these hydroxides are wholly basic in character. The salts derived from Bi(OH) 3 are called neutral, those from BiO -OH basic. The neutral nitrate, Bi(N0 3 ) 3 , is obtained by dissolving bis- muth in nitric acid. It crystallizes with five molecules of water in large translucent triclinic prisms. It is deliquescent. The addition of much water converts it into the basic nitrates, several of which are known. By treating the neutral nitrate with about 20 parts of boiling water a product is obtained whose composition is not perfectly constant for different preparations, but corresponds nearly to the formula (BijAOe-^OsVC^O^, or 2BiONO 3 175.] SUMMARY OF THE NITROGEN GROUP. 239 + Bi(NO 3 ) 3 + 3Bi(OH) 3 . This is the bismuth subnitrate, which is used in medicine. Bismuth sulphate, Bi 2 (S04)3, is obtained as an amorphous white substance when the metal is heated with concentrated sul- phuric acid. With water it forms a basic sulphate, Bi 2 (OH) 4 SO4. Sulphur Compounds. 174. Bismuth trisulphide is found in nature ( 170); artificially it can be prepared by heating bismuth with sulphur or by leading hydrogen sulphide into the aqueous solution of a bismuth salt. In the latter case it comes down as an amorphous black powder that is easily soluble in warm dilute nitric acid. It is insoluble in alkalies and their sulphides, hence forms no sulpho-salts. When heated with an alkali sulphide solution to 200 it takes on a crys- talline form similar to that of the natural mineral. SUMMARY OF THE NITROGEN GROUP. 175. Like the halogens and the elements of the oxygen group, the elements just discussed, viz. nitrogen, phosphorus, arsenic, antimony and bismuth, also form a natural group. Their family relation shows itself even in the formula types of their compounds. The hydrogen compounds have the type RH 3 (lacking with bis- muth), the halogen compounds RX 3 and RX 5 (the latter also lack- ing with bismuth), the oxygen compounds R 2 Os and R 2 0s. In other words, the elements of this group are trivalent or pentivalent. We find here, just as in the groups previously studied, that, as the atomic weight increases, a gradual change occurs in the physical properties. This is shown by the following small table: N. P. As. Sb. Bi. Atomic weight. . . Specific gravity. . . ( Water = 1) Melting-point. . . . Boiling-point Color 14.01 0.885 liquid -194.4 colorless 31.0 1.8-2.1 + 44.4 + 278 yellow or red 74.96 4.7-5.7 ca. 800 gray 120.2 6.7 629 1440 white 208.0 9.8 286 1420 pink In the chemical properties, also, regular variations are to be observed, all of which can be summed up in the general statement 240 IXORGANIC CHEMISTRY. [175- that the metalloid character gives way to the metallic character as the atomic weight increases. Nitrogen forms either indifferent or acid-forming oxides only; so does phosphorus; arsenic, on the contrary, displays a very feebly basic character in arsenious oxide, since this oxide forms the trichloride with hydrochloric acid, the trichloride reacting inversely with water, however, and breaking up into hydrochloric acid and arsenious oxide. In antimony trioxide this basic character is a little stronger; some salts and double salts of it with acids are known. The corresponding chlo- ride does not suffer an immediate hydrolytic dissociation with water, but oxychlorides are formed, which require a great deal of water to convert them entirely into the trioxide. While the highest oxides of arsenic and antimony have strictly acid prop- erties, with bismuth the acidic nature has practically disappeared; the oxide Bi 2 Os has exclusively basic properties and the higher oxide Bi 2 O5 acts like a peroxide, giving off oxygen readily (it generates chlorine with hydrochloric acid) and going over into the lower oxide Bi 2 O3. Bismuth trichloride, Bids, gives the oxy- chloride, BiOCl, with water and this is not decomposed by an excess of water. In the hydrogen compounds, too, the gradual change of the properties is very apparent. Consider the stability for example: ammonia requires a very high temperature for decomposition; phosphine and arsine a much lower temperature; stibine is unstable at ordinary temperatures when it comes in contact with oxygen, and the hydrogen compound of bismuth is so unstable that the conditions for its formation and existence have not yet been ascertainable. A similar change is noticeable in their ability to form XELj ions in aqueous solutions; it is strong in ammonia, much weaker in phosphine and wholly absent in arsine and stibine. In the sulphur compounds a progressive change of color is observed. P 2 S 5 is bright yellow, As 2 S 5 deep yellow, Sb 2 $5 red and Bi 2 Ss black. The first three are sulpho-anhydrides of suipho- acids ( 164); bismuth sulphide is not, however, thus displaying again the more basic nature of bismuth. 176.J ALLOTROPIC FORMS OF CARBON. 241 CARBON. 176. Carbon occurs in nature both free and combined. In combination it is found in large quantities in the salts of carbonic acid, above all in calcium carbonate, limestone, which is of the widest occurrence and is even known to form great mountains. Farther, carbon is one of the constituent elements of animals and plants. It is found in these in numerous compounds. Still larger is the number of artificially prepared carbon compounds. The com- pounds of carbon exceed in number all other compounds together. For this reason and because of the peculiarities of the carbon compounds it is customary to treat them by themselves, as " organic chemistry." However, that we may be able to obtain a general survey of the elements, it is deemed advisable to discuss certain compounds of carbon in inorganic chemistry as well. Allotropic Forms of Carbon. We know of three: diamond, graphite and amorphous carbon. (a) Diamond. LAVOISIER found, in 1773, that this mineral can be burned to carbon dioxide. In 1814 DAVY proved that, when diamond burns, nothing else than this gas is formed, so that diamond must be pure carbon. Furthermore when the carbon dioxide given off by the combustion of diamond is absorbed by sodium hydroxide, a soda is produced which is in every respect identical with ordinary soda. Indeed, it has been found possible to manufacture diamonds from amorphous carbon (see below). The diamond crystallizes in the isometric system. Usually it is colorless, but yellow and black diamonds are also known; the black ones are called carbonado. The specific gravity of diamond is 3.50-3.55. It is a poor conductor of heat and electricity. The refractive index is very high: n=2A2. The, diamond is so hard that it scratches all other substances. If it is subjected to a very high temperature in the absence of air, it gradually turns to graphite. It resists the action of the strongest oxidizing-agents, e.g. a mixture of nitric acid and potassium chlorate. In 1893 MOISSAN succeeded in making diamonds artificially, although they were very small, the largest being about 0.5 mm. in 242 INORGANIC CHEMISTRY. [176- > diameter. His method consists essentially in dissolving carbon in molten iron at a high temperature and then cooling it rapidly. FIG. 39. ARTIFICIAL DIAMONDS (MAGNIFIED). This is accomplished as follows: Iron is brought in contact with pure carbon (sugar charcoal) in the electric furnace at a high tem- perature. After the iron has become saturated with carbon at about 3000, the fused mass is suddenly cooled by pouring it into a hole drilled in a copper block, which is kept cold by water, and at once covering, the cavity with an iron stopper. When the iron is all cold it is dissolved away by acids, leaving the carbon which did not combine with the iron. This residual carbon consists partly of small diamonds, which are identical with the natural diamond in hardness, crystal form, etc. Fig. 39 presents an enlarged view of some artificial specimens; they display the same properties as the rough natural diamonds, particularly the rounded edges and angles and the stria tions. The formation of the diamond by this method has been explained by BAKHUIS ROOZEBOOM as follows : In all probability the transition of diamond into graphite is endothermic. For this reason diamond is the more stable form at lower temperatures, graphite at higher ones, in analogy to the rhombic and monoclinic modifications of sulphur. But, while % the velocity of transformation of monoclinic sulphur is fairly great at low temperatures and the monoclinic sulphur can thus exist only for a short time below its transition point, the transition velocity of graphite into diamond is practically zero for temperatures below 1000. Carbon that has crystallized from molten iron in the form of graphite cannot, therefore, pass over into diamond. The rapid cooling of the molten iron, however, 176.] ALLOTROPIC FORMS OF CARBON. 243 has the effect of bringing the carbon into the region of temperature in which diamond is the stable modification; it can therefore separate in this form from its solution. The electric furnace that MOISSA.N used for these and numerous other experiments is very simple in construction. It consists of two blocks of unslaked lime that fit tightly together. In the lower block FIG. 40. MOISSAN'S ELECTRIC FURNACE (CROSS-SECTION). there is a trough in which the carbon terminals are laid. The upper block is slightly hollowed out on its lower side so as to reflect the heat rays on to the crucible. Fig. 40 shows a cross-section of an electric furnace, Fig. 41 a picture of the same apparatus in operation. The temperatures obtained in the electric furnace are as follows: Current of 30 amperes and 55 volts with a steam-engine of 4 H.P., 2250 100 450 45 70 50 2500 3000 FIG. 41. MOISSAN'S FURNACE IN OPERATION. (AFTER MOISSAN.) 244 INORGANIC CHEMISTRY. [ 176- The last-named temperature can however only be maintained for .a brief period, as the unslaked lime soon melts and flows like water. At 2500 the lime becomes crystalline in structure after a few minutes. (6) Graphite is also crystallized carbon. Unlike diamond, it is very soft and opaque and a good conductor of heat and electricity. Sp. g. =2.09-2.23. As was stated above, graphite can be pre- pared artificially by the crystallization of carbon from molten iron and by heating diamond strongly. There are various kinds of graphite. If graphite is treated with a mixture of perfectly dry potassium chlorate and very concentrated nitric acid, it turns to a yellow crystallized substance containing hydrogen and oxygen, in addition to carbon, and called graphitic acid. This substance is peculiar in that it decomposes explosively on heating and yields a large volume of extremely fine amorphous carbon. Graphite is used in the manufacture of lead pencils, crucibles, electrodes, polishes, etc. (c) Amorphous Carbon. This is obtained !n the purest state by charring sugar. The resulting mass is boiled with acids to remove the mineral matter and finally heated red-hot in a current of chlorine for quite a while to remove all the hydrogen. It can also be prepared from soot. Amorphous carbon is opaque, black and infusible. At the highest temperature that MOISSAN could reach with his furnace by employing a current of 2000 amperes and 80 volts (obtained with a 300 horse-power engine) it was barely possible to make carbon sublime. The sublimate was graphite. Amorphous carbon has a specific gravity of 1.5-2.3. Various sorts of amorphous carbon are known. They are probably different allotropic modifications, or mixtures of such. Gas carbon and coke are obtained as residues in the dry distilla- tion of coal. They conduct heat and electricity. Wood charcoal is very porous and can condense large quantities of gases in its pores, e.g. 90 times its own volume of ammonia (see also 111). When warmed or when the pressure is reduced, these gases all escape again. Bone-black is obtained by heating bones away from air; the resulting black mass is treated with hydrochloric acid to remove the phosphates and carbonates present. It has the power of absorbing coloring-matter and certain salts, e.g. lead salts, from liquids. The charcoal obtained from the dry distilla- tion of sugar is noted for its peculiar lustre. These different 177.] ALLOTROPIC FORMS OF CARBON. 245 sorts of charcoal do not consist of pure carbon but contain other substances in small proportions. It is a general rule that carbon conducts heat and electricity better the longer it has been exposed to a high temperature. 177. The various kinds of carbon all find their respective uses. Soot, or lampblack, serves for the preparation of India ink and black paint. Gas carbon (coke), being a good conductor of electricity, is used in the electrical industry. Wood charcoal is used in the manufacture of gun- powder ; animal charcoal, or bone-black, as a water-filter to remove color- ing-matter, ill-smelling gases or injurious salts (lead salts) from drink- ing-water ; it is also employed in enormous quantities in sugar refineries to decolorize sugar liquids. By far the most important use of carbon is as a f u e 1 . The heat generated by the burning of coal warms our houses, drives our steam- engines, etc. The principal kinds used as fuels are charcoal, coke, anthracite coal, bituminous coal, brown coal (lignite) and peat. Charcoal (wood charcoal) is made on a large scale by the colliers. Long sticks of wood are piled in a large heap, covered with sod and ignited at the bottom. The wood smolders away slowly and becomes completely charred. This " charcoal-pit " process is not at all economical, inasmuch as all the volatile products are lost; it is carried on exten- sively (Fig. 42), but it is being more and more replaced by the dry disr filiation of wood from iron retorts, in which process the gaseous and tarry products are recovered. FIG. 42. CHARCOAL PIT. Coke is the residue in the retorts of the gas factories after the coal has been deprived of its volatile products by heating. It is also manu- factured on a large scale for metallurgical and other purposes. Coke is thought by many to have a great future as a fuel, since it is a hard- burning smokeless fuel, manufactured from the cheap soft coal. 246 INORGANIC CHEMISTRY. [ 177- Peat and the various coals owe their origin to the same geological process, the slow decay of plant-remains in the absence of air. Peat is the youngest formation and anthracite coal the oldest. During this transition carbon dioxide and methane, CH 4 , are given off and the residue becomes richer in carbon and poorer in hydrogen and oxygen than the corresponding chief constituent of plant tissues, cellulose. The follow- ing table shows this: Carbon. Hydrogen. Oxygen. Cellulose 50.0% 6.0% 44 0% Peat 60.0 5.9 34 1 Brown coal 67.0 5.8 27 2 Cannel coal 85 8 5 8 8 3 Anthracite coal 94 3 4 2 6 The plants of which these formations originally consisted are different. Peat appears from its structure to have come chiefly from swampy growths, mosses and the like; mineral coal from extinct plants, gigantic horsetails (equiseta), lepidodendra and sigillariae. Molecular and Atomic Weight of Carbon. Chemical Properties. 178. The carbon molecule probably contains a large number of atoms. It has not yet been possible to determine how large this number is. It is supposed that graphite has a larger number of atoms to the molecule than amorphous carbon, and diamond more than graphite, since graphite and diamond are less easily attacked by chemical reagents and because they are denser. A determination of the vapor density of carbon is of course out of the question. The measurement of the melting-point depression that carbon produces in iron is also impracticable; however, it is known that even a small percentage of carbon causes a considerable lowering of the melting-point of iron (see 304). It can be shown in the following way, however, that the number of atoms in the carbon molecule must be very great. By the oxidation of amorphous carbon with potassium permanganate mellitic acid is formed, which contains 12 carbon atoms to the molecule. This makes it quite probable that the carbon mole- cule contains at least 12 atoms, for in the oxidation of organic substances the products almost always contain either a smaller or the same number of carbon atoms to the molecule. For the 178.] MOLECULAR AND ATOMIC WEIGHT OF CARBON. 247 following reason it is, however, to be supposed that the number of atoms in the carbon molecule is much greater than 12. When marsh-gas, CH4, is passed through a red-hot tube, ethylene, C2-H4, is formed among other things. If this is then treated in the same way, acetylene, C2H 2 , is obtained, and from this again benzene, C 6 H 6 . On conducting benzene vapor through a glowing tube, naphthalene, CioH 8 , pyrene, Ci 6 H 10 , etc., are formed. If either of the latter is heated still higher (in the absence of air) carbon is deposited. We thus see that as the temperature rises the num- ber of carbon atoms in the molecule steadily increases. The final product of these operations, carbon, will therefore probably contain a considerably larger number of atoms in its molecule than naphthalene or pyrene. Carbon can unite directly with many elements. At ordinary temperatures it combines with fluorine only. MOISSAN intro- duced lampblack into fluorine gas, and the carbon commenced to glow; when fluorine was present in excess carbon tetraftuoride, CF4, was formed. Hydrogen combines with carbon directly to form acety- lene and a small quantity of marsh-gas, when an electric arc is passed between two carbons in an atmosphere of hydrogen. Of all the many compounds consisting of only carbon and hydrogen these are the only ones which can be obtained by direct synthesis. Under analogous conditions carbon unites with chlorine to form perchloroethane, C2C16, and hexachlorobenzene, CeCle- Oxygen unites with carbon at an elevated temperature to form carbon monoxide, CO, or carbon dioxide, C02, according as carbon or oxygen is in excess. If sulphur vapor is passed over red-hot coals, carbon disulphide, CS2, is produced. The elements of the nitrogen group, N, P ; As, Sb and Bi, do not combine with carbon directly. Silicon and car- bon unite at the temperature of the electric furnace to form CSi, carborundum, which is so hard that it can be used as a powder for polishing glass and precious stones. MOISSAN also found that many metals are able to combine with carbon at a very high temperature, forming carbides. This was previously known to be true of iron and certain other metals. 248 INORGANIC CHEMISTRY. [ 179- The difference in the behavior of these carbides towards water is interesting. Iron carbide is unaffected by it; calcium carbide gives acetylene, C 2 H 2 ; aluminium carbide yields methane; other carbides give mixtures of the two hydrocarbons; uranium carbide produces methane and also liquid and solid hydrocarbons. 179. The atomic weight of carbon has been determined with great accuracy by DUMAS and STAS. The averages for the different series of experiments, each of which showed little variation, were as follows: Ratio by weight of carbon to oxygen in carbon di- oxide from the combustion of: Natural graphite 2.9994:8.0000 Artificial " 2.9995:8.0000 Diamond 3.0002:8.0000 The ratio of carbon to oxygen in carbon dioxide is thus very close to 3:8. As the specific gravity of carbon dioxide points to a molecular weight of 44 for this gas, it must contain, according to this ratio, 12.00 parts by weight of carbon and 32 parts of oxygen. The formula is therefore G/^. Inasmuch as no carbon compound is known whose molecular weight includes less than 12 parts of carbon, we have CO 2 as the formula; hence the atomic weight of carbon must be 12.00 for 0=16. Compounds with Hydrogen. 180. Carbon and hydrogen form a very large number of com- pounds (hydrocarbons), which are more fully discussed in organic chemistry. Two of them will be treated here briefly. Methane, also called marsh-gas and fire-damp, is the only hydro- carbon with just one atom of carbon. It occurs in nature in volcanic gases; moreover, it gushes out of the ground in the neigh- borhood of 'the oil-wells at Baku and various places in America. It is an important constituent of " natural gas." It owes the name "marsh-gas" to the fact that it arises from swamps, especially when the decaying vegetation at the bottom is stirred up. It is called " fire-damp " because it occurs in coal beds ( 177), from which it escapes when they are broken up. It forms a violently explosive mixture with air, which is frequently the cause of mine explosions. For its modes of formation and its physical and chem- ical properties reference should be had to OBG. CHEM., 29. 181.] COMPOUNDS WITH OXYGEN. 249 181. Acetylene, C 2 H 2 , is a colorless gas of a disagreeable odor. It is soluble in an equal volume of water at 18 and becomes liquid at 18 under 83 atmospheres. Its hydrogen atoms are replaceable by metals. It is manufactured by decomposing calcium carbide with water : CaC 2 +2H 2 0=Ca(OH) 2 + C 2 H 2 . Calcium carbide is prepared by heating coke with unslaked lime (CaO) in the electric furnace. The calcium formed by the action of carbon on lime unites with carbon at the high temperature of the fur- nace to form CaC 2 . Acetylene burns with a vivid flame on coming out of a small orifice under pressure. Since it can be prepared from calcium carbide pretty cheaply, it is used rather extensively in small systems for illuminating purposes. When mixed with air and ignited it explodes vehemently; the compounds with metals are also explosive. It is endothermic and can be exploded by fulminating mercury. The combustion of acetylene is another illustration of the rule of 137, that reactions are in most cases of a simpler nature than the chemical equations indicate. The equation here is: 2C 2 H 2 +50 2 =4C0 2 +2H 2 0. According to this equation the combustion should be septimolecular. BONE and CAIN proved, however, that the reaction has more than one stage, the first stage being represented by the bimolecular equation : CO and H 2 then burn further to CO 2 and H 2 O. From a kinetic standpoint, it is quite conceivable that polymolecular reactions should be rare, for the probability of a large number of mole- cules coming together in just such a way that a reaction can take place is indeed very slight. The reaction is more likely to proceed in a way which involves the interaction of only very few molecules. Compounds with Oxygen. Three oxygen compounds of carbon are known: carbon monoxide, CO, carbon dioxide, CO2, and carbon suboxide } CsC2. For the latter compound, see ORG. CHEM. 166. CARBON MONOXIDE, CO. 182. This gaseous compound is always formed when carbon burns in a limited supply of air or oxygen. A number of carbon compounds also yield carbon monoxide when burned under this same condition. It can also be obtained by the action of carbon on oxygen compounds, e.g., by heating zinc oxide, ZnO with 250 INORGANIC CHEMISTRY. [ i 82 _ carbon. On passing steam over red-hot coals a mixture of hydrogen and carbon monoxide is produced: This mixture goes by the name of water-gas. It is used on a large scale for heating and lighting, especially in America. For the latter purpose it is charged with the vapor of hydrocarbons rich in carbon, since its own flame is not luminous. The use of the incandescent gas- light (291) makes this " carburetting " unnecessary. Water-gas con- taining 50% of carbon monoxide is very poisonous (ORG.CHEM. 241). Carbon monoxide is also formed by the reduction of carbon dioxide with red-hot carbon: C + C0 2 =2CO. This reaction is limited by the reverse one and we have here a case of balanced action expressed by In view of the caloric effect of the reaction, 2CO=C + CO 2 + 3900CaL, an elevation of temperature must, according to LE CHATELIER'S rule ( 51), increase the amount of carbon monoxide; a depression of temperature, the opposite. Experience has shown this to be actually the case. As the temperature rises the quantity of carbon monoxide increases rapidly and at 1000 there is still a very small amount of dioxide. At 445, on the other hand, practically all the carbon monoxide is changed into carbon dioxide and carbon. This result is surprising, because the same change should also occur at lower temperatures; nevertheless, carbon monoxide seems perfectly stable at ordinary temperatures, even as high as 200. The cause of this phenomenon must, as in analogous cases, be sought in the very great retardation of the velocity of the reaction 2CO >CO 2 + C when the temperature sinks. On using certain catalyzers, e.g. finely divided nickel, the velocity of the reaction 2CO >CO 2 + C becomes measurable as low as 256. These measurements have shown that the decomposition of carbon monoxide into carbon dioxide and carbon is not a bimolecu- lar reaction, as would be expected from the above equation, but 183 j CARBON MONOXIDE. 251 a unimolecular one. To explain this it may be suggested that the decomposition takes place in two stages: I. CO=C-fO; II. CO + O=CO2- If we assume that the second stage has an infinite velocity, it is only the first that is really measured, i.e. a unimolecular reaction. The reduction of salts of carbonic acid also furnishes a method of preparing carbon monoxide. If chalk (CaCOs) or magnesite (MgCOs) is heated with zinc dust, pure carbon monoxide is formed: CaC0 3 + Zn = CaO + ZnO + CO. Physical Properties. Carbon monoxide is a colorless, odorless gas of a specific gravity of 0.967 (air = l). It is hard to condense, its critical temperature being 139.5 and its critical pressure 35.5 atmospheres. It boils at -190 and solidifies at -211. It is only slightly soluble in water. 183. Chemical Properties. Carbon monoxide burns with a characteristic blue flame to carbon dioxide. It can unite with chlorine directly to form phosgene, COC12, and also with sulphur (at an elevated temperature) to form carbon oxy sulphide, COS ? both compounds are gaseous. Again, it unites directly with nickel and iron, giving the compounds Ni(CO)4 and Fe(CO) 5 ( 214 and 311). On account of its tendency to combine with oxygen, it displays strong reducing power, especially at high temperatures. Thus metallic oxides, like Fe20s, CuO, etc., are easily converted into the metals when hot. Some compounds are reduced by carbon mon- oxide even at ordinary temperatures. Palladium is precipitated from an aqueous solution of palladious chloride and an ammoniacal silver solution (prepared by dissolving silver oxide in ammonium hydroxide to the point of saturation) is turned black by the gas on account of formation of the metal. Both of these reactions serve for the detection of carbon monoxide. An ammoniacal cuprous chloride solution absorbs the gas because of the formation of a compound, Cu2Cl 2 -CO + 2H 2 O, which can be isolated in the crystalline state but decomposes again very readily. The composition of carbon monoxide can be determined by exploding a mixture of the gas with oxygen. It is then found that 2 vols. CO unite with 1 vol. O 2 to form 2 vols. CO 2 . This together with the vapor density establishes the formula as CO. 252 INORGANIC CHEMISTRY. [ 183- CARBON DIOXIDE, CARBONIC ACID ANHYDRIDE, C0 2 . 184. This compound occurs not only by itself but also in com- bination. It is a regular constituent of the air ( 106); many mineral waters contain the free gas; in some places of the earth (in the Dog's Grotto at Naples and the famous Poison Valley in Java) it comes up out of the ground and it is also found in volcanic exhalations. The most minerals and rocks contain numerous extremely small cavities, partly filled with liquid carbon dioxide. Combined, it occurs in large quantity in the carbonates of lime and magnesia ( 176). Carbon dioxide results from the combustion of carbon in an excess of oxygen and also from the direct decomposition of many salts of carbonic acid (carbonates) by heat : 2NaHCp 3 =Na 2 C0 3 +H 2 O + CO 2 ; CaC0 3 =CaO + C0 2 . Sodium bi- carbonate. Moreover, it is formed when a carbonate is decomposed by an acid: Na 2 C0 3 + 2HC1 = 2NaCl + H 2 + C0 2 . By the action of oxygen at high temperatures all carbon com- pounds are burned with the formation of carbon dioxide. It is also produced by the action of carbon on oxygen compounds, e.g. by heating powdered charcoal with an excess of copper oxide; finally also by the interaction of carbon compounds and oxygen compounds. This latter action is the basis of the general method for determining the proportion of carbon in organic substances; they are heated together with copper oxide and the carbon dioxide formed is absorbed in a weighed amount of caustic potash. Physical Properties. Carbon dioxide at ordinary temperatures and pressures is a gas with a somewhat pungent odor and taste. gp. g. = 1.529 (air=l). It is thus about half again as heavy as air, so that in those places where it comes out of the earth, as in the Dog's Grotto at Naples, it stays in a layer close to the ground and a dog, for instance, is suffocated while a man can breathe with comfort. Carbon dioxide is easily condensed, becoming liquid at under 35 atmospheres pressure. Its critical temperature is 31.35 and its critical pressure 72.9 atm. Liquid carbon dioxide (" liquid carbonic acid ") is manufactured in great quantities and 184.] CARBON DIOXIDE. 253 brought on to the market in steel bottles (bombs). It is a very mobile liquid, which is not miscible with water in all proportions. If the liquid is allowed to escape from the bomb into a coarse linen bag (by inverting the bomb and opening the valve), part of it vaporizes, absorbing hereby so much heat that the remainder solidifies in white flakes. A mixture of this solid carbon dioxide with ether, alcohol or acetone is often used as a freezing-mixture; it enables us to obtain a temperature of 80, and even 140 in vacua. When liquid carbon dioxide is cooled down in a sealed tube, it congeals to an icy mass, which melts at -65. At 15 carbonic acid gas dissolves in its own volume of water (more accurately 1.0020 vol.); at in 1.7967 vol. In alcohol it is still more soluble. Chemical Properties. Carbon dioxide is a very stable com- pound; it is only decomposed by intense heat (see 182) or by the continued action of induction sparks, breaking up into oxygen and carbon monoxide. This decomposition never completes itself, for just so soon as a certain amount of these gases have been formed, they reunite with explosion. At the moment before the explosion the amount of carbon dioxide still present becomes no longer suffi- cient to dilute the mixture of oxygen and monoxide enough to hinder an explosion; the explosive limit is reached. Carbon dioxide cannot be farther oxidized; it is therefore not combustible. In general it cannot support combustion either. There are, however, certain substances that take up oxygen from it when hot; if carbon dioxide is mixed with hydrogen and passed through a red-hot tube, carbon monoxide and water are formed; when led over glowing carbon or when heated with phosphorus it is reduced to carbon monoxide. If a burning magnesium ribbon is lowered into carbon dioxide, the oxide of the metal is formed and free carbon is deposited ; the same thing happens when sodium or potassium is heated in dry carbon dioxide. The aqueous solution of carbon dioxide reacts slightly acid; it is supposed that this solution contains a compound H 2 CO3, of which many salts are known. This acid, carbonic acid, has not yet been isolated in the free state, however, since it gives off gaseous carbon dioxide ("carbonic acid gas") when its solution is boiled or frozen. If its salts (carbonates) are treated with an acid, no H 2 C0 3 is obtained either, for it breaks up forthwith into water and carbon dioxide. Carbonic acid is a very weak acid; it is 254 INORGANIC CHEMISTRY. [ 184- liberated from its salts by almost every other acid. By adding hydrochloric acid to a carbonate H* ions are introduced into the liquid and they unite with the COg" ions to form integral H 2 COs molecules. These, however, break up largely into water and carbon dioxide, the latter of which can only remain in solution up to a certain amount at a constant pressure, so that all in excess of this passes out. As a result the concentration of the H 2 COs molecules cannot exceed a definite and, in this case low limit. Since, however, the ionization of these molecules is very weak, in reality all of the carbonate is decomposed by the strong acid (73). ' The neutral carbonates of the alkalies are soluble in water, giving an alkaline reaction, as a result of partial hydrolysis ( 66). The acid, H 2 CO3, is a weak acid, although its salts, e.g. K 2 CO3, are strong electrolytes. A solution of such a salt, therefore, con- tains a large number of COs" ions, part of which must unite with the H* ions of the water in order to establish the equilibrium be- tween carbonic acid and its ions. The result of this is that other molecules of water must be split up into ions in order to com- pensate the loss of H* ions. This leaves in the liquid a certain number of OH' ions, which are not balanced by an equal number of H" ions. The liquid therefore acquires an alkaline reaction. The carbonates of the other metals are insoluble in water; however, the acid carbonates are mostly soluble. Calcium carbo- nate, e.g., dissolves in water containing carbonic acid. The solu- tions of such acid carbonates give off carbon dioxide on merely boiling, however, and the neutral carbonates are precipitated. In the solid state also the acid carbonates give off carbonic acid gas very readily on warming. Composition of Carbon Dioxide. In connection with what was stated in 179 it is an important fact that no change of volume occurs when carbon burns in an excess of oxygen: C+O 2 =CO 2 . 1 vol. 1 vol. When a very concentrated solution of potassium carbonate is elec- trolyzed with high current density at 30-40, potassium percarbonate, K 2 C 2 O 6 , is formed at the anode. In aqueous solution it sets free iodine from KI solution at once, which serves to distinguish it from H 2 2 , since a dilute solution of the latter liberates iodine only very slowly. 185.] OTHER CARBON COMPOUNDS. 255 Other Carbon Compounds. 185. Cyanogen, (CN) 2 , can be prepared by heating mercuric cyanide, Hg(CN) 2 , or by treating a solution of potassium cyanide with copper sulphate solution. It is possible that first cupric cyanide is formed and that this at once breaks up into cuprous cyanide and cyanogen: Cyanogen has a penetrating odor. When liquefied it boils at 20.7. It is unaffected by high temperatures. It dissolves in water, but the solution deposits amorphous brown flakes after a while. It burns with a purple-,tinged flame according to the equation The reaction, however, is not trimolecular, the first stage being i.e. a bimolecular process. This was proved by DIXON by determining the velocity of propagation of the explosion of mixtures of cyanogen and oxygen. When explosive gas mixtures are introduced into a long tube and their explosion started at one end (by an electric spark, for example) a flame results, which is propagated through the tube with a definite and measurable velocity. BERTHELOT called this self-propagating flame the explosion wave. DIXON ignited a mixture of 1 vol. cyanogen and 1 vol. oxygen, obtaining after the explosion carbon monoxide and nitrogen; the velocity of the explosion wave was found to be 2728 m. per sec. Thereupon he mixed 1 vol. cyanogen with 2 vols. oxygen in one instance and with 1 vol. oxygen and 1 vol. of an indifferent gas in another instance; in both cases the velocity of the explosion wave was nearly the same, viz. 2321 m. and 2398 m. per sec. It is plain, therefore, that the second volume of oxygen influenced the explosion wave in the same way as the indifferent gas, viz. as a diluent. The conclusion may be drawn that in the explosion wave itself only carbon monoxide and nitrogen are formed, even in the presence of an excess of oxygen. However, since the tube contains only carbon dioxide and nitrogen after the combustion, it must be assumed that the combustion of carbon monoxide to carbon dioxide is a secondary process. 256 INORGANIC CHEMISTRY. 185- Hydrogen cyanide, HCN (prussic acid), is important in inor- ganic chemistry because of the numerous complex salts which it forms. Those of the alkalies are soluble in water and crystallize beautifully; see 308. The salts of the alkaline earths and mercuric cyanide are also soluble in water, the other salts in- soluble. The Flame. 186. A flame is produced by the burning of a gas; solids, like iron, carbon, etc., burn without a flame. If a flame is observed during the burning of mineral coal, a candle or the like, it is due FIG. 43. HEVERSE FLAME. FIG. 44. POTASSIUM CHLORATE FLAME. to the fact that at that high temperature gaseous decomposition- products are formed, which burn. If a gas burns in the air, it is called a combustible gas and the oxygen of the air is called the sup- porter of the combustion. These expressions in common use are only relative terms; it is possible to light the oxygen and have it burn with a flame in a gas which is ordinarily called combustible. This phenomenon is illustrated in a way by the reverse flame. This can be easily obtained with the aid of the apparatus of Fig. 43. A lamp-chimney is fitted with a two-hole cork at its lower end. Through 187-] THE FLAME. 257 the narrower hole of the cork a small tube a is inserted for conducting in the gas; through the wider hole a tube b for the admission of air. The chimney is first removed and the gas coming out of tube a lighted and so regulated as to produce a small flame. Then the chimney is replaced; the flame continues to burn quietly, inasmuch as plenty of air io supplied by the wider tube. Thereupon the gas supply is gradu- ally increased and at a certain moment the small flame at the end of a is extinguished and a large pale flame appears at the end of b; it is air burning in the gas which fills the chimney. This is the reverse flame of air in illuminating-gas. At the same time the excess of gas escaping at the top ignites in the outside air, so that the apparatus presents both a direct and a reverse flame at the same time. That it is really air that burns at the mouth of b is proved by introducing a tiny gas-flame by means of the tube c into the flame of the wide tube 6; the small flame continues to burn. Substances that give up oxygen are capable of burning when sur- rounded by a combustible gas. The experiment can be carried out with potassium chlorate as follows: Illuminating-gas is conducted into a glass cylinder (Fig. 44) and lighted at the top, where the cylinder is covered by a thin piece of metal with a hole in it. A little potassium chlorate is then lowered into the flame by means of a deflagrating spoon and heated till oxygen comes off freely. If the bowl is then dipped down in the cylinder, the oxygen burns with a very luminous flame, which is colored violet-blue by the vaporization of some potassium salt. We saw above (27) that a hydrogen flame continues to burn in chlo- rine with the formation of hydrochloric acid; on the other hand chlorine can also be made to bum in hydrogen. For this purpose a cylinder closed at the top is filled with hydrogen and lit at the lower edge. A tube through which chlorine is supplied is then brought in contact with this flame and inserted in the cylinder. The chlorine burns on. 187. A flame may be luminous or non-luminous. It gives light when solid particles are suspended in it. An ordinary gas- flame is luminous because particles of carbon, set free by the com- bustion, are made to glow. On introducing a cold object into the flame they are deposited as soot. The light of the WELSBACH incandescent gas-light is produced by the glowing incombustible mantle ( 291). Such flames give a continous spectrum ( 263). Many gases, which yield only gaseous products on burning, give either a very faint light or none at all, e.g. hydrogen, carbon monoxide, etc. However, when hydrogen burns in oxygen of 20 atmospheres 258 INORGANIC CHEMISTRY. [ IBS- pressure, its flame is strongly luminous. Other incandescent gases, such as the vapors of certain metals ; can render a flame luminous even at ordinary pressure, imparting to it a definite color. Colored flames of this sort give aline spectrum ( 263). A gas-flame, whose light is due to incandescent particles of car- bon, is made non-luminous by mixing the gas with air before the combustion. This is the principle of the BUNSEN burner (Fig. 45), which is used in all laboratories and quite extensively also, with some variation or other, in heating and cooking apparatuses (gas stoves'). The Bunsen burner consists of a base in which is a tube for supplying the gas, which escapes from a narrow orifice at a. Here it mixes with air that enters through the lateral holes in c, the proportion of air being regulated by the collar b. This mixture burns with a colorless flame when ignited at the top of c. The opinion was originally held that the loss of luminosity of the flame is due to the oxygen of the air, the latter causing the complete combustion of the carbon particles. As has since been shown, however, the dilution of the burning gas with nitrogen also has a part in it : if illuminating-gas is mixed with two or three times as much nitrogen, ^^~~ the former burns with a colorless flame. JT IG> 45^ BUNSEN BURNER. With the aid of a wire gauze a burning gas mixture can be cooled so low that the combustion cannot propagate itself through the gauze; in other words, the flame does not get through the gauze (Fig. 46). If gas is allowed to flow out of a BUNSEN burner and a wire gauze is held across the current a short distance from the orifice, the gas can be lit above the gauze without the flame springing back to the burner. It was by experiments such as these that DAVY was led to discover his miner's safety -lamp. As Fig. 47 shows, this consists of an oil-lamp, the flame of which is surrounded by a wire cage. A combustible gas mixture may catch fire inside of the lantern, but the fire cannot pass through the gauze to the outside. 188-] THE FLAME. 259 188. The temperature of the flame is much lower than we might suppose. Since, when hydrogen burns in oxygen, 57.2 kg.-calories are produced by every 18 g. of the mixture, and the specific heat of FIG. 46. EFFECT OF A WIRE GAUZE ON A FLAME. steam is 0.48, this amount of heat ought to raise the 18 g. steam to 57 2 a temperature of ^ ' 6 "66QO. In reality the temperature U.UloX U 4o of the flame does not exceed 2500. This difference between calculation and observation is due to the fact that on account of dissociation only a partial combination of hy- drogen and oxygen takes place in any part of the flame. The temperature of 6600 could indeed be obtained at any point, if the gases united there completely and instantaneously; but this is im- possible, for above 1300 the formation of the com- pound is checked by the opposite process, the dis- sociation of steam. Therefore what occurs must be this: oxygen and hydrogen, when brought to- gether at the aperture, combine and effect a certain rise of temperature; in the same measure as the system in equilibrium (hydrogen, oxygen, steam) FIG. 47. MINER'S cools off, fresh portions of the gases unite. Their SAFETY-LAMP, combustion cannot therefore take place at any particular point but must be gradual throughout the whole extent of the flame and at any one point the temperature cannot ex- ceed a certain limit, which is determined by the degree of dissocia- tion of the combustion product. 260 INORGANIC CHEMISTRY. [ 189- 189. Zones oj a luminous flame. Let us take a candle-flame, for example. In the central zone (1 in the diagram Fig. 48) FIG. 48. ZONES OF A LUMINOUS FLAME. there is no combustion. The stearin of the candle is here con- verted by the heat of the flame into volatile combustible products. In a large candle this can be proved in the manner shown in Fig. 48. The narrow tube conducts off the inflammable gases and they can be lit at the outer end. The hollo wness of a flame can be demonstrated in various ways; in a Bunsen burner, for instance, by placing a match-head in the center, where it does not ignite, or by holding a thin platinum wire across a flame; the wire only glows at the edges of the flame. The dark central zone of the flame is next surrounded by the luminous zone (2) . Here the volatilized hydrocarbon is decom- posed with the separation of carbon, because the air supply is insufficient for complete combustion. These carbon particles become incandescent and so make the flame luminous. Finally there is the blue outer zone (3), in which the glowing particles of carbon are burned by direct contact with the air. It radiates very little light. The amount of solid carbon in a flame which is raised to incandes- cence and hence gives light is very small, as the following calculation shows. The substances in burning illuminating-gas which break up with the liberation of carbon are chiefly benzene and ethylene. The former makes up about 1, the latter about 4, per cent by volume of the gas. If we assume that the benzene is completely broken up and 190.] SILICON. 261 the ethylene only half, then the total amount of carbon deposited by 1 liter of burning illuminating-gas is about 54 mg. The volume of the luminous part of a gas flame with a consumption of 150 liters per hour amounts to about 2 c.c. (reduced to 0), so that the mass of solid 2x54 incandescent carbon present in it is only =0.1 mg. SILICON. 190. This element in combination with oxygen is one of the principal constituents of the earth's crust ( 8). In the free state, however, it does not occur in nature, being found almost exclusively as silica, SiO 2 , or in the silicates. Sand and the many varieties of quartz are different forms of natural silicon dioxide; the number of silicates is very large. . Free silicon is obtained by heating sodium fluosilicate, Na2SiFg, with sodium: Na 2 SiF 6 + 4Na = 6NaF + Si, or by heating sodium in an atmosphere of silicon tetrafluoride: 4Na+SiF 4 =4NaF+Si. The sodium fluoride can be removed by water. Another method, which is far easier, is to mix 400 g. aluminium filings with 500 g. sulphur and 360 g. sand. This mixture is ignited, whereupon it burns with a large flame. The mass fuses and becomes white-hot. When cooled it consists principally of aluminium sulphide and free silicon. It is then treated with dilute hydro- chloric acid, which decomposes and dissolves the sulphide, leaving the silicon behind. (KUHNE method.) Allotropic Forms.- The silicon obtained by the two first-named methods is a brown amorphous powder; it can be fused under a layer of molten sodium chloride and obtained crystal- line on cooling. The latter form is best prepared by KUHNE'S method. The crystals are regular, black, and of a high lustre. If silicon is heated in the electric furnace, it vaporizes and con- denses again in small globules, mixed with a little gray powder and some silica. Chemical Properties. Silicon takes fire only when heated in 262 INORGANIC CHEMISTRY. [ 190- the air to a very high temperature, burning to silica. It unites with fluorine at ordinary temperatures, the combustion being marked by a glow; combination with chlorine takes place on gently warming. At an elevated temperature silicon combines with nitro- gen and some metals ; these silicides have been prepared mainly by MOISSAN in his electric furnace. It is indifferent towards sulphuric, nitric and hydrochloric acids. Hydrofluoric acid dissolves it, however, with the evolution of hydrogen. Hydrogen chloride gas reacts with it at a high tem- perature, forming silicon tetrachloride and sili co-chloroform. It dissolves in a hot solution of sodium or potassium hydroxide, pro- ducing hydrogen and a silicate: Si + 2KOH + H 2 = K 2 Si0 3 + 2H 2 . Hydrogen Silicide, SiH 4 . 191. This gas is obtained by adding freshly prepared magnesium silicide to hydrochloric acid. The magnesium silicide is prepared by heating sand with an excess of magnesium powder, or better by fusing 40 parts of anhydrous magnesium chloride with a mixture of 35 parts of sodium fluosilicate, 10 of sodium chloride and 20 of sodium. The hydrogen silicide so obtained is mixed with hydro- gen. A purer product results from heating an organic derivative of silicon, tri-ethyl silicof ormate : 4SiH(OC 2 H 5 ) 3 = 3Si(OC 2 H 5 )4+SiH 4 . Hydrogen silicide, or silicon tetrahydride, is a gas, which becomes liquid at 1 under a pressure of 100 atmospheres. It has a disagreeable odor. It takes fire in the air; each bubble that escapes from the generator forms a cloudy ring of hydrated silica. If, however the hydrogen silicide is perfectly pure, it does not ignite spontaneously in the air at ordinary temperatures except under reduced pressure. " The spontaneous ignition in the air is caused by the presence of small quantities of other compounds, probably also composed of silicon and hydrogen. We have therefore in this case phenomena similar to those in the case of hydrogen phosphide ( 136). Heat decomposes the hydrogen silicide readily into Si and 2H 2 . It burns in a chlorine atmosphere and is decomposed by an alkali solution according to the equation : SiH 4 + 2KOH + H 2 = 4H 2 + K 2 SiO 3 . 192] HALOGEN COMPOUNDS OF SILICON. 263 Silico-ethane, Si 2 H 6 , is formed by the decomposition of magnesium silicide by hydrochloric acid. It is a gas, which can be liquefied below 7 Halogen Compounds of Silicon. 192. Silicon tetrachloride, SiCl 4 , is prepared by heating silicon in a current of chlorine at 300-310. It is a colorless liquid with the specific gravity 1.5241 at and the boiling-point 59.6. It is instantly decomposed by water, forming hydrochloric acid and hydrated silica. Silico-chloroform, SiCl 3 H, is obtained, together with a large quantity of silicon tetrachloride, on heating silicon in a current of hydrochloric acid gas ( 190). From this mixture it is separated by fractional distilla- tion. It is a colorless, strongly smelling compound which fumes in the air, boils at 34, and is decomposed by water. By the action of dark electrical discharges on a mixture of dry hydrogen and silico-chloroform vapors chlorine-silicon compounds are formed of the order SinCl 2 7i+ 2 , e.g., perchloro-silico-ethane, Si 2 Cl 6 , etc. Silicon tetrafluoride, SiF 4 , can be obtained by warming a mixture of sand and calcium fluoride with concentrated sulphuric acid: 2CaF 2 + Si0 2 + 2H 2 S0 4 = SiF 4 + 2CaS0 4 + 2H 2 0. It is a colorless gas with a very pungent and suffocating odor; it condenses under 9 atm. pressure or by cooling to 160. When perfectly dry, it does not attack glass. Silicon fluoride is also formed by the action of hydrogen fluoride on silicates ; the silica is first set free from them and then attacked in the way just described. Glass-etching ( 53) depends on this action. By the repeated treatment of silicates with hydrous hydrofluon acid all the silicic acid is driven off as silicon fluoride. The bases which were in combination with the silicic acid are left behind in the form of fluorides. They can be transformed into sulphates by warming with sulphuric acid and then converted into a form suitable for analysis. We have here a very useful means of determining the metals present in the silicates. Water decomposes silicon fluoride as follows: 3SiF 4 + 3H 2 = H 2 Si0 3 + 2H 2 SiF 6 . INORGANIC CHEMISTRY. [ 192- The compound H 2 SiF 6 is called hydrofluosilicic acid; it is known only in aqueous solution. If the latter is concentrated by evaporation, silicon tetrafluoride escapes but hydrogen fluoride stays in solution. When the concentration corresponds to 13.3% H^SiFe the vapor contains 2HF to 1SLF4; but dilute solutions yield a vapor which contains much more hydrogen fluoride. If, therefore, a concentrated solution of hydrofluosilicic acid is par- tially evaporated, the residual liquid is able to dissolve silica because of the presence of free hydrofluoric acid. On the other hand, a dilute solution, after partial evaporation, leaves a residue, from which silicic acid is deposited, because the excess of silicon tetra- fluoride which it contains is decomposed by water according to the above equation. The decomposition of silicon fluoride by water is usually demon- strated in the following way: The compound is generated in the pre- scribed manner in a flask (Fig. 49), whereupon it is conducted through FIG. 49. PREPARATION OF HYDROFLUOSILICIC ACID. a doubly-bent glass tube into a cylindrical jar containing a little mer- cury (into which the tube opens) and on top of this. some water. Every bubble of gas that rises from the mercury into the water generates in the latter a cloud of silicic acid. If the glass tube opened directly in water, it would soon become stopped up because of this decomposition. 193.] OXYGEN COMPOUNDS OF SILICON. 265 The solution of hydrofluosilicic acid reacts acid; it dissolves metals with the evolution of hydrogen and behaves in all respects like an acid. A hydrate, H 2 SiF 6 + 2H 2 O ; is known in the solid state. It melts at 19, and is obtained by leading silicon fluoride into con- centrated hydrofluoric acid. Most of the salts of hydrofluosilicic acid are soluble in water; the potassium salt is difficultly so, how- ever, and the barium salt is insoluble. Hydrofluosilicic acid is used in hardening objects made of gyp- sum (this is due probably to the formation of calcium fluoride) and also in analytical chemistry. Oxygen Compounds of Silicon. 193. Only one such compound is known: silicon dioxide, or silica. SILICA, Si0 2 . This compound occurs in astonishingly large quantities and in a great number of varieties in the solid crust of the earth. It is found crystallized as rock crystal, quartz (when colored brown, called smoky quartz), amethyst (the more beautiful sorts being used for ornament), tridymite, onyx, cat's-eye, etc. Sand is largely silica; sandstone also belongs here and so does jasper (usually colored red with ferric oxide and having a conchoidal fracture). Opal is an amorphous variety, containing varying amounts of water. Silica can be prepared artificially as an amorphous white powder by heating silicic acid. Physical Properties. In the crystallized state silica is very hard and insoluble in water and has a specific gravity of 2.6. It is very difficultly fusible; in the oxyhydrogen flame it softens and passes over into a vitreous modification. When heated strongly this can be drawn out into extremely fine threads that are so tenacious and display so regular a torsion that they are frequently used in sus- pending magnets, etc., in physical instruments. It can be made to boil vigorously in the electric furnace; the vapor condenses in woolly flakes. Quartz that has been fused has a very small coefficient of expansion (about Vi7 of that of platinum); this explains why objects made of it can endure very sudden changes of temperature. They can be heated very hot and then thrust 266 INORGANIC CHEMISTRY. [ 193.. into cold water at once without cracking. They are attacked only by metallic oxides and at a high temperature. Recently it has become possible to utilize fused (vitreous) quartz for the manu- facture of chemical apparatus. It is interesting that quartz vessels are transparent to ultraviolet rays, which is not the case with glass vessels. Chemical Properties. Especially in the crystallized condition silica is very little acted upon by acids except hydrofluoric acid ( 193). Fused alkalies dissolve it, forming alkali silicates. It can be reduced by carbon in the electric furnace, carborundum ( 178) being formed. It is also reduced by heating with magnesium ( 190). Silicic Acids. 194. When a solution of potassium or sodium silicate (water' glass] is treated with hydrochloric acid, a very voluminous, gelat- inous mass separates out; this consists of hydrous silicic acid cor- responding to the general formula SiC^aq. After being washed with water and dried in the air it is a fine white amorphous powder of the approximate composition H 2 Si03. Freshly precipitated silicic acid is slightly soluble in water, but more so in dilute hydro- chloric acid If, therefore, water-glass is introduced into an excess of hydrochloric acid, the silicic acid stays in solution; it can be sepa- rated from the sodium chloride simultaneously formed, by the fol- lowing process : The solution is put into a piece of parchment tubing, which is tied at both ends, and the whole submerged in pure water, the latter being frequently renewed. It is found that the salt goes through the parchment, but that the silicic acid does not. This process is called dialysis and any arrangement for carrying it out is known as a dialyzer. GRAHAM found that crystallizable substances in solution (crystalloids) are able to pass through such a membrane, while other substances, which he called colloids, are not. In the latter class are glue, gums, gelatine, albumen in short, many amorphous substances occurring in the animal and vegetable kingdoms. The silicic acid which separates from the colloidal solution dries in the air to a white amorphous powder, still containing a 195.] SILICIC ACIDS. 267 good deal of water, however. The water can be slowly extracted in a sulphuric acid desiccator. Since silicon tetrachloride is changed to silicic acid by water, just like phosphorus pentachloride to phosphoric acid ( 145), we can consider the compound as the basis from which the remaining silicic acids are derived. The latter can in general be represented by the formula mSi(OH) 4 nH 2 0. These polysilicic acids themselves have not been isolated, but many of their salts and double-salts are known, which occur as minerals in nature. The silicates of potassium and sodium are soluble in water, those of the other metals insoluble, as are also most of the double silicates of the alkalies. In the soil hydrous silicates are found whose bases are usually lime and alumina. In contact with alkali salts these undergo a double decom- position, an insoluble potassium aluminium silicate, for example, being formed together with chloride of calcium, which is taken off by the under- ground water. This phenomenon is said to be caused by the absorptive power of the soil; it plays an important role in the determination of soil- values. It is this that holds back the potash, an invaluable nutrient, which is furnished to the soil in the form of potassium salts and would otherwise be quickly washed off by the rain because of its solubility. The soluble phosphates are "absorbed*' by the soil in the same way. This is mainly to be ascribed to the lime they contain, with which insoluble tri- or dicalcium phosphate is formed; to some extent this ab- sorption may be caused also by basic lime silicates. Silicon Compounds of Other Elements. 195. Silicon sulphide, SiS 2 , is produced when carbon disulphide vapor is led over a mixture of charcoal and silica at red heat. It forms long, silken needles, which are broken up by water into SiO 2 aq and hydrogen sulphide. ; Silicon nitride, Si 2 N 3 , a white amorphous substance, results from the heating of silicon in an atmosphere of nitrogen. (For metal silicides cf. 190.) 268 INORGANIC CHEMISTRY. [ COLLOIDS. 196. In silicic acid we have become acquainted with a sub- stance that occurs in a special form, viz., as a colloid. A con- siderable number of such substances is now known, and the study of them has latterly been so active and prolific that a brief recapitulation of the principal results is fitting at this point. GRAHAM discovered that in aqueous solution a number of substances, principally amorphous materials, such as the glues, albumen and dextrin, have a very small power of diffusion, quite contrary to most salts. Accordingly he distinguished between colloids and crystalloids. Subsequent investigations served to increase greatly the number of colloids, i.e., substances of small diffusibility. Gradually the view developed that the colloidal condition is not something peculiar to certain com- pounds, but that all sorts of substances, even the crystalloids, can be obtained colloidal by suitable treatment. Hence the colloidal state . is now regarded as a general property of matter. Just as we have substances in the solid, liquid and gaseous states, so we can also transform them into the colloidal state. The question that arises first is: How may this condition be brought about? The following methods serve the purpose: In the first place, colloids may be prepared by simply dis- solving certain substances, such as glue, in water. Secondly, they are formed in many cases instead of pre- cipitates, when no ions are present. For example, if hydrogen sulphide is passed into a solution of arsenic trioxide, there results, instead of a precipitate of As 2 Ss, a yellow liquid containing the arsenic sulphide in colloidal solution. However, if the arsenic trioxide solution is first acidified with a little hydrochloric acid (a highly ionized substance), the As2Ss separates out as a yellow precipitate. Again, we may take mercuric cyanide, a compound that is hardly ionized at all in aqueous solution. If a solution of it is treated with hydrogen sulphide, which is also a very feebly ionized substance, the mercuric sulphide that is formed is retained in colloidal solution; yet, the usual precipitate can be obtained by adding previously a small amount of a strong mineral acid. 190.] COLLOIDS. 269 A third way of preparing colloids is by dialysis, a process described in connection with silicic acid. In this way hydrosols (see below) of ferric oxide, aluminium oxide and many other substances can be obtained. Ferric oxide hydrosol, for instance, is formed when ferric chloride, FeCls, is dissolved in water, and just a little less ammonia added than would produce a precipitate, and the whole then dialyzed. The ammonium chloride, NH 4 C1, and hydrochloric acid (resulting from a partial hydrolysis of FeCls in aqueous solution) pass through the membrane, while Fe^Os aq. remains inside in colloidal solution. A fourth method is the comminution, or dusting, of metals under water. This is accomplished by connecting wires or rods of platinum, gold and other metals with the poles of a 110-volt circuit; if the wires are moved toward each other under water, a small arc is formed when they are a short distance apart, and dark clouds of the metal proceed out into the liquid from the cathode. The liquid is then filtered; the coarser bits of metal remain on the filter and the filtrate is a clear, dark-colored solu- tion containing the metal as hydrosol. Metals can often be converted into the colloidal state by treat- ing a very dilute solution of one of their salts with certain reduc- ing-agents at ordinary temperature. Thus from a very dilute gold chloride solution the colloidal gold can be prepared by the addition of phenylhydrazine hydrochloride or acetylene. Finally, it is worthy of note that by means of protective colloids many substances can be obtained colloidal when other means fail. For instance, if a silver nitrate solution and a potassium bromide solution, each containing about 1% of gelatine, are mixed together, the silver bromide is not precipitated, but comes out colloidal. " Collargol," a therapeutic preparation, is a silver colloid, made stable by a protective colloid. Colloids can be divided into two groups, reversible and irreversible. The reversible colloids comprise- among other sub- stances the agglutinants , as they are called, gelatine, agar-agar, albumins, starch, etc. When they are mixed with water they swell up and on being gently warmed form a solution. When cooled they gelatinize, i.e., they congeal to a soft, viscous mass which retains all the solvent water. The solution itself is called 270 INORGANIC CHEMISTRY. [ 196. a hydrosol (or, in case alcohol is the solvent, an alcosol) and the gelatinized mass a hydro gel. When the solvent water is extracted from a reversible colloid by evaporation at a low temperature a hydrogel is at first formed, which still contains a great deal of water. This water is partially lost on exposure to the air. More rapidly in a desiccator, and its vapor tension does not differ perceptibly from that of pure water. When, however, a certain stage of dehydration is reached the vapor tension begins to dimmish. If water is added to the hydrogel before this stage is reached, a hydrogel is again obtained with the same properties as originally. The process of s o 1- and g e 1-formation is thus a reversible one. Other interesting properties are attached to hydrosols. Crystalloid salts, for example, diffuse in them, even in the con- gealed mass, almost as easily as in water. If a piece of jellied agar-agar is immersed for some time in a dark blue ammoniacal solution of a copper salt, the agar-agar becomes stained through- out its entire mass. Colloids, on the contrary, do not diffuse. This can be shown by a colloidal solution of Prussian blue, which does not penetrate at all into the agar-agar, as above. The electrical conductance, too, is practically the same for a gel containing crystalloid salts in solution as for an aqueous solu- tion of the same salts at like concentration. Oftentimes large amounts of crystalloid salts can be added to a reversible hydrosol without the formation of gel. This is very different with the second class of colloids, the irreversible colloids, for they are in many cases very sensitive to additions of salts. When the salt is added the irreversible hydrosol begins to appear cloudy, and a precipitate is formed which cannot be reconverted offhand into hydrosol. Irreversible hydrosols can be prepared in the various ways already mentioned. They comprise the colloidal metals, sul- phides, hydrated oxides, etc. Most of them are mobile liquids, in contrast to many reversible colloids, such as glue. The quantity of an electrolyte that is just sufficient to pre- cipitate an irreversible hydrosol is connected with the valence of the electrolyte, the quantity decreasing rapidly with in- creasing valence. The As 2 S 3 hydrosol is just coagulated by 196.] COLLOIDS. 271 71 mill-equivalents of NaCl per liter, 2.0 of MgCl 2 and 0.39 of A1C1 3 . Certain irreversible colloids are capable of mutually pre- cipitating each other; others are not. The hydrosols of ferric oxide and arsenious sulphide give a precipitate when mixed, but the hydrosols of gold and arsenious sulphide mix without precipitation. These phenomena have been shown to be connected with the behavior of the substances toward the electric current. If a solution is introduced into a U-tube supplied with electrodes at the upper ends and a strong current (say 110 volts) is passed through it, the colloid is seen to separate out and wander either to the anode or to the cathode. At one of the two electrodes an aqueous layer appears, which is entirely free from colloid and is separated sharply from the hydrosol. This convective trans- ference, or electrical endosmose, is by no means to be confused with the ionic migration in electrolytes. For, while in the elec- trolytes there is an electrical opposition between the dissocia- tion products of the dissolved substance, the electrical opposi- tion exists in this case between the colloid and the solvent. In general, mutual precipitation is only possible with colloids whose electrical charges are opposite with respect to that of a common solvent. The colloidal state must be regarded as a very fine divi- sion, or distribution, of one substance in another. This follows from the great analogy which exists between suspensions (e.g., clay and water) and colloids. Both can be separated out by centrifuging and both display the TYNDALL effect. This effect may be described as follows: When a beam of light passes through a body of air that is free from dust it is invisible transversely; the gas is " optically a vacuum/' But, so soon as dust particles enter the air, the path of the beam can be followed through the dispersion of the light by the particles. Optically vacuous liquids and optically vacuous solutions of crystalloid salts can also be prepared. But if a beam of light is passed through a hydrosol the path of the beam can be seen. The hydrosol is therefore not an optical vacuum; it must con- tain floating particles, but these are so small that they can not be seen even with the best microscopes. 272 INORGANIC CHEMISTRY. [ 196- However, SIEDENTOPF and ZSIGMOXDY have succeeded in rendering these sub microscopic particles visible with an apparatus that they call the ultramicroscope. In it the hydrosol is illuminated transversely so that the luminous rays do not blind the eye of the observer. The submicroscopic particles bend (dif- fract) the light rays " in all directions, so that with sufficiently intense illumination the light effect produced by each individual particle comes within the range of microscopic visibility and can be separately observed without however revealing its form. Suspensions resemble colloids further in that they exhibit electrical endosmose and can be precipitated by the addition of electrolytes. When a liquid is distributed through another in exceedingly gmall drops we have an emulsion. The most familiar example is milk, an emulsion of butter fat. There is reason for assuming that many reversible colloids are extremely fine emulsions; for instance, an emulsion, like a gelatine solution, cannot be coag- ulated by the addition of an electrolyte. The knowledge that in the colloidal state we have to do with a very fine distribution of one substance in another has led to the introduction of a new set of terms. The substance distributed as a colloid is now generally spoken of as the disperse phase, distributed in the dispersion medium. Further, the words dis- persoids and emulsoids are replacing the word "colloids." It was formerly thought that a sharp distinction must be drawn between colloids and real solutions. GRAHAM spoke of two different worlds of matter. In contrast to the true solu- tions colloids exhibit practically no diffusion, no vapor pressure lowering, no boiling-point elevation or freezing-point depres- sion, in short, no osmotic phenomena. The researches of recent years have, however, shown that essential differences do not really exist. To begin with, we have come upon many cases of transition between colloidal and real solutions. Furthermore, it was previously observed by LOBRY DE BRUYN that salt solu- tions can be separated by centrifugal force into portions of unlike concentration. More important still, the investigations of EINSTEIN, PERRIN, SVEDBERG and others have shown that colloids, just like true solutions, are subject to the osmotic laws. 197.] GERMANIUM. 273 From the molecular-kinetic point of view there is no difference, according to these investigations, between a " dissolved molecule ' and a " suspended particle"; consequently a mechanical sus- pension must exert exactly the same osmotic pressure as a " true solution " of the same number of particles per unit volume. The fact that the colloidal solutions display no properties cor- responding to osmotic pressure is simply due to the fact that at the same concentration the number of freely moving particles in solutions is enormously greater than with the colloids, i.e., an individual colloid particle has gigantic dimensions as compared with those of a molecule. This makes the freezing-point lower- ing, etc., so slight that it cannot be measured by present exper- imental means. In order to test the applicability of the osmotic laws to colloidal solutions we are therefore forced to employ indirect methods. The methods employed are associated with four phenomena: (1) the translatory and rotatory movements of the particles. (2) diffusion ; (3) the change of concentration under the influence of gravity; and (4) the local temporary changes of concentration. The possibility of testing the osmotic laws by such measurements is a result of developing formulae for these phenomena that are deduced upon the assumption that the osmotic laws are applic- able. These investigations also serve to corroborate the reality of atoms and molecules, supporting the information gained in many other ways, as has been set forth in 35. GERMANIUM. 197. This element is of extremely rare occurrence. It was discovered by WINKLER in an argentiferous mineral, argyrodite, GeS 2 -4Ag 2 S, found in Freiberg, in Saxony. Germanium forms grayish-white octahedrons with a metallic lustre and a specific gravity of 5.469 at 20. It melts at 900. At ordinary temperatures it is unaffected by the air; at red heat it burns, forming white fumes of germanium oxide, GeO 2 . Two series of compounds of this element are known, which are derived from the oxides GeO and GeO 2 ; the ous compounds are easily oxidized to the higher form, germanic acid. The hydrogen compounds, GeH 4 and GeHCl 3 , are known. Germanic chloride, GeCl 4 , can be prepared directly from the elements. It is broken up by water forming Ge(OH) 4 . 274 INORGANIC CHEMISTRY. [ 197- Germanium dioxide, GeO 2 , is produced by heating the corresponding hydroxide, or by roasting the element or its sulphide or by treating it with nitric acid. It is a white powder of a specific gravity of 4.703 at 18 and is unaffected by heat. Germanium disulphide, GeS 2 , separates as a white precipitate when hydrogen sulphide is passed into the solution of germanium dioxide in strong hydrochloric acid. In moist air it decomposes, giving off hydrogen sulphide. It dissolves in alkalies and alkali sulphides to form sulpho-salts. For germanium cf. also 218. TIN. 198. This metal is not very widely distributed on the earth; in some places, however, it is found in quite large quantities. The principal tin mines of Europe are those in Cornwall; even the Phoenicians obtained tin there. The most important present locali- ties are on the group of islands lying east of Sumatra (Banca, Bil- liton, Sinkop, etc.). There the metal occurs in the form of tin- stone (cassiterite, SnC^); it is found in quadratic crystals, which are usually colored brown or black by a small amount of iron. In order to extract the metal, the ore is at first roasted, to eliminate any sulphur or arsenic it may contain, and then reduced with car- bon. The tin thus obtained is refined by liquation, i.e. by fusing again at a low temperature and pouring it off from the less fusible alloy of tin with iron and arsenic. It is then melted once more and stirred with a wooden pole (branch of a tree), whereby the oxide still remaining is reduced. The Banca tin is nearly chemic- ally pure. Physical Properties. Tin is a silvery-white metal, melting at 232.7 and volatilizing between 1450 and 1600. Sp. g. = 7.293 at 13. It has a crystalline structure which can be made visible by moistening with hydrochloric acid, whereupon peculiar frost- like etch-figures are produced on the surface (tin-moiree). When tin is bent, a characteristic crackling sound (cry of tin) is heard, which is probably caused by the grating of the crystal faces on each other. Tin is very malleable and ductile; it can be beaten into very thin leaves (tin-foil) at the ordinary temperature, and at 100 it can be drawn out into wire. At a very low temperature and in contact with an alcoholic pink-salt solution ( 201), tin passes spontaneously into another modification, gray tin, which has a lower specific gravity, 198.J TIN. 275 5.8. Above 20 this form changes back to white tin. If the latter is brought in contact with gray tin at ordinary temperatures (below + 20), it turns very slowly into gray tin, falling to powder, probably because of the increase in volume (this phenomenon is called the "tin-disease")- If it is not in contact with the gray modification, the transformation does not take place at all at ordinary temperatures, or at least not for centuries. Evidently there is a transition point of the two forms at 20, and we are forced to the odd conclusion that, except on warm summer days, tin is in the metastable condition. The reason why tin, even in contact with gray modification, passes so slowly into that form at ordinary temperatures is that the velocity of transformation is small in the neighborhood of the transition point; it is accelerated on moving away from that point. When the temperature sinks this acceleration is counteracted, however, by the retardation that all reactions undergo by a lower- ing of temperature. In many cases, therefore, there must be a maximum of the velocity of transformation, such as we have here at 48; below that temperature the transformation again becomes slower. Ordinary tin crystallizes in the tetragonal system. In addition to tlie gray modification there is also a third one, the rhombic modification. The transition point tetragonal ^rhombic is about 170. This point was determined in a unique way, namely, by measuring the velocity of flow of the metal under high pressure. For this purpose the solid metal was placed in a cylinder having a hole in the bottom, and the quantity of metal was measured that was forced out under constant pres- sure in the unit of time. In general, this quantity increases rapidly with rising temperature, but with tin it was found to diminish considerably when the temperature reached about 200. This may be taken as a proof that the metal has another (third) modification. At 200 tin is so brittle that it can be easily pulverized. Chemical Properties. Tin is unaffected by the air at ordinary temperatures; if heated strongly, it burns with an intense white light to tin oxide, Sn0 2 . Hydrochloric acid dissolves it, forming stannous chloride and hydrogen. It is also attacked by nitric acid ( 201). A boiling solution of caustic soda or potash converts it into a stannic acid salt (s t a n n a t e) with the evolution of hydro- gen: Sn + 2KOH + H 2 O = K 2 SnO 3 + 2H 2 . 276 INORGANIC CHEMISTRY. [ 199- In the presence of weak acids (acetic acid) and alkalies it is very stable. 199. Uses. On account of its permanence tin is used as a pro- tective covering for metals which are attacked by the air and the above-named agencies. Many kitchen utensils are " tinned." Sheet iron is covered with a layer of tin, to protect it from rusting ( 279), and is then known as tin-plate, or sheet-tin. This is done by simply dipping the sheet iron, which has been cleaned by hydrochloric or sulphuric acid, in molten tin. Many alloys of tin are in use. Solder consists of tin and lead (in the ratio 2:1 or 1:1 or 1:2), and is harder than either of its components but more easily fusible. The alloys of c o p p e r and t i n are called bronzes; their composition varies according to the purpose they serve. At present the bronzes usually contain a little lead and zinc as well. Bronze is hard and tough, can be easily worked and fuses to a mobile liquid, hence it is particularly suitable for casting. Gun metal contains 90% copper and 10% tin; bell metal 20-25% tin, the rest being copper. Phosphor bronze is prepared by fusing copper with tin phosphide ( 202). The result- ing mass is remarkably homogeneous and contains 0.25-2.5% phosphorus and 5-15% tin. Its great hardness and firmness render it especially valuable for certain parts of machines (axle-bearings). Silicon bronze contains silicon in place of phosphorus, is very hard and conducts electricity well, hence it is used for making telephone wire. Tin amalgam forms the metallic coating of mirrors. Compounds of Tin. Tin forms two sets of compounds; they correspond to the oxygen compounds, stannous oxide, SnO, and stannic oxide, Sn0 2 . STANNOUS COMPOUNDS. 200. Stannous chloride, SnCl2, is prepared by dissolving tin in hydrochloric acid: It crystallizes with two molecules of water, which are given off at 100. : - It is very readily soluble in water (1 part in 0.37 at ordinary temperatures). Anhydrous stannous chloride is white and trans- parent; it melts at 250 and boils at 606. A little above the 200.] STAXXOUS COMPOUNDS. , 277 boiling-point the vapor density corresponds to the formula Sn 2 CLi; above 900, however, to SnCl 2 . The aqueous solution acts strongly reducing. It absorbs oxy- gen from the air with the partial formation of basic chloride (a white powder), if the liquid is not too acidic: 3SnCl 2 + H 2 + = SnCl 4 + 2Sn(OH)Cl. Basic chloride. But if the liquid is strongly acid the tetrachloride SnCU is also formed in this oxidation. This same basic chloride also results from hydrolytic dissociation, when a neutral stannous chloride solution is strongly diluted. SnCl 2 +aq =Sn(OH)Cl + HC1 +aq. The reducing power of stannous chloride is further seen in its action on potassium permanganate, potassium dichromate, cupric chloride, mercuric chloride, etc., all of which are converted into lower stages of oxidation in acid solution. It may be remarked here that, from the ionic point of view, oxidation amounts in many cases to raising an ion to a higher positive potential, and reduction to the reverse. Let us consider, for instance, the reaction between stannous chloride and mercuric chloride. This can be expressed by the equation SnCl 2 + HgCl 2 = SnCl 4 + Hg. Stannous chloride is oxidized to stannic chloride; at the same time mercuric chloride is " reduced " to the metal. Written in ions, this equation becomes that is, the electrical charge of the mercury ion is taken by the bivalent tin ion, the former losing its electrification. Another example is the action of chlorine on stannous chloride, by which the latter is "oxidized" to stannic chloride: = SnCl 4 . The ionic reaction is 278 INORGANIC CHEMISTRY. [ 200- Tin takes up two more positive charges, but this necessitates that the two Cl atoms become ions; they thus require two negative charges; but when these are formed two positive charges are ob- tained at the same time. However, the Sn*"' and Cl' ions unite to form stannic chloride, SnCU, which is a very weak electrolyte (cf. 201). In the preparation of chlorine, hydrochloric acid is " oxidized " by manganese dioxide : Mn0 2 + 4HC1 = MnCl 2 + 2H 2 + C1 2 , or the positive charge of the four H' ions is thus transferred, half to the manganese and the rest serving to discharge two chlorine ions, i.e. to equalize their negative charges. Various double salts of stannous chloride are known, e.g. SnCl 2 2KC1 ; SnCl 2 2NH 4 C1. Stannous hydroxide, Sn(OH) 2 is precipitated when a solution of stannous chloride is treated with soda : SnCl 2 +Na 2 C0 3 +H 2 O = Sn(OH) 2 +2NaCl+C0 2 . This hydroxide is insoluble in ammonia, but soluble in alkalies; when the latter solution is boiled, tin is deposited and alkali stan- nate, e.g. K 2 SnO 3 , formed. The hydroxide is also soluble in acids, thus displaying a basic as well as an acidic nature. Such compounds are able to give hydroxyl ions (Sn"+2OH / ) on the one hand and hydrogen ions (Sn0 2 " + 2H') on the other. They are termed amphoteric compounds. Stannous oxide is obtained by heating the hydroxide in a cur- rent of carbon dioxide; it is a dark-brown powder, which takes fire in the air, burning to stannic oxide, Sn0 2 . Other salts of stannous oxide than the above-mentioned stannous chloride are also known. The sulphate, for instance, is obtained by dis- solving the hydroxide or the metal in dilute sulphuric acid. It forms a basic salt readily. Stannous sulphide, SnS, is precipitated as an amorphous brown powder when hydrogen sulphide is passed into the solution of stan- nous salts. It is insoluble in potassium sulphide, K 2 S, but it 201.] STANNIC COMPOUNDS. 279 dissolves to form a sulpho-stannate when brought in contact with the p o 1 y sulphide of ammonium or potassium, K 2 S x (x=2-5). SnS + K 2 S 2 = K 2 SnS 3 . Stannous sulphide can also be prepared by fusing tin with sul- phur. It then forms a bluish-gray crystalline mass. STANNIC COMPOUNDS. 201. Stannic chloride, SnCl 4 , was prepared as early as 1605. It was named spiritus jumans Libami, after its discoverer. It is obtained by the action of chlorine on tin or stannous chloride. Stannic chloride is a liquid which fumes strongly in the air; it boils at 113.9, and has a specific gravity of 2.234 at 15. When brought in contact with a little water or on taking up moisture from the air, it goes over into a semi-solid, crystallized mass, SnCl 4 -3H 2 O, the so-called tin-butter. A fresh solution of stannic chloride is a very poor conductor of electricity. However, the con- ductivity increases slowly at ordinary, faster at higher, tempera- tures; after several days -it reaches a maximum. In the case of more dilute solutions this maximum is much higher. These facts can be explained by assuming that stannic chloride is but feebly ionized and that it reacts with water in the following way; SnCl 4 +4H 2 <=> Sn(OH) 4 +4HCl; in other words, that it undergoes hydrolytic dissociation. It is the liberated hydrochloric acid that causes the conductivity. The solution contains tin hydroxide in the colloidal state. The water has thus split up the stannic chloride into a basic hydroxide and an acid. Stannic chloride forms well-crystallized double salts with the alkali chlorides, e.g. SnCl 4 -2KCl and SnCl 4 -2NH 4 Cl. The latter is known as pink salt (because of its color) and is used as a mor- dant in dyeing. Tin tetrachloride also unites with the chlorides of the metalloids to form crystallized substances, e.g. SnCU-PCls; SnCl 4 - POC1 3 ; SnCl 4 SC1 4 , etc. It combines with hydrochloric acid, forming a leafy-crystalline mass, H 2 SnCl 6 6H 2 0, which melts at 9. Tin fluoride, SnF 4 , itself is not known, but there is a compound, K 2 SnF 6 , which corresponds to potassium fluo-silicate; the salts of hydro- fluostannic acid are isomorphous with the analogous silicon compounds. 280 INORGANIC CHEMISTRY. [201- Stannic oxide, SnO 2 , can be prepared synthetically by heating tin in air. It is an amorphous white powder, insoluble in acids and al- kalies; the latter, however, dissolve it when f used, forayngstannates. Stannic Acid and Metastannic Acid. The -hydroxides corre- sponding to SnO2 have only very weakly basic properties; here the acidic properties are prominent. The normal hydroxide, Sn(OH) 4 , is unknown, but there is a hydroxide of the empirical composi- tion H 2 SnO3( Sn(OH) 4 H 2 O), corresponding to carbonic acid, H 2 CO 3 . Strangely enough this exists in two modifications, which differ from each other both chemically and physically; they are called stannic and metastannic acids. The stannic acid is precipitated when ammonia is added to an aqueous solution of stannic chloride or hydrochloric acid to a potas- sium stannate solution. This precipitate reacts acid when moist and is soluble in concentrated hydrochloric and nitric acids, as well as in alkalies. It gradually changes into metastannic acid. Metastannic acid is generally prepared by treating tin with strong nitric acid; it is then formed in a vigorous reaction as a dense white powder. Metastannic acid is insoluble in sodium hydroxide, but nevertheless unites with it to form sodium metastan- nate ; this is dissolved by water, although with difficulty, but is insolu- ble in the caustic soda solution. When boiled with hydrochloric acid, metastannic acid goes over into a chloride, which is insoluble in the concentrated acid but soluble in water. This solution does not contain the ordinary tin chloride, but another one, meta-tin chloride, having, however, the same composition, SnCl 4 . It is distinguished from the ordinary stannic chloride by giving a yellow coloration with stannous chloride solution; the solution of the ordi- nary chloride does not do this till after some time, during which the metachloride is formed in it. Stannic acid and the corresponding chloride thus pass over into the meta-compounds spontaneously; on the other hand, metastan- nic acid can be converted into the ordinary tin compounds by boil- ing it for some time or fusing it with a caustic alkali. The difference between stannic and metastannic acids was pointed out by BERZELIUS as early as the beginning of the nineteenth century. They are both colloids. The salts of metastannic acid have in general a very com- plicated composition, similar to the polysilicates ( 195), for which reason metastannic acid is regarded as a polymer of the ordinary stannic acid, i.e., that its molecule is represented by (H 2 SnO 3 )z, stannic acid itself being H 2 SnO 3 . 203.] LEAD. 281 Of the salts of stannic acid, the sodium stannate, Na 2 SnO 3 + 3H 2 O, is especially well known. It comes on the market under the name of " preparing-salt " and is used as a mordant in dyeing. It is made by fusing tin-stone with caustic soda and crystallizes in hexagonal crystals, which are more soluble in cold than in warm water. Purple of Cassius is obtained when a mixture of the hydrosols of tin dioxide and gold is precipitated by adding some such electrolyte as am- monium chloride. This mode of formation proves that the substance is not a compound of the two components, as was formerly believed, but only a mixed gel. 202. Stannic sulphide, SnS 2 , falls out as a yellow amorphous powder, when hydrogen sulphide is passed into the acid solution of a stannic compound. It can be synthesized by heating tin amalgam with sulphur and ammonium chloride, being thus obtained in the form of transparent golden leaves and known as aurum musivum, or mosaic gold; it is used for gilding. Stannic sulphide is a sulpho- anhydride; the corresponding sulpho-acid, H^SnSs, is not known in the free state, but exists in the form of salts. Sodium sulphostannate, Na 2 SnSs + 2H 2 O, crystallizes in color- less octahedrons. When its solution is treated with an acid, stan- nic sulphide is precipitated. Tin phosphide serves, as was stated above, for the manufacture of phosphor bronze. Of the various compounds of tin and phosphorus, the best known is the compound Sn 9 P. It forms a coarsely crystalline mass, which melts at 170. LEAD. 203. Among the lead ores the most important is qalenite (PbS) ; it occurs in isometric crystals (cubes) of a graphitic color. Other ores are cerussite (PbCOs), crocoite (PbCrO^, wulfenite (PbMoO 4 ), etc. For the extraction of the metal galenite is used almost exclu- sively. This is roasted to convert the sulphide partially into oxide, and partially into sulphate: PbS + 3O = PbO + SO 2 ; PbS + 2O 2 = PbSO 4 . In roasting care is taken that a considerable portion of the ore remains as sulphide. On farther heating, the latter reacts with the oxygen compounds in the following way: = 3Pb+SO 2 ; and PbSO 4 +PbS=2Pb+2S0 2 . 282 INORGANIC CHEMISTRY. [ 203 - Physical Properties. Lead is a soft ductile metal of a bluish color. On exposure to the air it loses its lustre rapidly, becoming coated with a very thin layer of oxide. It has a specific gravity of 11.254, melts at 327, and boils at 1525. Chemical Properties. The thin coating formed by the oxide on the brilliant surface of the metal protects the lead from further attack by the air. If, however, it is prepared in a very finely divided state, e.g., by heating lead tartrate or citrate in the absence of air, it takes fire in the air even at ordinary temperatures. (Other metals can be reduced in a similar way to a fine state of division, whereupon they ignite spontaneously in the air. A substance which exhibits this phenomenon is called a pyrophorus.) When lead is melted, it becomes coated with red oxide of lead; by constantly removing the latter, the lead can be entirely oxidized. A compact mass is unaffected by sulphuric or hydrochloric acid, but, when finely divided, it reacts to form the corresponding salts. Nitric acid easily dissolves it to form the nitrate. Acetic acid and various vegetable acids attack it; since all lead salts are very piosonous and very serious effects result from chronic poisoning with insignificant but successive amounts, it is not admissible to use tin containing lead in tin-plating vessels for use in the kitchen. Zinc and iron precipitate the metal from solutions. A piece of zinc becomes covered with a dendritic crystalline mass (" lead- tree "). This reaction can be expressed by: Zn+Pb"=Zn" i.e. zinc is changed into the ionic condition, and the lead ions are discharged. How it comes about that one metal thus assumes the electrical charge of another may be explained by a hypothesis of NERNST. His supposition is that every metal on coming in con- tact with water or a solution tends to send positive ions into it. This emission of ions continues until the positive charge acquired by the solution and the negative charge created on the metal balance by their mutual attraction the tension (called the electrolytic solution-tension) with which the ions are driven into the solution. This tension differs considerably for different metals; for zinc it is much greater than for lead. When, therefore, a strip of zinc is dipped in a lead solution it forces zinc ions into the solution and the zinc thus becomes much more negatively charged than would 204.] OXIDES OF LEAD. 283 a piece of lead by the emission of lead ions. The lead ions are therefore attracted by the zinc and discharged, i.e. lead is precipi- tated from the solution. This process stops only when all the lead of the solution has been replaced by zinc. Distilled water, from which the air has been entirely removed by boiling, has no effect on lead, but the simultaneous action of air and water produce lead hydroxide, which is somewhat soluble in water. This hydroxide is converted into insoluble basic carbonate by carbonic acid. From a hygienic standpoint these properties of lead are of vast impor- tance, because drinking-water is almost universally conducted through pipes made of lead or material containing lead ("compo-pipes"). The absorption of lead from such pipes by water and the continuation of the process depends in a large measure on the proportion of salt in the water. As a rule, the less of salts it contains, the more lead it takes up. Rain-water, which is almost entirely free from solid matter, but contains oxygen, carbon dioxide and traces of ammonia, is, therefore, most likely to dissolve lead. The lead eave-troughs, etc., which were once extensively used, should, therefore, be rejected, in case the rain- water is used for drinking. Well-water usually contains acid calcium carbonate and gypsum; as a result, the lead pipes soon become coated .with an insoluble layer of lead sulphate and basic carbonate (as well as calcium carbonate), so that after a while the lead can no longer be absorbed by the water. Lead is used for many purposes, not only in the elemental con- dition, but also in the form of alloys (see 199). Oxides of Lead. 204. The following oxides of lead are known: Pb2O, PbO, Pb 2 O 3 , Pb 3 4 , Pb0 2 . Lead oxide, PbO, is the only one of these oxides with basic prop- erties. It is formed by direct synthesis from its elements ( 203). It is fusible, and congeals again to a reddish-yellow mass called litharge. By carefully heating lead, lead hydroxide or lead nitrate, the oxide is obtained as an amorphous brown powder (massicot). It is somewhat soluble in water, forming the hydroxide. It dis- solves in caustic potash, and crystallizes out in rhombic prisms on cooling. 284 INORGANIC CHEMISTRY. [204- Lead hydroxide, Pb(OH) 2 , is formed by precipitating a lead solution with an alkali. It is amphoteric, since it is soluble in caustic alkalies; ammonia, however, does not dissolve it. On being warmed to 145 it gives up water and turns to oxide. It is somewhat soluble in water, imparting to the latter an alkaline reaction, and absorbs carbon dioxide from the air. Minium, or red lead, Pb 3 O4, is prepared by heating lead oxide or white lead in the air for quite a while at 300-400. Because of its pleasing red color it is used as a pigment in painting. Gentle heating makes the color a brighter red at first; stronger heating turns it violet and finally black; on cooling, however, the original color returns. By treating it with dilute nitric acid, lead nitrate and lead peroxide are formed, hence minium may be regarded as 2PbO-PbO 2 . Lead peroxide, PbO 2 , is obtained in the way just stated; more easily, however, by passing chlorine into an alkaline lead solution or adding a hypochlorite to a lead salt, thus: 2PbCl 2 + Ca(OCl) 2 + 2H 2 O = 2PbO 2 + CaCl 2 + 4HC1. Milk of lime is then added to neutralize the free acid. Lead peroxide is an amorphous dark-brown powder. It has the property, common to most peroxides, of giving up oxygen easily. At an elevated temperature it splits up into lead oxide and oxygen. On warming it -with sulphuric acid, lead sulphate and oxygen are formed; on warming with hydrochloric acid, lead chloride and chlorine are produced. Lead peroxide, like the oxides C0 2 and SnO 2 , has the character of an acid anhydride; it is soluble in hot concentrated potassium hydroxide and this solution, on cooling, deposits crystals of the com- position K 2 PbOa + 3H 2 O (which are thus entirely analogous in com- position to potassium stannate) . This plumbate is easily decomposed by water into potassium hydroxide and lead peroxide. If we regard lead peroxide as an acid anhydride, minium can be considered as the lead salt of the normal plumbic acid, Pb(OH) 4 , i.e. Pb 2 -Pb04. This idea is confirmed by the following method of formation: If a solution of lead oxide in potassium hydroxide is added to a solution of the plumbate K 2 PbOs, a yellow substance is precipitated having 205.] HALOGEN COMPOUNDS. 285 the composition PbsC^-H^O, which gives off water readily and forms minium. If a mixture of litharge and calcium carbonate is heated in a current of air at 700, carbon dioxide is given off, oxygen absorbed and cal- cium plumbate, CaPbO 3 , formed. If this plumbate is treated with carbon dioxide at about the same temperature, calcium carbonate and lead oxide are again formed, while oxygen escapes. This process (discovered by KASSNER) serves for the commercial manufacture of oxygen. The latter is brought on the market compressed in iron bottles (c/. also 262). The oxide Pb 2 O 3 is obtained by adding sodium hypochlorite to a solution of lead oxide in potassium hydroxide. It can be regarded as the lead salt of a lead acid, H 2 PbO3, i.e. as Pb-PbOs, for, on treatment with dilute nitric acid PbO2 (the anhydride of and lead nitrate are formed. Halogen Compounds. 205. The halogen compounds of lead having the formula are difficultly soluble in cold water; lead fluoride is almost insoluble and the solubility of the three others decreases with increasing atomic weight of the halogen. Lead chloride, PbCl2, is obtained as a white precipitate when dilute hydrochloric acid is added to the solution of a lead salt. At 12.5 it dissolves in 135, at 100 in less than 30, parts of water and crystallizes from the hot solution in the form of white silky needles or lamellse. If an aqueous solution of lead chloride is treated with dilute hydrochloric acid, lead chloride is precipitated, for by the addition of Cl-ions the solubility product of lead chloride is exceeded ; nevertheless, lead chloride is easily soluble in concentrated hydrochloric acid. This must be due to the formation of a com- pound of the chlorides of lead and hydrogen, an analogue of which has been found in PbI 2 -HI + 10H 2 O, which has been isolated. A characteristic compound of lead is the iodide PbI 2 , which is pre- cipitated from lead solutions by potassium iodide. It is scarcely soluble in cold, but moderately soluble in hot water. It crystallizes out of a solution in dilute acetic acid in beautiful crystal flakes with a golden lustre. 2S6 INORGANIC CHEMISTRY. [206- Lead tetrachloride, PbCl 4 , is formed when a solution of lead dichloride in strong hydrochloric acid, cooled by ice 2 is saturated with chlorine. From this liquid ammonium chloride precipitates a lemon-yellow crys- talline substance, 2NH 4 C1 PbCl 4 , having a composition analogous to pink-salt ( 201). Analogous double salts of PbCl 4 are also formed with the alkali metals, such as potassium and rubidium (Rb 2 PbCl 6 ). If one of these double salts is treated with ice-cold concentrated sulphuric acid, lead tetrachloride gradually separates out as a heavy yellow pil (sp. g. 3.18), which is stable at a low temperature and crystallizes at 15. At as high a temperature as room temperature, and more rapidly on warming, it breaks up into lead dichloride and chlorine. Other Lead Salts. 206. Lead nitrate, Pb(N0 3 )2, is obtained by dissolving lead in dilute nitric acid. It is colorless, crystallizes isometric and is soluble in 8 parts of water. Heating decomposes it ( 122). Several basic lead nitrates are known Lead sulphate, PbSCU, is practically insoluble in water and can therefore be obtained by precipitating a lead solution with dilute sulphuric acid or a soluble sulphate. It occurs as a mineral in crystallized form under the name of anglesite; it is isomorphous with the sulphate of barium, barite. Lead sulphate is soluble in concentrated sulphuric acid; hence the crude acid which is con- centrated in lead pans ( 186, 3) contains lead sulphate; this is precipitated on diluting the acid with water. It is dissolved by concentrated alkalies. Ignition on charcoal reduces it to sulphide. Lead disulphate, " plumbic sulphate," Pb(SO 4 ) 2 , separates from the acid around the anode when sulphuric acid of 1.7-1.8 specific gravity is elect roly zed between lead electrodes. It has not been obtained quite free from lead sulphate. It is a white granular sub- stance of strong oxidizing properties. Water decomposes it readily into sulphuric acid and lead peroxide. It is isomeric with lead persulphate, PbS2Og, a salt of the dibasic persulphuric acid. Lead carbonate, PbC0 3 , is deposited when a solution of the nitrate is treated with ammonium carbonate. White lead, a basic carbonate, is used extensively as a pigment. However, it soon turns black, if any hydrogen sulphide (from drainage pipes, etc.) comes in contact with it; moreover, it is injurious to the health, because it comes off of the painted walls in the form of dust and 207.] SUMMARY OF THE CARBON GROUP. 287 gets into the lungs. White lead is particularly valuable for its covering-power, i.e. the painted surface appears perfectly white when covered with only a very thin layer of the pigment; it is much greater than that of other white pigments, such as white zinc and barite, which are frequently substituted for white lead because they are harmless. The manufacture of white lead is still carried on exten- sively after the Dutch method. This consists in placing rolls or "buckles" of lead-plate into jars containing a little acetic acid. The vessels are loosely covered with a leaden lid and buried in a heap of horse-manure. The heat generated by the decaying manure causes a part of the acetic acid to evaporate and converts the lead into basic lead acetate. The latter is then transformed to white lead by the car- bon dioxide given off from the decaying heap. After about five or six weeks the plates are almost entirely changed to white lead. This is then ground moist, washed out (to remove any acetate) and dried whereupon it is sent to the market. Lead sulphide, PbS, is black and comes down amorphous when hydrogen sulphide is passed into a lead solution. A liquid con- taining only traces of lead is colored brown by sulphuretted hy- drogen; this is a very delicate means of testing for lead. Strong nitric acid oxidizes it readily to lead sulphate. SUMMARY OF THE CARBON GROUP. 207. The elements carbon, silicon, germanium, tin and lead form a natural group, as may be seen from a comparison of their physical and chemical properties. In the following table the most important physical constants are summarized ; here, as in other natural groups, the gradual change of these constants with the rise of the atomic weight is very evident: C. Si. Ge. Sn. Pb. Atomic weight. 12.00 28.3 72.5 119.0 207.10 Specific gravity. f 2.251 1 3.6 ) 2.49 5.5 7.29 11.39 Melting-point. . Boiling-point. . . above 3000 1420 circa 900 233 circa 1500 327 1525 288 INORGANIC CHEMISTRY. [207- With respect to the chemical properties we note in the first place that all these elements have the same compound-types, MX2 and MX 4 ; in other words, that they are bi- or quadri-valent; this is even true of lead (PbO 2 , PbCl 4 , etc.), which does not fit into the table of physical properties with its boiling- and melting- points. Moreover, there is to be noted in general a transition from metalloid to metallic character, as is plainly shown by the following facts: 1. Only carbon and silicon are known to form hydrogen com- pounds (of an indifferent nature). 2. Of the oxygen compounds of the MO type, that of carbon is indifferent and the others (no such compound of silicon is known) grow more ba^ic in character as the atomic weight increases, lead hydroxide having rather strongly alkaline properties. 3. The oxygen compounds, RO 2 , however, are decidedly acidic in character in the cases of carbon and silicon and also in the case of germanium, while in that of lead the salts of the acid H 2 PbO3 are immediately decomposed by water, so that here the acid proper- ties appear much weakened. 4. As to the halogen compounds, those of carbon (CX 4 ) are unaffected by cold water perhaps because of their insolubility in it; the other halogen compounds, MX 4 , are decomposed by water. Lead, in some of its physical and chemical properties, does not dis- play the gradation which is ordinarily met with in the elements of a, group. This phenomenon is quite often observed in elements of very high atomic weight. In the nitrogen group we saw it in the case of bismuth. METHODS OF DETERMINING ATOMIC WEIGHTS. 208. So far only one method of determining the atomic weight has been mentioned ( 34). This consists in investigating as large as possible a number of gaseous compounds of the element as to their vapor density and empirical composition and then calculating how many grams of the element are contained in a mole of the various compounds. The smallest figure thus found is taken as the atomic weight. Although this method is quite general, 208.J METHODS OF DETERMINING ATOMIC WEIGHTS. 289 it has the drawback of affording only a certain degree of proba- bility, a probability which becomes greater as the number of in- vestigated compounds increases and which lessens the chance of finding a compound that contains per mole only a simple fraction of the previously accepted atomic weight. There are, however, other methods. None of them is so generally applicable as this, but they are of a more absolute char- acter and have been of great value in the many cases in which they could be used. They furnish a very valuable check on the determinations made by the general method. These methods are based on the following laws : 1. The law of DULONG and PETIT. The product of the atomic weight of a s oli d element and its specific heat is about 6.4. This is evident from the table on the opposite page. Most of the values of the product lie as the table shows, very close to 6.4; the maximum is 6.9, the minimum 5.0. Calling this product the atomic heat, we can express the law of DULONG and PETIT in the following simple way : The atomic heat of the solid ele- ments is approximately constant and is about 6.4. A few deviations have been pointed out in the last column; in such cases it has frequently been found, however, that at an increased temperature the atomic heat approaches the value 6.4. This is probably due to the fact that at the temperature (room- temperature) at which the measurements of the specific heat of the elements have been mainly carried out, the elements are not all in the proper physical condition for comparison. It is notably the elements with atomic weights under 35 that show the greatest deviations. It is an interesting fact that there is a certain regularity to be found in these irregularities. The latter become more marked as the valence increases. Element ..... Li Valence ...... 1 Atomic heat.. 6. 6 Be B 23 3.7 2.8 C 4 1.9 Na 1 6.7 Mg 2 6.1 Al 3 5.8 Si 4 4.6 P S 3(5) 2(4,6) 5.9 5.7 It is easy to see how the law of DULONG amd PETIT can be made use of for the determination of atomic weights. Inverting, we have - - =At. wt. b. H. 290 INORGANIC CHEMISTRY. [ 209- Element. Sp. H. At. Wt. Product. Remarks. Hydrogen. Lithium. . , Beryllium. Boron. ^ , f amorphous. Carbon diamond. Sodium. . . . Magnesium. Aluminium. Silicon. . Phosphorus. Sulphur. Potassium. . . Calcium Scandium. . . Chromium. . . Manganese. . . Iron Cobalt Nickel Copper Zinc Gallium Germanium. . Arsenic Selenium. . . . Bromine Zirconium. . . Molybdenum. Ruthenium. . Rhodium. . . . Palladium. . . Silver Cadmium. . . . Indium Tin. Antimony. . . Tellurium. . . , Iodine Lanthanum. . Cerium Tungsten. . . Osmium Iridium Platinum. . . . Gold Mercury Thallium. . . . Lead Bismuth Thorium. . . . Uranium. . . . 6 0.941 0.408 0.254 0.174 143 0.293 0.250 0.214 0.165 0.189 0.178 0.166 0.170 0.153 0.121 0.122 0.114 0.107 0.108 0.095 0.094 0.079 0.077 0.082 0.080 0.084 0.066 0.072 0.061 0.058 0.059 0.057 0.054 0.057 0.054 0.051 0.047 0.054 0.045 0.045 0.033 0.031 0.032 0.032 0.032 0.032 0.033 0.031 0.030 0.027 0.027 1.008 6.94 9.1 11.0 I 12. 00 23.00 24.32 27.1 28.3 31.04 32.07 39.10 40.09 44.1 52.0 54.93 55.85 58.97 58.68 63.57 65.37 69.9 72.5 74.96 79.2 79.92 90.6 96.0 101.7 102.9 106.7 107.88 112.40 114.8 119.0 120.2 127.5 126.92 139.0 140.25 184.0 190.9 193.1 195.2 197.2 200.0 204.0 207.10 208.0 232.4 238.5 6 6.6 3.7 2.8 2.1 1.7 6.7 6.1 5.8 4.6 5.9 5.7 6.5 6.8 6.7 6.3 6.7 6.4 6.3 6.4 6.0 6.1 5.5 5.6 6.9 6.3 6.7 6.0 6.9 6.3 6.0 6.0 6.1 6.0 6.5 6.5 6.1 6.0 6.8 6.2 6.3 6.1 5.9 6.1 6.2 6.3 6.4 6.7 6.4 6.2 6.2 6.5 Liquid. Sp. H. at 257 =0.58; prod. = 5.2. Amorphous. Sp. H. at 400 = 0.58; prod. =6.4. Above 900 Sp. H.= 0.459; prod. =5.5. Crystallized. Sp. H. above 200 = 0.204; prod. =5.8. Yellow. Sp. H. of redP- 0.1698; prod. =5.24. Rhombic. Crystallized Do. Solid. Solid 209.] METHODS OF DETERMINING ATOMIC WEIGHTS. 291 Of course the result thus obtained is only approximately correct, for the product 6.4 is not strictly constant. The method is, however, reliable enough to determine what multiple of the equivalent weight ( 22), the exact value of which can be found by analysis, is the atomic weight. 209. 2. Closely connected with the law just enunciated is that of NEUMANN", which has been more carefully investigated by REG- NAULT and KOPP. This law says that in solid compounds each element has a constant atomic heat, which varies but little from that of the free element. The molecular heat is therefore equal to the sum of the atomic heats. If the molecular heat of solid com- pounds is divided by the number of atoms, the quotient must be about 6.4. In reality this quotient proved to be: for bromine compounds RBr, 6.9; for RBr 2 , 6.5; for iodine compounds RI, 6.7; RI2, 6.5. The law of NEUMANN likewise holds for many elements whose specific heat in the solid state has not been susceptible of measurement, thus e.g. for chlorine compounds: for RC1 com- pounds the quotient referred to was 6.4, for RC12, 6.2, for certain double chlorides, 6.1-6.2. For other elements, like oxygen, the atomic heat found from the molecular heat of the compounds is constant, but it is about 4.0 instead of about 6.4. The same is true of hydrogen, whose mean atomic heat in solid compounds is 2.3. These figures were found by determining the molecular heat of various oxygen or hydrogen compounds and subtracting from it the known atomic heats of the other elements. If the atomic heat thus obtained is divided by the atomic weight, we have the specific heat of the element in its compounds. The way in which the law of NEUMANN can be applied to atomic weight determinations is illustrated by the following example: The problem is to determine the atomic weight of calcium with the help of the specific heat of sulphate of lime, CaS0 4 , which amounts to 0.2 according to REGNAULT. Analysis has shown that anhydrous calcium sulphate contains 29.4% calcium, 23.5% sulphur, and 47.1% oxygen. Since sulphates contain the SO 4 group in combination with a metal, it follows from the above .analysis that there must be associated with this group 40 parts by weight of calcium. The next question is, whether 40 is really the 292 INORGANIC CHEMISTRY. [ 209- atomic weight of calcium or a multiple or submultiple of the atomic weight. Now the molecular weight of calcium sulphate must be 40 + 32 +4 X 16 = 136, no matter whether 40 is the relative weight of one or of more than one calcium atom. The molecular heat is therefore 136 X 0.2 =27.2. The atomic heat of sulphur in compounds is about 5.4 and that of oxygen about 4.0; consequently the molecular heat of the SO 4 group in its solid compounds is 5.4+4X4.0 = 21.4. For the atomic heat of calcium we have the remainder, 27.2-21.4 = 5.8. It therefore follows that the formula of calcium sulphate must be CaS0 4 , which means that 40 is the atomic weight of calcium; for, if the atomic weight were a multiple or submultiple of this number, we should have found for the atomic heat of the metal a number much farther from the average atomic heat of the elements, 6.4, than 5.8. The value of the atomic weight calculated from NEUMANN'S law therefore serves merely to decide what multiple of the equiva- lent weight must be taken ; for this purpose the number so obtained is sufficiently accurate. 210. 3. The law of Mitscherlich. The crystal form of com- pounds having analogous chemical composition is the same: or, in other words, compounds of analogous chemical composition are is omor phous . The compounds KC1, KI, KBr, e.g. are analogous in composition; they all crystallize in cubes. H 2 KP04, H 2 KAsC>4, H 2 (NH 4 )P04 also have an analogous composition and all crystallize in the tetragonal system. The analogous compounds KC104 and KMn(>4 both crystallize rhombic. If two compounds have been proved to be isomorphous, it is very probable that their composition is analogous, whereupon the atomic weight is readily found. Let us, for example, take the case of manganese, supposing its atomic weight to be unknown; now potassium permanganate is isomorphous with potassium per- chlorate, which latter is known to have the formula KC104. Analysis has shown the formula of potassium permanganate to be KMn x 04, x being unknown, for 39 parts (by weight) of potassium (1 atom) are combined with 64 parts of oxygen (4 atoms) and 55 parts of manganese (x atoms). From its isomorphism with KCICU it follows that its formula must be KMn(J4 (i.e. x = l), hence 55 is the atomic weight of manganese. 211.] DETERMINATION OF EQUIVALENT WEIGHTS. 293 In determining the atomic weight of zinc we could use the iso- morphism of the crystallized sulphates of magnesium and zinc. The formula of the former is MgSO 4 4-7H 2 O. On the basis' of the analysis of zinc sulphate and the isomorphism mentioned we have the formula ZnS04+7H 2 O, from which the atomic weight is ob- tained in the same way as above. The law of isomorphism was discovered as early as 1819. Since at that time the law of AVOGADRO received little attention and the determination of the specific heat was in many cases impossible, the phenomena of iso- morphism were the most important means of getting information regarding the value of the atomic weight. Subsequently its importance for this pur- pose lessened, mainly because simpler means were found, but also because it proved to be very difficult in many cases to decide whether two sub- stances are isomorphous. Moreover it was found that certain substances of entirely different composition are isomorphous. A very delicate test for isomorphism is the fact that a supersaturated solution can be made to crystallize, not only by an extremely small amount of the dissolved substance itself (" sowing," or " inoculation "), but by bodies that are isomorphous with it. Experimental Determination of Equivalent Weights. 2ii. In the methods above described the question is one of determining "which multiple of the equivalent weight is the atomic weight. In order to establish the atomic weights with accuracy the equivalent weights must be determined with the greatest possible precision. The solution of this problem, which is one of fundamental importance, since all the numerical relationships of chemical reactions are based on the atomic weights, has been the object of numerous investigations in the preceding century and to-day it is still only partially accomplished. The first atomic weight table dates from DALTON in 1805. The figures given in it were scarcely more than rough approximations. BERZELIUS (1779-1848) in the first and second decades of the century determined a long series of equivalent numbers, after having been first obliged in most cases to work out reliable analytical methods. The atomic weights at which he arrived were in general use for many years and really differ from the more accurate ones now employed by hardly more than a fraction of a per cent. Exceedingly accurate "atomic weight determinations" were undertaken by STAS (1813-1891). The ten atomic weights determined by him, viz. those of Ag, Cl, Br, I, K, Na, Li, S, Pb, and N, are in most cases accurate to within a few units in the second decimal place. The researches called for most exhausting and persistent labors during a long period of years. 294 INORGANIC CHEMISTRY. [211- In the last decade atomic weight determinations have been carried out on a scale of much greater refinement by MORLEY, RICHARDS, GUYB and others and the accuracy of the values has been extended another decimal place, so that now not a few of the atomic weights are established with certainty to within a few units in the third decimal place. In determining atomic weights either purely chemical or physico- chemical methods may be employed. Both have been greatly perfected in these latter investigations and they will now be described in a few para- graphs. As for the purely chemical methods, there are four conditions which are essential to an accurate determination of an atomic weight: (a) A suitable substance must be found which can be prepared perfectly pure. (6) This compound must contain in addition to the element under study only elements of accurately known atomic weight, (c) The valence of the elements in this compound must be well denned. It is not permissible, for example, that the substance be a mixture of two stages of oxidation. (d) The compound selected must be adapted to an exact analysis, or else its exact synthesis from the weighed elements must be possible. Notwithstanding the simplicity and legitimacy of these demands it is often difficult to satisfy them. The preparation of a compound in the pure state is among the most difficult of operations, if by purity we mean the reduction of the impurities to a 10 ~ 4 part of the whole. It was formerly believed that this could be readity accomplished by recrystallization, but now we know that every substance that separates out in a solid phase has a tendency to retain upon its surface or in its interior a part of the other substance contained in the phase out of which the solid separated. All precipitates or crystals from aqueous solutions contain water that is not in chemical combination. Even the splendid glistening silver crystals that are obtained in the electrolysis of a silver nitrate solution and are apparently perfectly dry and pure contain not only water but silver nitrate as well. Silver chloride, precipitated from a solution of sodium chloride by silver nitrate, may have included traces of NaCl, AgNO 3 or NaNO 3 , even after a thorough washing. Potassium chlorate, though much less soluble than potassium chloride, contains nevertheless 0.027% of the latter after repeated recrystallizations. One of the most troublesome sources of error in all quantitative researches is the unsuspected presence of hygroscoDically held water, since it is not at all easy to detect by chemical tests and causes no essential change in the external appearance of the substance containing it. The analysis of a substance resolves itself in most cases into a separation cf its components in the form of other compounds and weighings of the latter. For example, in order to determine the silver content of silver nitrate the metal is thrown down as silver chloride and the latter is weighed, whereupon the quantity of silver can be calculated from the known silver content of the chloride. The analyst generally finds it also necessary to convert one compound into another quantitatively. The modern investiga- 212.] DETERMINATION OF EQUIVALENT WEIGHTS. 295 tions of atomic weights have also taught us that this is often a very difficult problem. Among other sources of error in this connection are the solu- bility of the so-called " insoluble" substances and the solubility of glass. It has long been known that substances like silver chloride, barium sul- phate, etc., are not strictly insoluble; but their solubility has first received proper attention in connection with the recent atomic weight determinations. In working with glass vessels it is impossible to avoid silicic acid as an impurity. Recognition of this fact has led to the use of vessels of quartz or, better still, of platinum, which has proved to be an important refinement of method. 212. Physico-chemical methods have found application in the determination of the volume weight of gases. One of the most fruitful of modern physical concepts is that of the ideal gas, whose expansion at constant pressure or pressure increase at constant volume both have a coefficient for a tempera- ture change of one degree of exactly 1/273.08. Moreover, the ideal gas is in strict accord with BOYLE'S law. A gram molecule of such a gas at 6 and 760 mm. Hg pressure would occupy a volume of 22.412 1. However, the actual gases are more compressible and expansible than the ideal gas; hydrogen and helium are the only ones that are less compressible. For this reason 22.412 1. of an actual gas at and 760 mm. Hg contains a little more than one gram molecule. If we let 1 + A represent the number of gram molecules of an actual gas which are contained in 22.412 1., the molecular weight of the gas becomes 22.412 G where G is the weight of a liter of the gas under normal conditions. The establishment of the atomic weight of a gas thus resolves itself into the accurate determination of the magnitudes G and A. The methods for ascertaining the exact weight of a given volume of a gas have undergone important improvements in recent years. The agreement of the values found by the different investigators is within 0.0001. While formerly the gases were weighed in huge globes, some containing as much as 21 1., later investigators have been able to reduce this volume to between one liter and half a liter, or even less. Nevertheless the concordance between the various series of determinations was improved, because the corrections for the small globes were much less. An additional correction was applied for the contraction of a globe on evacuation, due to the external pressure of the atmosphere reducing the volume slightly; the buoyancy effect of the air is somewhat less for an evacuated globe than for one filled with gas. In order to remove completely the layer of air that has been condensed on the inner surface of the globe, it is necessary to evacuate the latter repeatedly to as low a vacuum as possible and to fill it with the gas whose density is to be determined, great care being taken meanwhile to exclude the air. Furthermore, the purification of the gases to be weighed is much better 296 INORGANIC CHEMISTRY. [ 212- accomplished by first liquefying them and then removing the impurities by fractional distillation at a lo\? temperature. ' The determination of the quantity A can be accomplished in four different ways, which are found described in the larger physics manuals. It should be. noted, however, that they have not yet attained the exactness that characterizes the methods of determining G. In the cases of the less easily condensed gases, like H, N, O, and Cl, however, very accurate determinations have already been made. From an experimental standpoint these physico-chemical methods have a decided superiority over the purely chemical methods in that ph^Gica' measurements only are carried out after the gas has been obtaiivxi pure, All the uncertainties that are involved in chemical transformations are thus avoided; and upon such transformations every purely chemical determi- nation of an atomic weight is based. THE PERIODIC SYSTEM OF THE ELEMENTS. 213. In studying the elements which we have considered so far, we have found that they can be arranged into groups of elements according to their valence, the elements of each group showing great similarity in the types of their compounds. The physical and chemical properties of the elements of such a group are found to change progressively as the atomic weight increases. The question now arises whether all elements can be thus arranged into groups; the reply is affirmative. In the course of the last century there was no lack of attempts to arrange the elements into groups of similar elements. DOEBEREINER called atten- tion to a simple relation between the atomic weights of kindred elements as early as 1817, and in 1829 he presented the doctrine of triads, i.e. he showed that there are different groups of three elements each, which have a great similarity among themselves and a constant difference in the atomic weights, e.g. Cl, Br, I; Ca, Sr, Ba t etc. In the year 1865 the law of octaves was proposed by NEWLANDS, he having discovered that, if the elements are arranged according to increasing atomic weight, after an interval of seven elements an element follows which has properties analogous to those of the first, i.e. the first, eighth, fifteenth, etc., are similar. In 1869 MENDELEEFF and LOTHAR MEYER almost simultaneously reached con- clusions which are comprehended by the term " periodic system." If we arrange the elements according to increasing atomic weight, thus: H 1 Li 7 Be 9.1 B 11 C 12 N 14 O 16 F 19 Na23 Mg24 Al 27.1 Si 28.4 P 31.0 S 32 Cl 35.4 213.] THE PERIODIC SYSTEM OF THE ELEMENTS. 297 we see that there is a gradual variation in the properties of elements in a horizontal line; after fluorine, however, a small increase in the atomic weight involves a sudden change of properties. Moreover those elements which are in the same vertical column show great similarity, as we saw above in the cases of carbon and silicon, nitrogen and phosphorus, etc. This regular change makes itself evident in the valence toward oxygen, which rises from one (with Li and Na) to two (Be, Mg), hree (B, Al), four (C, Si), five (N, P), six (S) and seven (Cl in C1 2 O 7 ). The valence toward' hydrogen v or a 'halogen increases, however, from one (Li) to four (C) and then falls again to one (F). A similar regular change is to be observed with reference to the physical properties, e.g. specific gravity and atomic volume. Na MR Al Si P (red) S Cl (liq.) Sp. gravity 0.97 1.75 2.67 2.49 2.14 2.06 1.33 AT. volume 24 14 10 11 14 16 27 by atomic volume we understand the atomic weight divided by the weight of the unit volume (based on water of 4 as 1); it is therefore tLe number of cubic centimeters occupied by a gram-atom. Here we observe an increase of the specific gravity up to alu- minium, then again a decrease to chlorine, while the atomic vol- ume, on the other hand, decreases from the beginning of the series to aluminium and then increases. This steady change of the same physical properties is also observed in the compounds of the above elements. For the oxides, e.g. we have: Na 2 O M-rO Al 2 Oa SiO 2 P 2 O 5 SO 3 Cl 2 Or Sp. gravity 2.8 3.7 4.0 2.6 2.7 1.9 ? At. volume 22 22 25 45 55 82 ? Moreover, if we write down a series of elements according to increasing atomic weight, beginning with another univalent metal, we discover irregularities of exactly the same sort as the above. The following series may serve as an example of this: Ag Cd In Sn Sb Te I Atomic wt 107.8 112.4 115 119.0 120.2 127.6 127.0 Sp. gravity 10.5 8.6 7.4 7.2 6.7 6.2 4.9 Heie also we find the same gradual rise of valence from silver, which is univalent, to septivalent iodine, the progressive transi- 298 INORGANIC CHEMISTRY. [ 213. tion from metal to metalloid and a continuous decrease in specific gravity. But more; if we put this last row under the first two: H 1 Li 7 Be 9.1 B 11 C 12 N 14 0. 16 F 19 Na 23.1 Mg 24.4 Al 27.1 Si 28.4 P 31.0 S 32.1 Cl 35.5 Ag 107.9 Cd 112.4 In 115 Sn 119.0 Sb 120.2 Te 127.6 I 127.0, it is apparent that the elements in the same vertical columns belong to a group. This has been demonstrated for the last four columns in preceding chapters; it will be proved for the others later on. In the light of these facts, we are led to the conclusion that the physical and chemical properties of the elements are junctions of their atomic weights; and when we consider the series beginning with lithium and sodium, and note that in each instance, after a difference of about 16, there follows another element with corre- sponding properties, we are led to the supposition that these properties are periodic functions of the atomic weights. By a function we understand in general a dependent relation between two or more magnitudes, of such a sort that, when one changes, the other does likewise. In the equations y=ax; y=ax; y=x n , etc., y is a function of x. A periodic function requires that the same value appear for one magnitude in regular intervals as the other magnitude steadily increases. An example of this kind is presented by the gonio- metric functions, as y=sm x, etc., for every time x increases by 2?r, y comes to have the same value again. If we desire to substantiate the conclusion just stated, we shall have to investigate first the length of each period, in other words, determine how many elements intervene in the table, according to increasing atomic weight between two with analogous properties. It has already been shown that for the elements as far as chlo- rine, a period always includes seven elements. After chlorine comes potassium (39), which thus falls into the column under sodium. The following elements, K39.2 Ca40.1 Sc44.1 T148.1 V51.2 O52.1 Mn55.0, correspond very well with the preceding series, Na23.1 Mg24.4 A127.1 Si28.4 P31.0 S32.1 C135.5- 213.] THE PERIODIC SYSTEM OF THE ELEMENTS. 299 at least so far as the valence and the form of the compound are concerned (A1 2 3 and Sc 2 O 3 , TiO 2 and SiO 2 , K 2 CrO 4 and K 2 SO 4 , KMnCU and KCICU), although the similarity of these elements in other respects is not very marked. The elements following manganese, viz., Fe 55.9, Co 59.0, Ni 58.7, however, do not fit in at all under K, Ca, Sc; but if we pass these by there follows another series of seven elements, which corresponds to the one beginning with potassium: Cu 63.6 Zn 65.4 Ga 70 Ge 72.5 As 75.0 Se 79.2 Br 80.0. We therefore reach the conclusion that, after the first two periods of seven elements ending with chlorine must come one of seventeen elements (two of seven each, and three elements placed at the side), if the elements in the same vertical column are to correspond in their properties. This large period of seventeen elements can, therefore, be arranged under the preceding small period of seven elements in the following way: SMALL PERIOD. Na23.1 Mg24.4 A127.1 Si28.4 P31.0 S32.1 C135.5 LARGE PERIOD. K39.2 Ca40.1 Sc44.1 Ti48.1 V51.2 Cr52.1 Mn55.0 Fe55.9 Co59.0 M58.7 Cu63.6 Zn65.4 Ga70 Ga72.5 As75.0 Se792 BrSO.O. In order to arrange in periods the elements whose atomic weights exceed eighty, it is again necessary to assume large periods, and, moreover, to leave several places vacant. In this manner we arrive at the scheme known as MENDELEEFF'S table (see p. 301). As to the position of hydrogen in this table opinions are divided. MENDELEEFF placed this element in the first group, above lithium; its chemical properties indicate without doubt that it belongs with these j metals. On the other hand, ORME MASSON has presented arguments for placing it at the head of group VII, as is done in this table. These arguments are as follows: (1) The molecule of hydrogen contains two atoms, as does a halogen molecule, while the molecule of an alkali metal consists of one atom. (2) The very low boiling-point of hydrogen indicates a similarity to the halogens; moreover the boiling-points of the alkali metals fall with 300 INORGANIC CHEMISTRY. [215. increasing atomic weight. (3) The difference between the atomic weights of the elements of a horizontal series is, on the average, 3. By placing hydrogen in group VII it differs by 3 from the next element, helium; but it is then also in good agreement with fluorine, for the mean difference in atomic weight between the successive elements of a column is 16. The difference F -H =18, while Li -H =6, i.e., there is no analogy in the latter case. (4) Liquid and solid hydrogen have no metallic properties. (5) The most important argument for placing hydrogen in the first group is based on its relation to the acids, which may be regarded as salts of hydrogen. But in organic compounds chlorine can replace hydrogen without essentially altering the nature of the substance. This "organic" argument thus offsets the ''inorganic" one, based on the analogy of acids and salts. As to the elements discovered in the atmosphere since 1894, viz., helium (at. wt. 4), neon f20), argon (40 , krypton (81.6) and xenon (128), it is clear that they form a natural group, for their properties display great analogy (see 110). Since they are not able to form compounds with other elements, they can be regarded as nullivalent. In that case their group could find a place after the eighth, or before the first, group (compare the table, page 294), thus forming a bridge, or transition, from the strongest electro-negative, to the strongest electro-positive, elements. However, it must be noted that argon with an atomic weight of 40 precedes potassium with an atomic weight of 39. As may be seen from Plate I, their atomic volumes fit into LOTHAB MEYER'S curve very well. Group VIII, as has been said, owes its origin to the setting at the side of the elements included in it, for by this means the corresponding elements of groups I-VII could be brought under each other. It will thus be of importance to the system, if the nine elements of this group display so much analogy to each other that the grouping of them together appears actually justified. Now this is really the case, as is seen from the following study of their properties : 1. All these elements are of a gray color and difficultly fusi- ble; indeed, osmium is one of the hardest of all metals to fuse (2500); iridium melts at 1950, wrought iron at 1500, etc. The melting-point of iron is higher than that of cobalt, and the latter higher than that of nickel. A similar fall of this constant is found with ruthenium, rhodium and palladium, and also with osmium, iridium and platinum. 2. Their atomic volumes are small in comparison with those 213.] THE PERIODIC SYSTEM OF THE ELEMENTS. 301 CO Cl Q} O W fc l DO 01 8? 3 o. g 8 1 1 o ': 2 1 9 | i rH rH 00 00 l " ii O5 . O , t o *-> ri CS3 OJ 1 S o ^ 00 TH pq iO 00 r^ 00 ~i w Qi ^Q S O ^ *o ^ W O5 CO i-t T I rH t/5 I ^ O I <1 00 O5 O rH fi O o 1 302 INORGANIC CHEMISTRY. [ 213- of the neighboring elements. The atomic volume of molybdenum is 11.2, that of ruthenium, rhodium and palladium about 9; that of silver 10.3; that of cadmium 13.0. 3. They display in a marked degree the ability to let hydrogen pass through at red-heat, or to condense it on themselves at or- dinary temperatures. The former property is especially developed in iron and platinum, the latter in palladium. 4. It is only with these metals that we find RO 4 compounds, in other words, they are the only ones which can be octivalent. The compounds Os04 and RuO 4 , for example, are known, and also those with carbon monoxide, as Ni(CO) 4 and Fe(CO) 4 . We find here a general tendency to form compounds with carbon monoxide, e.g. PtCl2-3CO. In this last compound platinum can also be con- sidered octivalent. 5. They form stable and, in most cases, well-crystallizing double salts with potassium cyanide. Iron, ruthenium, and osmium give compounds of the type K 4 R(CN) 6 ; cobalt, rhodium, and iridium form K3R(CN)6, while the elements of the last column, nickel, palladium, and platinum, give K 2 R(CN) 4 double salts. 6. They all form colored salts: those of cobalt are red or blue; the nickel salts are green; all the rest are of various shades of brown. 7. They all possess the ability to condense on themselves other gases than hydrogen in larger or smaller amounts ; especially is this true of the platinum metals. Platinum and palladium absorb carbon monoxide greedily. Let us now examine the seven other groups (vertical col- umns) more closely. If we bear in mind what was stated with reference to the large periods, it is apparent that not all the ele- ments in such a group display perfect chemical analogy. Such analogy is found, however, when we compare with each other only those elements that belong to the even or the odd rows of the large periods. The similarity of the elements of these divisions is seen, in the case of the large periods, from the following facts: 1. Only the elements of the odd rows give hydrogen or alkyl compounds. 2. In the even rows the basic properties of the hydroxides are prominent, in the odd rows the acidic properties. In general it may be said that, passing from left to right in the 215.] THE PERIODIC SYSTEM OF THE ELEMENTS. 303 table, we meet first those that form the strongest bases and then gradually those that give acids. The latter property is most marked in the halogens, since they even form strong acids on combination with hydrogen. A similar change is observed in going from the top to the bottom of the system. As the atomic weight increases, the metallic (base- forming) nature of the elements in each group becomes more predominant. 214. The periodic system of the elements is of value in four different respects: 1. In constructing a system of the elements. 2. In ascertaining the atomic weights of elements whose equiva- lent weights only could be determined. 3. In foretelling the properties of elements as yet undiscovered. 4. In confirming or correcting atomic weights. Let us look at these various applications in some detail: The Use of the Periodic Law in Constructing a System of the Elements. 215. After a careful comparison of the elements MENDELEEFF reached the following important conclusion: The entire character of an element, as displayed in its physical as well as in its chemical properties, is determined by the position which it occupies in the sys- tem and particularly by the four adjacent elements, the atomic analogues. If an element is in an even series, the elements in the adjoining even series are its atomic analogues; the same is true of the odd rows. From this it follows that, when the properties of an element are exactly known, its place in the system can be assigned. A couple of illustrations will make this clear. The element beryllium possesses marked analogy to alumin- ium on the one hand and magnesium on the other; therefore it was a much discussed question whether its oxide should be given the formula BeO or ~Be 2 Oz- Since, according to analysis, 9.1 parts Be combine with 16 parts O, the atomic weight would be in the former case 9.1, in the latter |X 9. 1 = 13. 7 With the atomic weight 13.7 the element would stand between nitrogen and oxygen; as nitrogen and the elements of the sulphur group as well yield only acid-forming oxides, beryllium oxide would have to be an acid anhydride too, which is not the case, it being a basic oxide 304 INORGANIC CHEMISTRY. [ 215- Thus beryllium would not fit in the system with that atomic weight. If, however, it has the atomic weight 9.4, it falls in the horizontal series between the strongly basic lithium and the weakly acidic boron and over magnesium, with which it shows real analogy. This is indeed its fit place, i.e. its properties are those which must belong to an element in this position (see table). Farther, the properties of beryllium are to those of lithium as the properties of boron are to those of beryllium, or, in the form of a proportion: Be: Li: :B:Be. Just as lithium forms a more strongly basic oxide than beryllium, so the basic character of beryllium oxide is stronger than that of boron oxide; again, beryllium chloride is more vola- tile than lithium chloride, boron chloride more volatile than beryllium chloride. We also have the relation Be:Mg: :Li:Na: :B:A1, for beryllium oxide is less basic than magnesium oxide, lithium oxide than sodium oxide, boron oxide than aluminium oxide. Beryllium fluoride dissolves in water, magnesium fluoride does not; simi- larly boron fluoride, but not aluminium fluoride, is soluble in water. Finally we have Be:Al: :Li:Mg: :B:Si. The hydroxides of beryllium and aluminium are very similar to each other; they are gelatinous and soluble in alkalies. Both metals are scarcely acted upon by nitric acid and both dissolve in alkalies with the evolution of hydrogen. Their chlorides must be prepared in the same way from the oxides (by heating with charcoal in a current of chlorine). Likewise lithium and magnesium are analogous in certain respects; the carbonate and the phosphate of lithium, like the corresponding salts of magnesium, are very difficultly soluble, which is in marked contrast with the other metals of the lithium group. Boron and silicon both form very refractory oxides and salts; their fluorine compounds are decomposed by water in a similar manner, etc. The evidence in accordance with which beryllium was assigned its present position in the system was subsequently confirmed directly by the determination of the vapor density of the chloride, which led to the formula BeCl 2 . The vapor density of the beryllium compound of an organic substance (acety lace tone) also led to 9.1 as the atomic weight of beryllium. As a second example let us take thallium. This element dis- plays analogy with the alkali metals and also with aluminium, as 216.] THE PERIODIC SYSTEM OF THE ELEMENTS. 305 well as with lead and mercury. According to its atomic weight it must lie between the two latter and belongs in the aluminium group. This position is justified when the following relation is taken into consideration : Tl:Al::Hg:Mg::Pb:Si. The highest stages of oxidation show unmistakable analogy, since the oxides have the same properties by pairs. The oxides of aluminium and thallium are weakly basic, those of mercury (ic) and magnesium more strongly basic, while lead dioxide and silica are slightly acidic. Thallium, mercury and lead all form lower oxides of strongly basic character (T^O, Hg2O, PbO,) which alumin- ium, magnesium and silicon are unable to do. The oxide T^O is comparable with l^O in its properties, the lead monoxide with calcium oxide (T1:K: :Pb:Ca); see the table. In regard to the physical properties, it should be observed that in the matter of volatility thallium lies between mercury and lead. Use of the System in Ascertaining the Atomic Weights of Ele- ments whose Equivalent Weights only are known. 216. A good instance is that of the very rare element indium. When the periodic system was established only the equivalent weight of indium was known, the analysis of the chloride having shown it to be 38.3, i.e. it was known that 38.3 parts of indium com- bine with 35.5 parts of chlorine. If this were the atomic weight and consequently In2O the formula of the oxide, the element would have to occupy the place which potassium now has. Not only is this place taken, but the oxide is only weakly basic, while it should have strongly basic properties if it belonged to the first group. If the atomic weight were 76.6, corresponding to the oxide InO, the element would stan T between arsenic and selenium in the table; its oxide could not then have the formula InO, but would have to be In2Os or In2Os and have acid properties. On the other hand, there is no place for a metal with the atomic weight 76.6 and an oxide InO hi the second group, where the oxides have this type. If the oxide be In2Gs, the atomic weight of indium must be 115. Now, as a matter of fact, there is a vacant place in the system between Cd = 112 and Sn=119 for an element with an I^Os oxide. In 306 INORGANIC CHEMISTRY. [ 216- the same way as we have shown it for the oxides R 2 O and HO, it can also be demonstrated that indium could not be located in the table with an oxide InC>2, In 2 O 5 , InOs, etc. There remains, thus, no other possibility than to give the oxide the formula In 2 O 3 and the element an atomic weight of 115. Let us now see whether the properties of the element and its compounds conform to this location. Two of its atomic analogues for this position are cadmium and tin. The oxides of both are easily reduced; this is true also of indium oxide. If, now, we consider the metals : Ag, Cd, In, Sn, Sb (7th series), we notice first that the melting-point of silver lies higher than that of cadmium; likewise antimony melts higher than tin: Ag>Cd; Sn 223- the pressure of the ammonia over the solution is diminished or the tem- perature raised, the solution gives off ammonia and deposits copper-red, crystalline masses. When there is no longer a liquid solution these masses also lose ammonia and the metal is left behind in the crystalline form. RUFF has demonstrated that we do not have here a case of compounds being formed between the metal and the ammonia, but that the copper-red masses are mixtures of metal and saturated liquid solution; for the solution can be pressed out, leaving a compact piece of metal. The molecule of sodium contains only one atom, as is proved by the depression of the freezing-point of its solution in tin. A great many metals have this same property. In moist air the bright surface of a freshly cut piece tarnishes rapidly, but in air that has been dried with phosphorus pentoxide it keeps its metallic lustre for days. Sodium can be heated in the air to melting and even still higher without catching fire. It ignites only when heated strongly, whereupon it burns with a very bright yellow light (especially in an atmosphere of oxygen). With water it generates hydrogen, sodium hydroxide being also formed. If it is held firmly in one place during this process (e.g. by laying it on a piece of filter-paper floating on water, or upon ice), the hydrogen takes fire because of the localization of the heat produc- tion. Sodium finds extensive use in the laboratory and in the arts. Because of its strong reduc ing-power it is often used to obtain the elements from their oxides; magnesium and aluminium were formerly obtained thus. In organic chemistry it is also frequently employed for various purposes. OXIDES AND HYDROXIDES OF SODIUM. 224. On burning sodium in dry oxygen a mixture of two oxides, Na^O and Na2O 2 , results. Sodium oxide, Na2O, is obtained pure by the partial and slow oxidation of sodium with oxygen under reduced pressure and removal of the excess of metal by distillation in a vacuum. The oxide dissolves slightly in the metal and after distilling off the latter the oxide is left in the crystalline form. It is white; it dissolves in water, forming sodium hydroxide, NaOH, and giving off much heat. The peroxide, Na 2 2 , is obtained by heating sodium in a current of oxygen till no more oxygen is absorbed. With 8 mols. 224.] OXIDES AND HYDROXIDES OF SODIUM. 313 water it forms a hydrate, Na 2 02 + 8H 2 O. Since it yields hydrogen peroxide with dilute acids and is a vigorous oxidizing-agent it is manufactured commercially. Sodium hydroxide, NaOH ; caustic soda, is formed, together with metallic sodium, when sodium monoxide is reduced in a cur- rent of hydrogen. The ordinary method of preparing caustic soda consists in boiling soda with slaked lime: Na 2 CO 3 + Ca(OH) 2 =2NaOH+CaCO 3 ; or 2Na- + CO 3 " + Ca" + 20H' = 2(Na- + OH') + CaCO 3 . As the solubility product ( 73) of CaC0 3 molecules is very small, the ions Ca" and C0 3 " must unite and the carbonate of lime sinks to the bottom. In order to make the decomposition of sodium carbonate complete, a slight excess of slaked lime is added. Never- theless, the solution does not contain an appreciable quantity of calcium hydroxide. The reason of this is clear: In the solution there are a large number of OH-ions; as a result the number of Ca-ions can only be very small, for the value of the solubility product of calcium hydroxide is reached with even a very low concentration of the latter ions. Sodium hydroxide is now manufactured on a large scale by the electrolysis of concentrated brine. This method yields an almost chemically pure hydroxide and it dominates the market with users of high-grade caustic. Three types of electrolytic processes are in operation: the diaphragm type, the amalgamation type, and the bell type. In the first type the cathode and anode compartments are separated by a diaphragm. In the Griesheim process, a suc- cessful representative of this type, the cathode is of iron and the anode ferrous-ferric oxide, Fe 3 04> that has been fused at 2000- 3000 and cast into plates. (This makes an anode unaffected by chlorine.) In a freshly charged bath containing only chloride solu- tion the current is carried mainly by sodium and chlorine ions; but as fast as the sodium is liberated at the cathode and reacts with water to form hydroxide and free hydrogen, the ions of the hydroxide participate in the transport of the current. The sodium atoms are reliberated and again react with water to form hydrogen, while the discharged hydroxyl ions yield water and oxygen. As a net result of the electrolysis we have, so to speak, an intentional decomposition of alkali chloride accom- 314 INORGANIC CHEMISTRY. [ 224- panied in an increasing measure by an unintended decomposition of water. On this account the chloride electrolysis cannot be carried to completion; in practice the process is interrupted as soon as the alkali hydroxide concentration gets up to 8%. The caustic cathode liquid is then replaced by fresh brine and the former is evaporated in vacuum pans to a concentration of 50%, whereupon the undecomposed chloride separates out and is returned to the electrolytic cell. The diaphragm process most favorably known in America is the TOWNSEND process. The compartment containing the (graphite) anodes occupies the center of the cell and on each side i: a diaphragm of tliin asbestos. The cathodes, of woven wire, :est closely against the outside of the diaphragm and the cathode compartment is filled with warm oil, that is kept in lively cir- culation by the escaping hydrogen. The freshly formed caustic liquor trickles down the side of the cathode and is drawn off from beneath the oil. The constant removal of the hydroxide enables the electrolysis to be carried p/actically to completion and the yield approaches the theoretical. A high current density is employed. The CASTNER process is the most extensively used of the amalgamation type. Its cell is divided into three compartments, the two outside ones containing brine and the carbon anodes, while the middle one contains the caustic liquor. A layer of mercury covers the bottom of the whole cell. In the brine com- partments the mercury acts as cathode, taking up the sodium to form amalgam. The amalgamated mercury is transferred to the middle compartment where it is decomposed by water to form sodium hydroxide. This caustic solution is drawn off and fresh water introduced at a regulated rate. On evaporation a very pure sodium hydroxide results. The efficiency of the proc- ess is enhanced by conserving the electrical energy liberated in the decomposition of the amalgam. In one successful bell process the anode consists of some carbon supported in a bell which is suspended in the brine and has an exit tube at the top for piping off the chlorine. The cathode is a cylindrical piece of sheet iron encircling the bell. For critical discussions of the relative merits of these and 225.] SALTS OF SODIUM. 315 competitive processes the reader must refer to the technical journals or the most recent works on industrial chemistry. The hydroxide is obtained by evaporation, whereupon it is generally cast into sticks for the market. It is radiate-crys- talline and very hygroscopic. It dissolves in water with the evolution of considerable heat. Sodium hydroxide is a very strong base; it is used in the arts for numerous purposes, among others the manufacture of soap. SALTS OF SODIUM. 225. The sodium salts are of great industrial importance; many of them are prepared in enormous quantities. The starting- point for their manufacture is usually common salt. Sodium chloride, NaCl, common salt, is found in large masses as rock-salt, e.g. at Stassfurt and Reichenhall and in Galicia, where it ,is dug out by miners. Farther, large, amounts are obtained from sea water and the water of salt wells. Three methods are employed to remove salt. In sufficiently warm countries (e.g. Mediterranean coast, central New York State) the brine is led into flat basins of very large surface area (" salterns," or "salt covers "). In these the water is removed by solar evaporation and the salt crystallizes out. Any gypsum that may be present separates out first, whereupon the brine passes to further basins and yields pure salt. Later on, the other salts separate out; these are sometimes worked up commercially. In countries with a cold climate (e.g. on the shores of the White Sea) the water is allowed to freeze in flat basins. The ice that forms is free from salt so that the remaining liquid is more con- centrated. In countries of the temperate zone the sea water is con- centrated by letting it evaporate spontaneously from a greatly enlarged surface. This is done by the "graduation" process (Fig. 50). Bundles of fagots are piled up together in a "rick," above which a trough with small outlet-holes runs from end to end. The brine is pumped up into the trough and trickles down from along the entire length of the latter upon the brush; in this way the surface of the salt solution is greatly enlarged and the evaporation is made 316 INORGANIC CHEMISTRY. [225 much more rapid. A very concentrated brine flows out at the bottom. The salt is obtained from this concentrated solution by boiling ("salt-boiling"). Common salt is almost equally soluble in hot and cold water, hence it does not crystallize out on cooling but falls out at the same rate as the saturated brine evaporates, even while FIG. 50. GRADUATION PROCESS. boiling hot. The salt obtained from the first crystallization is of course impure, containing small amounts of magnesium salts, which render it hygroscopic. In order to purify it, it is redissolved in water and again precipitated by evaporation. Chemically pure sodium chloride is obtained by passing hydro- chloric acid gas into a saturated solution of the salt or treating the solution with the concentrated acid. The sodium chloride is deposited because it is less soluble in hydrochloric acid than in water ( 205). Common salt crystallizes in cubes; when the solution evaporates 225.] SALTS OF SODIUM. 317 slowly the well-known hollow four-sided pyramids, or hopper- crystals (Fig. 51), are formed. Sp. g.=2.16. M.-pt.=776. 100 parts of water dissolve 36 parts NaCl at 0. 39 parts at 100 ; a saturated solution contains about 26% of salt. The crystals frequently enclose some of the mother-liquor; for this reason they decrepitate on heating. On cool- ing below 10 a saturated solution depos- its crystals of the composition NaCl + H 2 O, which lose their water at 0. Chemically pure sodium chloride is not hygroscopic. It i 8 insoluble in absolute alcohol. Sodium bromide, NaBr, and sodium iodide, Nal, are more soluble in water than the chloride. From hot solutions they crystallize in anhydrous cubes, below 30 in monoclinic crystals with 2 mols. H 2 0. Sodium bromide is difficultly soluble, sodium iodide easily soluble, in alcohol. Sodium thiosulphate, Na 2 S 2 3 -5H 2 ; spoken of sometimes as sodium hyposulphite or " hypo/' is employed in photography ( 247) , in titrating iodine ( 93), and as an antichlor ( 82). It melts in its water of crystallation at 45 and forms super- saturated solutions very easily. Sodium sulphate, Na 2 SO4-10H 2 (sal mirabile Glauberi, Glau- ber's salt so called after the discoverer), can be obtained in various ways: (1) by heating common salt with concentrated sulphuric acid; (2) by conducting a mixture of air, sulphur dioxide and steam over hot sodium chloride (HARGREAVE'S method): 2NaCl + S0 2 + O + H 2 O = Na 2 SO 4 + 2HC1 ; (3) by the double decomposition of magnesium sulphate and sodium chloride at a low temperature (winter- temperature) : MgSO.i + 2NaCl = MgCl 2 + Na 2 SO 4 . This last process is carried out at Stassfurt, where large masses of magnesium sulphate occur. At ordinary temperatures Glauber's salt crystallizes with ten molecules of water ; above 33 it goes over into a mixture of water 318 INORGANIC CHEMISTRY. [ 225- and anhydrous salt; below 33 the hydrous salt, is again formed. The system Na2SO4-10H 2 O^Na 2 SO4-}-10H 2 O thus has a transition point at 33. This is confirmed by the fact that the solubility of Glauber's salt in water suddenly changes at 33 from a rapid increase with rising temperature to a slow de- crease. This salt, like the preceding one, forms supersaturated solutions easily. The solution saturated at 33 can be cooled down to room- temperature without any deposition if care is taken to exclude even the tiniest crystal of the salt. When a crystal of Glauber's salt is exposed to the air it effloresces, i.e., it gives off water vapor and becomes opaque, like chalk. This is evidently due to the fact that the vapor tension of its water of crystallization is greater throughout than that of the water vapor in the air. Inversely, we say that a salt deliquesces when it takes up water vapor from the air, as a result of the vapor tension of its saturated solution being less than the mean tension of the water vapor in the atmosphere. It is found that a perfectly sound crystal of Na 2 SO-4 10H 2 does not effloresce, but that when efflorescence has begun at any point it spreads over the crystal. The phase rule gives a satisfactory explanation of this phenomenon. We have in the Glauber's salt two substances, Na 2 S04 and H 2 0; in the case of a perfectly bright crystal exposed to the air we have only two phases, Na 2 S04 10H 2 O and H 2 O (moisture of the air). According to the equation F+P=S+2 (71) we still have in this system two degrees of freedom, i.e., the pressure of the water vapor and the temperature can both be selected arbitrarily (within certain limits). If, however, some dehydrated salt is present, the number of phases is three; then there is only one degree of freedom, or ; in other words, every temperature has only one corresponding pressure and inversely every pressure only one corresponding temperature. Accordingly it is only permissible to speak of the well-defined vapor tension of a salt with water of crystallization when a second solid phase is tacitly admitted to be present; for then only is the number of degrees of freedom reduced to one. Glauber's salt is used in medicine; in the arts it is employed chiefly in the manufacture of soda. Sodium nitrate, NaNO 3 , called Chili saltpetre because of its extensive occurrence in Chili, crystallizes in rhombohedrons and 223.] SALTS OF SODIUM. 319 melts at 318. It is somewhat hygroscopic. Large quantities of it are used in fertilizing and also in the preparation of nitric acid and potash saltpetre. Sodium nitrite, NaNO 2 , is obtained by heating the nitrate, the addition of a reducing-agent such as lead or iron aiding the process. It is very soluble in water and is consumed in large quantities in the aniline-dye industry. Sodium phosphates. (See 146.) Trisodium phosphate, Na 3 P0 4 12H 2 0, is split up in aqueous solution chiefly into sodium hydroxide and the secondary salt, for the solution reacts strongly alkaline and absorbs carbonic acid from the air. The ordinary "sod'.um phosphate" is the disodium phosphate, Na 2 HPO4 12H 2 O. It separates from its aqueous solution at ordinary temperatures in large crystals which soon effloresce. 100 parts H 2 O dissolve 9.3 parts of the salt at 20 and 24.1 parts at 30. The solution is feebly alkaline. By leading in carbonic acid gas a liquid of amphoteric reaction (cf. 224) is obtained, which turns blue litmus red as well as red litmus blue. Monosodium phosphate, NaH^PCVRsO, reacts acid. It is converted by heat into the metaphosphate. 226. Sodium carbonate, Na 2 CO 3 -10H 2 O (soda, sal-soda), is, next to the chloride, the most important sodium salt and it is manufactured on an enormous scale. It occurs in nature in Egypt, in America (Wyoming), on the Caspian Sea, and elsewhere; the ashes of many marine plants contain it. The manufacture of soda is carried on mainly by two different methods : (1) The LE BLANC soda process. This process consists of three parts. In the first place common salt is warmed with strong sulphuric acid (chamber acid) ; hydrochloric acid and sodium sulphate " salt cake " are formed. In the second part of the process sodium sulphate is heated with coal and limestone. The third section of the process consists in lixiviating the mass last obtained (" black-ash ") with water, whereby sodium carbonate is dissolved out. The latter is then obtained from this solution by crystallization. After the black ash has been leached out as far as practicable, it is cast aside as " tank waste." The most valuable constituent of the latter is calcium sulphide ; it is conserved by treating the 320 INORGANIC CHEMISTRY. ( 226. waste with water and carbon dioxide to form hydrogen sulphide, which is then oxidized to sulphur and the latter to sulphuric acid. The process in its entirety is thus represented by the following equations: . 2NaCl + H 2 SO 4 =Na 2 SO4+2HCl; Na 2 SO 4 + 2C = 2C0 2 + Na 2 S ; Na 2 S + CaCO 3 = CaS + Na 2 CO 3 ; G0 2 + H 2 O + CaS = CaC0 3 + H 2 S ; = H 2 SO 4 ; or, summed up : 2NaCl + CO 2 + H 2 O = 2HC1 + Na 2 CO 3 . The process is noted for its high efficiency, since all the by-products are worked up. Nevertheless, this process, which for a long period of years practically controlled the industrial market, is now almost wholly superseded by the other one and a few years more will probably see its entire abandonment. (2) The ammonia-soda process ofSoLVAY. This process, which originally presented numerous technical difficulties, is now so perfected that about ninety-five per cent of the total soda production is by the SOLVAY process. The chemistry of the SOLVAY process is very simple. Ammonia and carbon dioxide are led alternately into a cold concentrated salt solution under pressure. The following reaction then takes place : NaCl + (NH 4 ) HCO 3 = NaHCO 3 + NH 4 C1. The acid sodium carbonate (" bicarbonate ") so formed sepa- rates out, inasmuch as it is very difficultly soluble in the cold concentrated ammonium chloride solution. It is broken up, on heating, into soda and carbon dioxide, the latter of which is carried back to be used over. The ammonium chloride solution is distilled with lime, whereby ammonia is recovered. The process as a whole may be represented by the following equations : 226.] SODIUM CARBONATE. 321 2NaCl + 2NH 3 + 2CO 2 + 2H 2 O = 2NH 4 C1 + 2NaHCO 3 ; 2NaHC0 3 = H 2 O + CO 2 + Na 2 CO 3 ; 2NH 4 C1 + CaO = 2NHg + H 2 O + CaCl 2 ; CaC0 3 = or, summed up : 2NaCl + CaC0 3 = Na 2 C0 3 + CaCl 2 . In. the SOLVAY process there is formed together with the soda an equivalent amount of calcium chloride, for which there is only a limited market (chiefly for refrigeration; cf. 258), so that one valuable con- stituent of salt, the chlorine, is largely lost. Attempts to substitute magnesia for the lime so as to be able to utilize the resulting magnesium chloride for the recovery of hydrochloric acid or chlorine have not been commercially successful. Some soda is also manufactured by carbonating the electrolyt- ically prepared sodium hydroxide. Sodium carbonate crystallizes at ordinary temperatures with ten molecules of water of crystallization in large transparent monoclinic crystals, which soon turn, white and dull from loss of water (efflorescence). They melt at 60 in their own water; on continued warming the hydrate Na 2 CO 3 + 2H 2 O is deposited; the latter loses one molecule of water in dry air and the remainder at 100. At 30-50 rhombic prisms of the composition Na 2 CO 3 + 7H 2 crystallize out of the aqueous solution. 100 parts H 2 O dissolve 6.97 parts of the anhydrous salt at 0, 51.67 parts at 38. The aqueous solution reacts strongly alkaline ( 184) because ' hydrolysis. As was set forth in 146, this phenomenon m, always occur when a salt of a weak base and a strong acid or u. salt of a weak acid and a strong base is formed. In case both acid and base are weak the hydrolysis will be all the greater. Which reaction the solution of such a salt will give depends on the relative strengths of the acid and base. Soda is used commercially on a large scale, particularly in the soap and glass industries. It is the " washing-soda " of the household. 322 INORGANIC CHEMISTRY. [226- Acid sodium carbonate, NaHC0 3 (bicarbonate of soda), is obtained as a primary product in the SOLVAY process. It dis- solves in 10-11 parts of water at room temperature and reacts alkaline. On being gently warmed it breaks up into carbon dioxide, water and soda; this decomposition occurs even on warm- ing the aqueous solution, and when a current of air is passed through the concentrated solution at ordinary temperatures car- bon dioxide escapes. It is used extensively in baking-powders and is the saleratus of commerce. Sodium silicate (sodium water-glass) is prepared, among other ways, by fusing sand with Glauber's salt and charcoal. This yields a vitreous mass, which is dissolved by boiling water. The concentrated solution has the consistency of glue. It finds use as a fixative in calico printing, as well as for impregnating inflam- mable textiles like theater decorations, etc.; it is also used for " filling " soaps. The sulphides of sodium correspond to those of potassium and are prepared in the same way (see 231). Sodium borate: cf. Borax (283). POTASSIUM. 227. Compounds of potassium occur in nature very extensively but not in such large quantities as those of sodium. Potassium exists principally in the silicates, especially feldspar and mica. Upon the decay of these minerals it is carried into the soil and thence into the plants, to which potassium compounds are indis- pensable. Potassium salts are also found in sea- water. The largest source of them, however, is the Stassfurt " Abraum salts " ( 44), mainly double-salts of potassium and magnesium, such as carnallite, MgCl 2 KC1 6H 2 0, kainite, MgS0 4 Kd 3H 2 O, etc. The large amounts of potassium in the feldspars makes its recovery from them a very enticing problem. The metal was first obtained by DAVY by the electrolysis of molten caustic potash. One of the commercial methods is to ignite a mixture of carbonate of potash and powdered charcoal (preferably charred acid potassium tartrate). The extraction of the metal is thus analogous to that of sodium; in the preparation of potassium, however, potassium carbonyl, Co(OK) 6 , may be 228. ] OXYGEN COMPOUNDS OF POTASSIUM. 323 formed under certain circumstances, a substance which acquires explosive properties on exposure to the air. Potassium has a silvery-white metallic lustre and is almost as soft as wax at ordinary temperatures. Sp. g. =0.875 at 13. It melts at 62.5 and boils at about 720, forming a green vapor. The mirror-like surface of the metal immediately becomes dull in the air; when heated in the air it burns with an intense violet light. Water is decomposed by it with great vigor, the heat evolved being sufficient to ignite the escaping hydrogen and drive the piece of potassium around on the water. Oxygen Compounds of Potassium. 228. Potassium oxide, K 2 O, is formed by oxidizing potas- sium by the method described in 224. It is a white substance, which unites with water to form the hydroxide with the evolution of much heat. Potassium peroxide, KC>2, is produced together with the mon- oxide on burning potassium in the air. It is dark yellow. In con- tact with water it yields potassium hydroxide, hydrogen peroxide and free oxygen. Potassium hydroxide, KOH, results from the action of potas- sium on water and is generally prepared in the same manner as sodium hydroxide, viz., by treating potassium carbonate solution with milk of lime, Ca(OH) 2 . It can also be obtained by heating saltpetre with powdered copper (forming copper oxide and potas- sium oxide), and adding water; the copper oxide can be removed by filtration. The hydroxide usually comes on the market in sticks. The commercial product (" caustic potash ") is obtained chiefly by the first method and usually contains a little sulphate, chloride, etc., besides the carbonate which is gradually formed by the action of atmospheric carbon dioxide. It can be purified by treating with strong alcohol, which dissolves only the hydroxide; after filtering, the alcoholic solution is evaporated in a silver dish. Caustic soda is also purified in this way. Potassium hydroxide is one of the strongest bases. In the solid state it greedily absorbs water and carbon dioxide from the air and finally deliquesces to a concentrated solution of potassium carbonate, while sodium hydroxide under these conditions turns to a solid 324 INORGANIC CHEMISTRY. [228- white mass of soda. For this reason caustic potash is a much more suitable absorptive agent for carbon dioxide in analyses than caustic soda, for the use of the latter might easily cause a stopping up of the apparatus. Caustic potash is used especially in the manufacture of soft soaps. Potassium Salts. 229. Potassium chloride, KC1, occurs at Stassfurt in the min- eral sylvite. It crystallizes in cubes and melts at 730. It is easily volatilized at elevated temperatures. 100 parts H 2 O dissolve 25.5 parts KC1 at 0, 57 parts at 100. Like its sodium analogue, potassium chloride is precipitated from its saturated solution by hydrochloric acid. It unites with many salts to form double salts. Potassium bromide, KBr, is important therapeutically. It is prepared by mixing bromine with a potassium hydroxide solution, the bromide and bromate being formed; the bromate is reduced by heating the salty product with powdered charcoal. Potassium bromide crystallizes in cubes and dissolves readily in water. Potassium iodide, KI, also of medicinal value, can be prepared like the bromide and also in the following manner: Iodine and iron filings are mixed together under water, whereupon a solu- tion of the compound Feslg is formed; on treating this with a potash solution the oxide Fe304 is precipitated, carbon dioxide escapes and potassium iodide is left in solution; the salt is then obtained by filtration and evaporation. It crystallizes in cubes and is very soluble in water: 1 part H 2 dissolves 1.278 parts KI at 0. On exposure to light or the air the crystals gradually turn yellow because of the separation of iodine. It was remarked in 46 that iodine, though only slightly soluble in water, dissolves to a much greater extent when the water contains potassium iodide. This is due to the formation of JV ions in the latter case. That the iodine has entered into combination may be concluded in the first place from the fact that the addition of iodine to an aqueous solution of potassium iodide does not cause a further depression of the freezing-point; the number of molecules is thus unchanged, or, in other words, iodine has combined with potassium iodide: in the second place, from the fact that carbon disulphide takes up nearly all tHe iodine from an aqueous solution of the latter when it is shaken with the solution, but only a small proportion when the same operation is 229. POTASSIUM SALTS. 325 performed with a solution of iodine in a dilute aqueous solution of potassium iodide. The distribution ratio for iodine between water and carbon disulphide is 1:410. If, therefore, we divide the con- centration of the iodine in carbon disulphide by 410 we obtain the concentration of the free iodine in the potassium iodide solution. Subtracting this from the total concentration of the iodine in this solution, we have the amount of combined iodine. It is found that 1KI, or, rather,!!', has taken up 1I 2 . Is' ions are thus formed in .the solution. Nevertheless, a solution of iodine and potassium iodide in water behaves in many cases as if all the iodine were present in the free state, e.g. when it is titrated with sodium thiosulphate. This must be explained by the supposition that in the liquid we have the equilibrium: If the free iodine is removed, the equilibrium is disturbed; a new portion of 1% must therefore split up, and so on till it is entirely consumed. Potassium fluoride, KF, possesses a peculiar property, which is lacking with the other halogen compounds of potassium: it com- bines with hydrofluoric acid eagerly, forming the double halide KF-HF. Potassium cyanide, KCN (often also written KCy), is manufac- tured on a large scale by fusing yellow prussiate of potash with potash : K 4 Fe(CN) 6 + K 2 C0 3 = 5CNK + KCNO + C0 2 + Fe. The cyanate of potassium KCNO is reduced by the iron to potassium cyanide also. It is very soluble in water, forming a strongly alkaline solution. On account of its great tendency to form double- salts, it is employed in electro-metallurgy. It is also used in extracting gold from its ores ( 248). Potassium chlorate, KClOs, can be obtained by passing chlo- rine into a h o t solution of caustic potash ( 56). It is now pre- pared almost exclusively by the electrolysis of a hot solution of sodium chloride. If in the GRIESHEIM process ( 224) for the manufacture of caustic alkali the electrolysis is continued after a certain amount of hydroxide has formed, the oxygen liberated 326 INORGANIC CHEMISTRY. [ 229- at the anode oxidizes the sodium chloride to chlorate, NaClO 3 . The latter is converted into the potassium chlorate by treatment with potassium chloride. The advantage of this method is that sodium chlorate is much more soluble and does not retard the electrolytic process by separating from the solution, as does potassium chlorate. Potassium chlorate is a well-crystallized salt, which is used for the preparation of oxygen ( 9) ; furthermore, it is used in the manufacture of matches and fireworks, and also medicinally as a remedy for sore throat. On being heated it gives up oxygen, part of the salt being at the same time converted into potassium perchlorate, KC10 4 . The last-named salt is difficultly soluble in water. It is sometimes found in crude Chili saltpetre, rendering the latter unfit for use in fertilizing various cultivated plants. Potassium sulphate, K 2 S04, is obtained by the action of sul- phuric acid on potassium chloride. It crystallizes in beautiful, lustrous rhombic prisms and dissolves with some difficulty in cold water (1 part in 10 parts H 2 at room temperature). It is used principally for the preparation of potash according to the LE BLANC method. Acid potassium sulphate, KHS0 4 , is very soluble in water; it melts at 200, losing water and going over into potas- sium pyrosulphate, K 2 S 2 O7. The latter breaks up into potassium sulphate and sulphur trioxide on heating. Potassium nitrate, KNOs, is widely distributed in nature, although usually found only in small amounts, for it is formed wherever nitrogeneous organic bodies decay in contact with potas- sium compounds. This is the basis of an artificial method of preparing saltpetre, which method was formerly much used. Another process of manufacture depends on the double decom- position of Chili saltpetre with potassium chloride, which is obtained in large quantities at Stassfurt: KC1 + NaNO 3 = KN0 3 + NaCl. For this purpose hot-saturated solutions of the two salts are brought together. As sodium chloride is much less soluble than saltpetre 231.] POTASSIUM SALTS. . 327 at the temperature of boiling water, it is the first to crystallize out on evaporation, but when the solution is cooled the saltpetre comes out first, for it is much less soluble than sodium chloride in cold water. Potash saltpetre crystallizes in anhydrous prisms, either rhom- bohedral or rhombic according to the temperature. In the neigh- borhood of the melting-point the former is the stable variety at ordinary temperatures the latter. The location of the transi- tion point of the two forms has not yet been determined. 100 parts H 2 dissolve 13.3 parts KN0 3 at 0, 247 at 100. It melts at 338; farther heating breaks it up into potassium nitrite and oxygen. It has a cooling taste. 230. Potash saltpetre is consumed in large quantities in the manufacture of gunpowder. This is a mixture of sulphur, charcoal and potash saltpetre, the proportions varying in different countries, but being in most cases 75% KN0 3 , 10% S and 15% charcoal. 231. Potassium phosphates. The three potassium salts of phosphoric acid are known. They are very soluble in water. Potassium carbonate, K 2 CO 3 , potash. This salt was formerly obtained solely from wood-ashes, these being soaked in water and the strained liquor evaporated. At present it is manufactured extensively from potassium chloride after the LE BLANC process. Another source of potash is the molasses of the beet-sugar fac- tories, that contains the potassium salts in which the sugar-beet is rich. At the Neustassfurt salt mine it is made from potassium chloride by a patented process as follows: Magnesium carbonate, MgC0 3 -3H 2 0, is suspended in a solution of potassium chloride, and carbon dioxide is led in, whereupon the following reaction takes place: 3MgC0 3 .3H 2 O +2KC1 +C0 2 =MgCl 2 + 2MgC0 3 KHC0 3 H 2 0. The potassium magnesium carbonate separates out and is broken up by heating to a temperature not exceeding 80 into magnesium car- bonate and potash. The former salt is again obtained with three mole- cules of water of crystallization, which form is the only one suited for the above reaction. Potassium carbonate is a white powder, which deliquesces in the air and is very soluble in water (1.12 parts K 2 COs in 1 part 328 INORGANIC CHEMISTRY. [ 231- H 2 O at 20) ; the solution has a strong alkaline reaction. The salt melts at 838. It is used in the preparation of soft soaps and hard glass (potash-glass). Potassium silicate, potassium water-glass, is formed when sand is fused with potash. Different salts of this sort are described. They dissolve in water, forming a thick, mucilaginous mass which on drying turns to a vitreous, and finally opaque, product. Potassium water-glass is used for the same purposes as sodium water-glass. Sulphides of Potassium. Potassium monosulphide, K 2 S, is prepared by reducing potassium sulphate with charcoal. It dissolves in water very readily and crystallizes out with five molecules of water. It absorbs oxygen from the air ; going over into the thio- sulphate and hydroxide: 2K 2 S + H 2 + 20 2 = K 2 S 2 3 + 2KOH. Acids react with it, liberating hydrogen sulphide. Potassium hydrosulphide, KSH, is obtained by saturating a caustic potash solution with hydrogen sulphide: KOH+H 2 S=KSH + H 2 0. It is very soluble in water, the solution reacting alkaline; on evaporation in vacuo the solution deposits crystals of the com- position 2KSH + H 2 O. With potassium hydroxide it forms the monosulphide : KSH+KOH=K 2 S + H 2 0. Potassium polysulphides. When a solution of potassium mono- sulphide is boiled with sulphur, we obtain the compounds K 2 Ss, K 2 $4, K 2 S 5 . A mixture of these substances is also obtained by fusing potash with sulphur; besides these it contains the sulphate and the thiosulphate and is called hepar sulphuris (" liver of sul- phur ") because of its liver-brown color. These polysulphides are 232.] RUBIDIUM AND CESIUM. 329 decomposed by acids with the evolution of hydrogen sulphide and the separation of sulphur: K 2 S X + 2HC1 = 2KC1 + H 2 S + (x - 1)8. Rubidium and Caesium. 232. These elements are widely distributed, but always occur in extremely small amounts. The silicate lepidolite, or lithia mica, fre- quently contains a little rubidium. The exceedingly rare mineral pollux from the isle of Elba is a silicate of aluminium and caesium, and contains about 30% caesium oxide. In general these elements are found where- ever potassium salts are met with: in mineral springs, in the Stassfurt salts (carnallite contains rubidium), etc. They were discovered by BUNSEN and KIRCHHOFF in 1860 with the aid of spectrum analysis ( 264) and obtained their names from the most important lines in their spectra (rubidus = d&Tk red; ccesius = sky-blue.) The spectrum lines were used as a test in the separation of these elements from the others; after trying a possible method of separation the two sc,var:t3 would see which portion showed the lines of these elements the brightest ; this portion was then examined further. In order to separate them from the large amount of potassium salts with which they generally occur, they are converted into chlorides and the solution is evaporated, whereupon the dry residue is extracted with strong alcohol. Almost all the sodium chloride and potassium chloride remains behind, while the chlorides of rubidium and caesium dissolve. Platinum chloride is then added to precipitate K 2 PtCl 6 , Rb 2 PtCl e , and Cs 2 PtCl 6 ; the solubility of these double salts in water is quite different (at 10 100 parts H 2 dissolve 0.9 parts K 2 -salt, 0.154 Rby-salt, and 0.05 Cs 2 -salt), so that they can be very well separated by fractional extraction with boiling water. The rubidium iron "alum" is particularly well suited for the purification of rubidium salts, and especially for their separation from potassium salts, since it is readily soluble in hot, and only slightly soluble in cold, water, and moreover crystallizes beautifully; potassium iron alum, on the other hand, is very soluble even in cold water. The metals rubidium and caesium are best obtained by heating their hydroxides with calcium filings in a vacuum. The metals then distil 330 INORGA NIC CHE MIS TR Y. [232- off. Rubidium has a silvery lustre, melts at 38.5, and has a specific gravity of 1.522 at 15. The metal oxidizes very rapidly in the air or in oxygen, forming dark brown crystals of the peroxide, RbO 2 . On being heated in current of hydrogen it yields the hydroxide and free oxygen : 2Rb0 2 + 2H 2 = 2RbOH + H 2 + O. Rubidium oxide, when prepared in the same way as Na 2 0, is obtained as transparent, pale yellow crystals which turn golden yellow on heating but lose this color again on cooling. The hydroxide is a very strong base; its salts show much similarity to the analogous potassium compounds; they are in several instances less soluble, e.g., Rb-alum, Rb-perchlorate ( 60), etc. Csesiumisa silvery-white metal; sp. g. 1.85; m.-pt. 26.5; b.-pt. 670. It soon takes fire on exposure to the air. The oxide, Cs 2 0, obtained in the same way as the other alkali oxides, is crystallized and is orange-colored at room temperature but almost black at 250. The salts of csesium are very similar to those of rubidium ; some of them are even less soluble, and are therefore used for the preparation of pure csesium compounds. This is particularly true of the platinum double-salt already mentioned and the csesium alum and the acid tartrate. Rubidium bromide and iodide, and even more so the corresponding compounds of caesium, have the property of combining with two atoms of bromine and iodine, forming yellow or brown crystalline compounds, e.g. CsI 3 ; these metals can thus be trivalent. SUMMARY OF THE GROUP OF ALKALI METALS. 233. The gradual change of the physical properties of these metals with increasing atomic weight is made plain by the follow- ing table: Li Na K Rb Cs Atomic weight 7.00 23.00 39 10 85.45 132.81 Specific gravity 0.59 0.97 0.865 1.52 1.85 Melting-point . . . 180.0 97.6 62.5 38.5 26-27 Boiling-point <1400 877 757 696 670 Atomic volume 11.8 23.7 45.3 56.7 71.9 234.] AMMONIUM SALTS. 331 The specific gravity increases with the atomic weight, as does also the atomic volume; on the other hand, there is a fall in the melting- and boiling-points. From a chemical standpoint we notice, in the first place, the same general type in the compounds, showing that all these ele- ments are univalent. The hydroxides all have the formula ROH. the halogen compounds RX, etc. The salts of them all, even the carbonates and phosphates, are soluble in water (although in different degrees), the carbonates with basic reaction. The metals all oxidize very readily in the air. On the other hand, we cannot overlook the fact that the metals potassium, rubidium and cesium, which are very similar to each other, differ from sodium and lithium in many respects. The last- named metal, as we shall see in the sequel, displays analogy with magnesium in several important points, thus differing from the metals of its own group. A slight divergence in the behavior of the first members of a group from that of the rest is found to characterize almost all of the groups. We may recall carbon, for instance, the first member of the fourth group, which differs dis- tinctly from silicon and the rest in the ability of its atoms to unite with each other; also fluorine with its soluble silver compound. Still other examples of this sort will be met with later. Sodium differs from the sub-group, K, Rb, Cs, in the solubility of its salts. The sodium salts are almost all readily soluble in water; this is true even of the platinum double-salt, Na 2 PtCl6, the acid sodium tartrate and others. Soda crystals effloresce, while potash deliquesces in the air. The spectra of sodium, on the one hand, and the other alkali metals, on the other, are entirely dis- similar. Ammonium Salts. 234. In the description of ammonia ( 112) it was already observed that it combines with acids directly, forming salts which are very similar to those of potassium, and in which the group, NH 4 , the ammonium group, is assumed to exist. In connection with the alkali group a description of a few ammonium salts may find a place. The aqueous solution of ammonia must, because of its electrical conductivity and its alkaline reaction, contain NH4 and OH' ions 332 INORGANIC CHEMISTRY. [234- and hence also undissociated molecules of ammonium hydroxide, NH 4 OH. While solutions of the alkalies, KOH, NaOH, etc., conduct the electric current very well, this is not the case with an ammonia solution; it is a poor conductor. A 0.1 normal solution contains only 5% of ionized NH 4 OH molecules, while a solution of potassium hydroxide of the same concentration is 91% ionized. An aqueous solution of ammonia may be presumed to contain: (1) free ammonia, NH 3 ; (2) hydrates of ammonia, NH 3 -n aq.; (3) ammonium hydroxide, NH 4 OH; (4) the ions NH 4 and OH'. The existence of these hydrates in addition to free ammonia reveals itself in the behavior of ammonia solutions on being shaken with chloroform. According to BERTHELOT'S law (ORG. CHEM., 25) the distribution ratio of the ammonia between the two solvents should be constant. But this is not the case. Therefore, just as the deviations from HENRY'S law lead us to conclude that a dissolved gas exists in a special condition, so we can apply a simi- lar explanation to the exceptions to the BERTHELOT law; for HENRY'S law is really the expression of the distribution ratio of a gas between a liquid and a vacuum, while the other law has to deal with the distribution ratio between two liquids. Since in the case of ammonia this deviation is observed only when one of the two liquids is water, we are obliged to conclude that there is a combination of the ammonia with the water. The reason for assuming the existence of hydrates instead of ammo- nium hydroxide, NH 4 OH, is a double one. We have, on the one hand, the analogy between the behavior of ammonia and amines and, on the other, the entirely abnormal behavior of the organic quaternary ammonium bases. For, while the aqueous solutions of primary, secondary and tertiary amines are weak electrolytes, as is the case with ammonia, the solution of a quater- nary base, on the contrary, conducts electricity as well as a solu- tion of potassium or sodium hydroxide. We may thus conclude that if ammonium hydroxide, NH 4 OH, could reach the same concentration in solution as a quaternary base it would display just as great a conductivity as the latter. Unlike the quaternary base, however, it breaks up principally into ammonia, NH 3 , and water, for the quaternary base cannot be thus decomposed. That an aqueous ammonia solution really contains at least 234.] AMMONIUM SALTS. 333 an appreciable quantity of the hydroxide is evidenced by the existence of its ions. These necessitate the establishment of the equilibrium however far the point of equilibrium may be displaced toward the right. The great tendency of ammonium hydroxide to break up into ammonia and water is the reason for the very feeble basic reac- tion of an aqueous solution of ammonia, for undoubtedly ammo- nium hydroxide, so far as it is formed, is extensively ionized, like the strong bases. This view is supported by various observa- tions, among them the neutral reaction of the ammonium salts of strong acids, such as the chloride and the nitrate, and also the alkaline reaction of the carbonate and the cyanide, in har- mony with the similar alkaline reaction of the corresponding salts of the alkali metals. Ammonium chloride, NH^Cl, sal ammoniac, is obtained from the ammonia liquor of the gas factories ( 112). The ammonia is expelled by warming and absorbed in hydrochloric acid; this solution is evaporated and the solid residue sublimed, whereby the salt is obtained in compact fibrous masses. It dissolves in 2.7 parts of cold, and in 1 part of boiling, water and crystallizes out of the solution in small, usually feather-like groups of octa- hedrons or cubes. It has a sharp saline taste. Ammonium chloride vaporizes easily, dissociating into ammonia and hydrochloric acid, as is shown by the vapor density, which at 350 is only half as great as calculated. This dissociation can be easily demonstrated in the following manner : Introduce into a tube sealed at one end a little ammonium chloride and, not far from this, a piece of blue litmus paper. In front of the latter is pushed a plug of asbestos wool and finally a piece of red litmus paper. The chloride is then heated. Since hydrochloric acid has a smaller diffusion velocity than ammonia the latter passes through the wad first and colors the red paper blue ; as a result an excess of hydrochloric acid is left at the other end and it reddens the blue paper placed there. It is a remarkable fact, discovered by BAKER, that perfectly dry ammonium chloride (having stood for a long time in a desiccator over resublimed phosphorus pentoxide) has the normal vapor density. On the other hand, the same investigator found that 334 INORGANIC CHEMISTRY. [234, similarly dried ammonia gas and hydrochloric acid gas do not unite to form ammonium chloride ( 38). Traces of water thus produce a marked catalytical acceleration, both of the formation and of the decomposition of ammonium chloride. We have here an illustration of the general rule that when one part of the system in a reversible reaction is accelerated by a catalyzer the other must be likewise affected. The proof of this rule lies in the impossibility of the contrary being true, since that would necessitate a change in the equilibrium (see 49). In many other cases it is also observed that traces of water have a considerable influence on the velocity of chemical reactions. The follow- ing examples may be cited: (1) Phosphorus, that ordinarily takes fire in moist air at a little above room temperature, can be heated in oxygen to 150 without ignition, provided the oxygen has been carefully dried by phosphorus pentoxide. (2) Carbon monoxide burns in moist oxygen much more easily than in dry oxygen. (3) Very carefully dried detonating- gas can be heated in a tube to red-heat without exploding. Ammonium sulphate, (NH 4 ) 2 S04, crystallizes in large rhombic prisms and dissolves very readily in water. On boiling the aque- ous solution some ammonia escapes, acid sulphate being formed. Its solution in 30% hydrogen peroxide yields on evaporation crystals of the composition (NH 4 ) 2 SO 4 -H 2 O 2 . When these are heated under re- duced pressure a high per cent hydrogen peroxide distils off. Ammonium nitrate, NH 4 N03, deliquesces in the air; when heated it breaks up into water and nitrous oxide ( 119). This salt is known in three crystallized modifications, the transition points ( 70) of which have been determined. Ammonium phosphates. The tertiary salt, (NH 4 ) 3 PO4, is deposited in crystalline form on mixing concentrated solutions of phosphoric acid and ammonia. It cannot be dried, however, for it then loses ammonia and goes over into the secondary phosphate, (NH 4 ) 2 HPO4. On boiling the solution the salt again yields ammo- nia and is transformed into the primary phosphate. The best known of these salts is the sodium ammonium phos- phate, NaNH 4 HP04 4H 2 O, microcosmic salt. It forms large transparent crystals. On being heated it melts, loses water and ammonia, and passes over into a vitreous substance, sodium meta- phosphate, NaPO 3 . Ammonium carbonate was formerly obtained by the dry distil- lation of nitrogenous organic substances, such as hair, nails, leather, etc., hence the name " salt of hartshorn," which still 235.] SALT SOLUTIONS. 335 clings to it. At present, however, it is made by dry distilling a mixture of calcium carbonate and ammonium chloride or sulphate. The product is a mixture (molecule for molecule) of acid salt, NH 4 HCOs, and ammonium carbamate, NH 2 -CO 2 -NH 4 (this latter being the neutral salt minus 1H 2 O). From its composition, (NH 3 ) 3 (CO 2 ) 2 -H 2 0, it takes the name ammonium sesquicarbonate, On passing ammonia gas into a concentrated aqueous solution of it the neutral salt, (NH 4 ) 2 CO3, separates out as a crystalline powder; it smells strongly of ammonia and passes slowly over into the acid salt, NH 4 HCO 3 , a white odorless powder, which is scarcely soluble in water. This acid salt is also formed directly from the sesqui- carbonate, as the latter gives off carbon dioxide and ammonia in the air (hence the odor of ammonia) and goes over into the first- named salt. Ammonium sulphide is extensively used in analysis ( 73). A solution of ammonium hydrosulphide (or sulphydrate), NH 4 SH, is obtained by saturating aqueous ammonia with hydrogen sulphide; it is a colorless liquid, which soon turns yellow because of the for- mation of ammonium polysulphides. The oxygen of the air oxidizes part of the hydrogen sulphide and thus sets free sulphur, which combines with ammonium hydrosulphide to form polysulphides. These polysulphides are also obtained by dissolving sulphur in a solution of ammonium hydrosulphide. On mixing 2 vols. NH 3 gas and 1 vol. H 2 S gas at 18 a white crystal- line mass is obtained, which decomposes at ordinary temperatures into NH 4 SH and NH 3 . The compound NH 4 SH separates out crystalline when hydrogen sulphide is passed into alcoholic ammonia. As low as 45 it is completely dissociated into equal volumes of NH 3 and H 2 S. SALT SOLUTIONS. 235. Every solid substance is soluble in every liquid; however, the proportion which dissolves can vary all the way from zero to infinity. If only an infinitesimal amount of the solid goes into solution, we say ordinarily that the substance is " insoluble " in the liquid ; there can be no doubt, however, that, if our means of inves- tigation were sufficiently improved and large enough quantities of Jiquid were taken, the solubility would be perceptible. This has already been demonstrated in many cases of so-called insoluble substances ( 210). Even when we confine our attention to aqueous solutions of salts (including acids and bases) we find the 336 INORGANIC CHEMISTRY. [235- same infinite difference in solubility that is observed between sub- stances in general. Substances such as sand, barium sulphate ( 262), silver iodide, etc., are " insoluble "; others, like sulphuric acid, are able to dissolve in any given amount of water. With regard to the solubility of salts the following practical rules are worth remembering : Potassium, sodium and ammonium salts are soluble. Normal nitrates, chlorates and ace- tates are soluble. Normal chlorides are soluble (except AgCl, Hg 2 Cl 2 , and PbCl 2 ). Normal s ulp h a t e s are soluble (except those of Ba, Sr, Ca, and Pb) . H ydroxides are insoluble (except those of the alkalies and alkaline earths) . Normal carbonates, phos- phates, and sulphides are insoluble (except those of the alka- lies). Basic salts are insoluble. Acid salts are soluble if the acid iteelf is soluble. The solubility, i.e. the maximum relative amount of salt that can go into solution, is a function of the temperature and the pressure. In the great majority of cases the solubility increases with the temperature. If the temperature is plotted on the axis of abscissas and the amount of salt which dissolves in one hundred parts of water is plotted on the ordinate axis, a solubility curve is obtained (Fig. 52) which shows at a glance the variation of the solubility with the temperature. For some salts, e.g. potassium nitrate, the solubility increases very rapidly with the temperature;" for sodium chloride it remains practically constant. In certain cases, such as those of calcium hydroxide and calcium sulphate (within certain limits of tempera- ture) the solubility decreases with rising temperature. These phenomena are connected, as has already been explained, with the heat of solution, i.e. with the caloric effect which accompanies the process of solution, and in the manner expressed by VAN'T HOFF'S principle of mobile equilibrium ( 103). In fact saltpetre, for instance, whose solubility increases very rapidly with the temperature (see Fig. 52) dissolves in water with a considerable absorption of heat. $. 236. The term heat of solution has various meanings. We are obliged to distinguish between (1) the caloric effect of dissolving a salt in a very large amount of water; (2) the caloric effect of dissolving a salt in an almost saturated solution; and (3) the total heat of solution, i.e. the whole caloric effect of dissolving a salt in water until the solution is saturated. As a rule these three magnitudes will have dissimilar values, indeed their algebraic signs may be opposite. This is the case, 236.] SALT SOLUTIONS. 337 for instance, with the compound CuCl 2 -2H 2 0; 1 g.-mol. dissolved fa 198 g.-mols. H 2 O at 11 gives a caloric effect of +3.71 Cal.; 19.56 g.-mols. in the same amount of water, 3.129 Cal. The heat of solution to which VAN'T HOFF'S principle applies is that of the salt in its saturated solution. We have here the system: 10" 20 U 30 * 50 60 70 C FIG. 52. SOLUBILITY CURVES. 100 salt M- saturated solution; when the temperature changes, the equi- librium is displaced, i.e. salt either goes into solution or crystallizes out, the latter action producing just as large a thermal effect numerically as dissolving in the saturated solution, but with the opposite sign. Since this was not taken into consideration when the matter was first dis- 338 INORGANIC CHEMISTRY. [236- cussed, it was believed that there were exceptions to the principle, but closer investigation has proved the contrary. In some cases the solubility of a salt at first increases gradually with rising temperature and then steadily decreases, so that the solubility curve has a maximum (cf. Fig. 53). In full agreement with VAN'T HOFF'S principle the heat of solution is negative in the ascending portion of the curve, zero at the maximum and positive in the descending portion. In the case of gypsum, CaSO4-2H 2 O, for instance, the maximum was found to lie at about 38 and at that point the heat of solution was actually proved to be 0.00; at 14 it is -0.36; above 35, +0.24. The effect of pressure on the solubility is at the most very slight, but it is in entire accord with the principle of LE CHATELIER. ^ Ammonium chloride, for instance, dissolves with expansion; therefore its solubility lessens t IG. 53. . , . _, . . with increasing pressure (1% for an increase of 160 atm.). Copper sulphate, which dissolves with contraction, has its solubility increased 3.2% by an increase of 60 atm. pressure. 237. It was formerly thought that the terms " solvent " and " dissolved substance " (" solute ") should be kept distinct. How- ever, it has since developed that there is no essential difference between the components of a solution, and that aqueous solutions are therefore better defined as " liquid complexes, one of whose components is water," than as " water in which substances are dissolved." The interchangeability of the terms " solvent " and " solute " is evidenced first of all by the phenomena attending the cooling of salt solutions. Let us consider, for instance, a nearly saturated solution of potassium chloride at a definite temperature. . We have in it two substances (KC1 and H 2 0) and two phases ( p.), hence two degrees of freedom. We will suppose that the solution is then cooled; potassium chloride crystallizes out forthwith and, as three phases are then present, the system becomes univariant We recall that changes in the quantity of any phase have no effect on such a system; therefore, if more salt is introduced into the system, the concentrations of saturated solution and vapor are unaffected. 237.] SALT SOLUTIONS. 339 This is none the less true when water is added or the vapor volume increased, so long as the three phases remain. On cooling still farther, more potassium chloride is gradually deposited until a point is reached below which the entire liquid congeals to a mixture of salt and ice. This point is known as the cryohydric, or eutectic, point. There are now four phases, salt, ice, solution and vapor; hence the system has become nonvariant. The opinion was formerly held that at this point a chemical com- pound between the salt and water (a ' ' cryohydrate ") came into exist- ence. That it is only a matter of mixtures can be seen in the case of colored salts (K 2 OO 4 ), for instance, with a microscope; moreover the composition of these so-called hydrates may differ in case the solidifi- cation takes place under a different pressure. If we start with a dilute potassium chloride solution as another example, and cool it, we have ice formed at a definite temperature and a univariant system established, ice being the third phase required. Below this point the solution can be regarded as saturated in respect to ice, just as it could be considered saturated in respect to the salt in the previous case; for an increase of the solid phase (ice) does not now cause a displacement of the equilibrium ( 71) any more than the addition of the solid (salt) did in the previous instance. The addition of potassium chloride causes part of the ice to go into solution (i.e. melt) ; for the dissolving of more salt increases the concentration of the solution. Therefore, if the temperature is kept constant, ice must melt in order to restore the solution to its previous concentration. It is therefore perfectly analogous to the addition of water to a satu- rated potassium chloride solution in contact with the solid salt, in which case also the solid phase goes into solution. If the tem- perature rises, more ice dissolves; if it falls, more crystallizes out just as with rising temperature more potassium chloride goes into solution .^nd with sinking temperature more crystallizes out. On farther cooling more and more ice will be deposited until, in this case also, the cryohydric point is reached, below which the whole system solidifies to a mixture of salt and ice. The analogy is therefore complete. The cryohydric point is, according to this view, the point of intersection of two curves, viz.: the solubility curves of salt and of ice in the salt solution. 340 INORGANIC CHEMISTRY. [ 237. Another argument against the assumption of any essential difference between solvent and solute is found in the behavior of the solutions of certain hydrous salts, e.g. CaCl2-6H 2 0. A satu- rated solution of CaCl 2 in water at 30.2 has exactly the com- position CaQ 2 -6H 2 0. At this temperature, therefore, the hydrate melts to a homogeneous liquid. // either H 2 or CaCl 2 is added, there is a deposition of CaCl 2 - 6H 2 on cooling, for the addi- tion of either causes a depression of the point of solidification (freez- ing-point) of CaCl 2 -6H 2 0. In the first case this hydrate is in equilibrium with a liquid which contains more water than the hydrate does and which is therefore called an aqueous solution in the ordinary sense. In the second it is in equilibrium with a liquid which contains more CaCl 2 than CaCl 2 -6H 2 and must therefore be regarded as a solution in CaCl 2 . On examining the solubility curves of various salts (cf. Fig. 52) it is found that they are in general regular; however, in one of the curves (sodium sulphate) a sudden change of direction is noticed. This is often observed with salts that contain water of crystalliza- tion. Taking sodium sulphate as an example, the phenomenon may be explained thus: It has already been remarked ( 225) that this salt has a transition point at the temperature of 33, Na 2 S04- 10H 2 O being transformed into Na 2 S04 and 10H 2 0. Up to 33, therefore, we have the hydrous salt as the solid phase; above this temperature the anhydrous salt. This change must necessarily involve a sudden bend of the solubility curve. Below 33 the curve represents the solubility of Na 2 SO 4 -10H 2 O, above 33 that of Na 2 S0 4 . We can therefore also regard the point of inflection of the curve (at 33 Q ) as the point of intersection of the curves for Na 2 S04-10H 2 and Na 2 S04. In sodium sulphate the special case appears where the sol- ubility of the anhydrous salt decreases with rising temperature and hence the solubility curve falls as the temperature rises above 33. In the light of the above the solubility of a substance which has a transition point is the same for both modifications at this point. This must always be the case; it can be demonstrated in the same way as in 70, where it was shown that the vapor pressures become equal at the transition point. Indeed the same figure can be employed, if it is borne in mind that the solubility of a metastable modification is always greater than that of the stable modification at one and the same temperature. Inversely, more- over, we have here a means of determining the transition point. 237-1 SALT SOLUTIONS. 341 In general, as OSTWALD has pointed out, the solubility of any substance whatever is dependent on the condition in which it exists. The solid phase determines the equilibrium, not only in virtue of its chemical composition but also by the particular modification in which the solid substance is present. Thus, e.g., each of the various forms of the same polymorphous substance or different hydrates of the same salt has its own solubility, other things being equal. In a hydrous salt we may have the case where there are various hydrates, which are connected with each other by transition points. A salt with m + n molecules of water of crystallization passes over at a definite temperature into another with m molecules, for ex- ample. The latter may, at a higher temperature, have a second transition point (to anhydrous salt). At each of these points the solubility curve will show a bend, because the solid phase changes; the curve will therefore assume some such form as t.^at of Fig. 54. Let us examine such a solubility curve a little more closely. At (A in F.g. rj4) we will suppose that we have pure water and ice, to start with, and JQO that small portions of salt are then gradually 1 dissolved. If the ice Eg phase is to be preserved, 2 o the temperature must be allowed to sink, fora 2 salt solution has a lower freezing-point than pure water. We therefore TEMPERATURE FIG. 54. move along the curve AK. Soon a point K is reached when no more salt dissolves, since all the water has now turned to ice. Here, therefore, we have a mixture of ice and solid salt, or, in other words, the cryohydric point. If we wish to bring more salt into solution after K is reached the temperature must be raised. The ice phase then, of course, disappears and in its place we have the salt with m+n molecules of water of crystallization as solid phase. If the temperature is steadily raised and the solution is kept constantly saturated by the addition of this salt, we move along the curve KB. At B, however, we meet the transition point from the salt with m+n mols. H 2 O to the one with m mols. H^O; hence the solubility 342 INORGANIC CHEMISTRY. [ 237. curve must again bend here and in such a way that atthepoint B the solubility curve of the salt with m + n mols. H 2 O is steeper than that of the salt with m mols., no matter what the form of the curves KB and BC may be. This is readily understood by a course of reasoning entirely analogous to that given for the transition of ice to water or of rhombic to monoclinic sulphur ( 70). Finally at C we have a second transition point from salt with m mols. H^O to anhydrous salt, so that the solubility curve there shows one more bend. Where the curve CD ends depends on circumstances. In many cases, e.g. that of silver nitrate, it ends at the melting-point of the anhydrous salt (concentration of the solution = 100%). In other instances the anhydrous salt can form a second (fused) liquid layer under the saturated solution. Finally, mention may also be made of the case of copper sulphate, which at a given temperature loses its water of crystallization in contact with its saturated solution and from that point .on shows a decrease in solubility with rising temperature, which finally ends in almost total insolubility. If we draw a line kik 2 through K parallel to the ordinate axis, the figure is divided by this line and AKBCD into the following regions: To the right of the solubility curve is the region of the unsaturated solutions, AKk 2 is that of the superfused, kKBCD that of the supersaturated, solutions. To the left of k\k 2 only ice + solid salt can exist under ordinary pressures. A solubility curve, such as that represented in Fig. 54, can, on the other hand, be used to detect the existence of compounds between the salt and the water. From the cryohydric point upward every bend in the solubility curve shows that a salt with a different amount of water of crystallization has been formed. Each branch of the curve thus represents a separate salt, i.e. a different solid phase. The composition of these solid phases is by no means always self-evident. Such is, however, the case when the phase fuses without altering its composition, or, what amounts to the same thing, when it can exist in equilibrium with a liquid phase of the same composition. This does not often occur with salts in aqueous solution, but an example of it was seen above in the case of CaCl2-6H 2 0. The inspection of the solubility curve or melting-point curve is then especially valuable for the discovery of compounds. In order to understand this let us first examine a system of two substances, A and B, which do not combine. Fig. 237.] SALT SOLUTIONS. 343 55, represents the melting-point curve that one obtains on the addition of increasing amounts of B to A. At first the melting- point sinks until the eutectic point * E is reached. Along AE A alone separates out of the fused mass on freezing. At E, however, B also separates out. If more of B is now added the melting- point rises; we obtain the curve EB, which terminates in the melting-point of pure B. Along EB only B separates out of the fused mass. 100 FIG. 55. 20 so IB w FIG. 56. 100 Suppose we now assume that A and B form a compound AB in the molecular proportions 1:1 (Fig. 56). On the addition of B to A AB is formed and dissolves in the excess of A. This lowers its melting-point. When a certain amount of B has been added this point is lowered to E\. Here both A and AB separate out. EI is the eutectic point for mixtures of A and AB. If more of B is added the melting-point rises, just as in the case where there is no combination between the components. Only AB now separates out of the fused mass. The continued addition of B, however, increases the amount of the compound AB; at M free A has disappeared and the mass consists wholly of pure AB, whose melting-point is M. At this point the melting-tem- perature reaches its maximum, for the addition of either A or B lowers the melting-point of the pure compound. The further * The term "eutectic point" is more general than "cryohydric point," the latter term being usually restricted to aqueous solutions. "Eutectic mix- ture" and "cryohydric mixture" ("cryohydrate") are similarly related. 344 INORGANIC CHEMISTRY. [237 course of the curve is readily seen. As more and more B is added to AB the melting-point sinks, AB alone separating out, until the eutectic point E 2 is reached, where both AB and B crystallize out, and thereafter the melting-point again rises along E 2 B till it finally ends in the melting-point of the pure substance B. If more than one compound is formed between A and B, each one will cause a maximum point in the curve, i.e., each maximum will correspond to a compound. The following examples will serve to make this clear: 1. The system S0 3 +H 2 0. Here there are a number of hy- drates, which are indicated by the melting-point curve (Fig. 57). /\ \ / 1 62 05 -35 FIG. 57. A mixture of 62% SO 3 + 38% H 2 O has a freezing-point of 20. On the addition of sulphur trioxide this point rises till at the com- position SO3 + 2H 2 O=H 2 SO4-H 2 O it reaches its first maximum. At this temperature (8) the whole mass solidifies, yielding crystals of the above composition. The further addition of sulphur tri- oxide has the same effect on the melting-point of the hydrate H 2 S04-H 2 O as the ordinary addition of a foreign substance to a pure substance. At first the melting-point falls; that which crystallizes out is the hydrate H 2 S04-H 2 0. At a composition of about 3H 2 O + 2SO 3 a eutectic point is reached (237). The mass solidifies at 35 to a mixture of H 2 S04-H 2 O and H 2 S04. Continued addition of sulphur trioxide causes a rise of the melting- point till at +10 a second maximum is reached, where the whole solidifies to a homogeneous mass, consisting of pure sulphuric 237.] SALT SOLUTIONS. 345 acid, H 2 SC>4. Along this ascending branch of the curve H 2 S0 4 crystallizes out. The melting-point curve then proceeds to a third maximum point, corresponding to the hydrate H^SC^+SOs = H 2 S 2 O7 (pyrosulphuric acid), and comes to an end in the melting- point of the asbestine form of sulphur trioxide at +40. 2. In non-aqueous liquids the relations are almost exactly the same, as may be seen from a consideration of the system S + C1. It was remarked in 75 that, while the compound SCU could not be isolated, the form of the melting-point curve left nc 100 90 80 70 60 50 40 30 20 10 S Atomic Percentage of Sulphur FIG. 58. doubt as to the existence of such a compound. This curve has a maximum at the point C 2 (Fig. 58), corresponding to 20 atomic per cent, sulphur, or to the molecular formula SCl^ The points EI, E 2 , E$ are the eutectic points for S + S 2 C1 2 , S 2 C1 2 + SC1 4 and SC1 4 -1-C1 2 , respectively, while the maximum C\ corresponds to the compound S 2 C1 2 . Supersaturated solutions. A sodium sulphate solution satu- rated a little below 33 can, if carefully guarded from contact with any of the solid salt, be cooled down to room temperature without anything crystallizing out, but contact with the tiniest crystal 346 INORGANIC CHEMISTRY. [5 237 fragment of Na 2 S0 4 -10H 2 O is sufficient to cause a sudden crystalli- zation of this salt. Sodium sulphate is only one of a large number of salts capable of forming solutions of this nature. Sodium thiosulphate and many of the nitrates are other good examples. Such solutions are called supersaturated. They are perfectly stable; neither rubbing with a glass rod nor shaking (which treatment ordinarily tends to induce crystallization) causes the formation of crystals, provided no trace of the solid salt comes in contact with the solution. Such a system, which is unstable under only one condition, is called a metastable system. (See page 108.) If a supersaturated solution of sodium sulphate is cooled down below room temperature, another hydrate crystallizes out, viz., Na 2 SO 4 -7H 2 0; the resulting system is still metastable, however, for contact with the slightest trace of Na 2 S0 4 -10H 2 O suffices to convert it entirely into the stable system, with the deposition of Na 2 S0 4 -10H 2 0. The smallest amount of salt (crystal nucleus) that is sufficient to disturb and thus cause the disappearance of a metastable system, such as is represented by a supersaturated solution, is a quantity of about the order 10~ 10 g.> according to OSTWALD. The extreme minuteness of this amount explains why a spontaneous disappearance of the meta- stable condition was formerly regarded as possible. Inasmuch as very small bits of crystals are always floating in the air (especially in labora- tories), it is usually only necessary to open a bottle containing a super- saturated solution or to rub the sides with a glass rod (which always has crystal fragments on its surface) , in order to excite crystallization into the stable system. 238. For the reasons stated in 65 and 66 it is assumed that acids, bases and salts in aqueous solution are split up into ions. This dissociation can be more or less complete, according to the nature of the solute, the temperature of the solution and its con- centration. Examples of this have already been mentioned here and there in the text; hydrochloric and nitric acids in tenth-normal solutions are almost completely dissociated, carbonic and silicic acids scarcely at all. Among the bases the hydroxides of potas- sium, sodium and the alkaline earth metals are almost completely dissociated at this dilution. A similar difference is shown by 238.] SOLUTIONS. 347 salts ; those of the alkalies are practically completely ionized, while mercuric chloride is very slightly so. We shall return to this sub- ject more in detail in the discussion of the metals. The principle that solutions containing equivalent amounts of different electrolytes differ greatly in conductivity and hence in degree of ionization can be demonstrated in an elegant manner with the aid of an apparatus devised by WHITNEY (Fig. 59). FIG. 59. Four glass cylinders (3 cm. diam.) are fitted each with two horizontal platinum disks (copper can be used but is less satisfactory) to serve as electrodes, the upper ones being movable. Each lower electrode is connected with an incandescent lamp and the apparatus as a whole with the terminals of a (preferably) alternating 110-volt circuit. After placing in each of the tubes 120 cc. distilled water they are filled with 5 cc. of half-equivalent-normal hydrochloric, sulphuric, monochlo- racetic and acetic acids respectively. On making the distance between the electrodes alike in all the cylinders, the lamp beneath the hydro- chloric acid is found to glow brightest, since the resistance of this solu- tion is the least. The other lamps follow in brightness in the order given above. The electrodes are next adjusted so that all the lamps are equally bright, when it is seen that while the electrodes in the hydro* 348 INORGANIC CHEMISTRY. [ 238 chloric acid are farthest apart, those in the acetic acid are almost in contact. In order to show that the alkali salts of these acids, unlike the acids themselves, have nearly the same conductivities and degrees of disso- ciation the solutions are just neutralized with potassium hydroxide and the lamp test repeated. The lamps are equally brilliant when the electrodes are at approximately the same height. 239. In the solution of an extensively ionized salt we should expect to find the properties of the cation and the anion. It must exhibit the sum of the properties of the two ions, or, to use other words, its properties must be additive with reference to those of both ions. This is actually the case, both physically and chemically. As for the chemical properties, we observe that solutions of different salts of the same metal all give the same reactions ; from the solutions of all lead salts, for instance, hydrogen sulphide precipitates black lead sulphide, sulphuric acid white lead sulphate, etc. Similarly the solutions of salts of the same acid are all characterized by the same reactions; sulphates, for example, by the precipitate they give with barium chloride solution. All this appears very strange when we recall that the solid salts are markedly different from each other in their properties, but we are forced to just such a con- clusion when we assume that the salts are ionized in solution. Among the physical properties additivity is very apparent in the color of salts of colored acids and bases. OSTWALD found that all permanganates with a colorless base, when prepared in equiva- lent solutions, give exactly the same absorption spectrum ( 263). All dilute copper solutions are blue. In the permanganates it is the anion MnO 4 ', in the copper solutions the cation Cu", which is to be regarded as the color-carrier. When the solvent is one in which ionization does not occur, the salts of the same base may differ widely in color. For instance, a solu- tion of cobalt nitrate in alcohol is purple, that of the chloride is bluish violet; but if both are poured into an excess of water, the solution becomes pink in each case. Another example is found in the alcoholic solutions of cupric chloride and nitrate; the former is dark green, the latter blue; on the addition of water both become blue. This additive nature manifests itself in various other physical properties also. But since it cannot usually be shown directly 239.J SALT SOLUTIONS. 349 (as in the case of colored salts), we have to approach the matter somewhat indirectly, as the following example will illustrate. The specific gravity of a sugar solution can be represented fairly accurately by the formula n being the number of moles per liter and K a constant. Similarly in the case of the solution of a highly ionized salt whose specific gravity is raised to l+an by the anion, to 1 +pn by the cation, a and /? being constants, the specific gravity of the solution, if we assume additivity to exist, must be The values of a and /? are as yet unknown. For salts with the same anion the specific gravity is expressed by For salts with the same anions as in the former case but with a different cation the specific gravities of their solutions are repre- sented thus: Si'=l+w(ai+/?i); S 2 '=l+n(ai+p 2 ); S 3 '=l +n(>i +&), etc., whence it follows that the differences Si Si', 8282', Sz Ss'r = n(a ai), must always have the same value in case additivity really exists. The equality of these differences can therefore be used as a proof of additivity. A concrete example of the above reasoning is to be found in the specific gravity values of the solutions KC1, NaCl, NH 4 C1 and KBr, NaBr, NH 4 Br. Here we actually have the relationship: KC1- KBr - NaCl- NaBr = NH 4 C1 -NH 4 Br ; however, not simply in specific gravity but with reference to other physical constants as well. Compressibility, capillarity and re- fractive index, for example, have been found to conform to this same additive scheme. The ionization hypothesis also leads us to predict that when dilute solutions of strong acids and bases, each containing one 350 , INORGANIC CHEMISTRY. [ 239- mole, are mixed, the same caloric effect will be observed. This is found to be the case (13.8 cal.). The only change that takes place in the mixing is the formation of water from its ions ( 66). Further, the so-called law of thermo-neutrality, which says that when two dilute salt solutions are mixed there is no caloric effect, is a natural consequence; for the ions of the two salts exist in the free state both before and after the mixing. ACIDIMETRY AND ALKALIMETRY. THEORY OF INDICATORS. 240. The amount of acid or base present in a liquid can be determined most simply by volumetric analysis ( 93). Those parts of volumetric analysis which comprise the methods used for this purpose are known as acidimetry and alkalimetry. Suppose that we wish to determine the amount of hydrochloric acid present in a given volume of liquid. A known volume of this liquid (50 cc., 10 cc., or less, according to the supposed concentration) is meas- ured out and sodium hydroxide solution of known concentration is slowly added from a burette. When the point has been found at which the liquid becomes neutral, it is easy to calculate the con- centration of the acid from the number of cubic centimeters of sodium hydroxide consumed. Example. Determine the amount of nitric acid present in a liter of a solution of this acid if 10 cc. are neutralized by 7.3 cc. of a normal alkali solution. These 7.3 cc. are equivalent to the same number of cubic cen- timeters of normal nitric acid. Therefore the 10 cc. contain 7.3 milli- gram molecules of nitric acid or 63x7.3 mg. One liter must contain a hundred times as much, or 45.99 g. Before we can determine the concentration of an acid or an alkali in this manner, we must first possess an alkali or base solu- tion of known concentration and further have a delicate means of detecting when the liquid is exactly neutralized. 1. Preparation of an acid and an alkali of known concentration. This can be done in various ways. Oxalic acid, C2H 2 O4-2H 2 O, succinic acid, C 4 H 6 O 4 , or tartaric acid, C 4 H 6 6? can be used as the basis, for all of these are crystallized solids and can be easily obtained in a state of sufficient purity; hence the amount of acid dissolved 240,] ACIDIMETRY AND ALKALIMETRY. 351 can be very accurately determined by previously weighing the substance on an analytical balance. We thus weigh out 1 g.-equiv- alent (J g.-mol.) of one of these acids, dissolve it in water and dilute to exactly a liter. Thereupon with the help of this normal acid a normal alkali is prepared; a little more than 1 or ^ or ^, etc., gram-equivalent of sodium hydroxide or potassium hydroxide (barium hydroxide is also very satisfactor}^) is dissolved in water and this solution is standardized according to the normal acid, i.e. the concentration is determined by titration with normal acid and then diluted so that it is just normal. Sodium carbonate can also be used as a basis. After being first heated in order to expel all moisture it is weighed out and dissolved in water. This solution is heated to boiling and covered with a glass plate with a hole in it, through which the nozzle of a burette is passed. The solution of the acid whose concentration is to be determined is then allowed to flow from the burette into the boiling liquid till neu- tralization is effected. Carbon dioxide escapes, but the glass plate prevents any loss of the liquid by spurting. The standardizing can also be accomplished by adding the acid solution that is to be standardized to a mixed solution of potassium iodide and potassium iodate. Hydriodic and iodic acids are set free and they react at once in the following manner: 5KI + KI0 3 + 6HX =5HI +HI0 3 + 6KX; SHI +HI0 3 =3H 2 + 61. Thus for every equivalent of acid one atom of iodine is set free. By titrating with sodium thiosulphate the amount of iodine liberated can be determined. This method gives very accurate results. 2. Determination of the point at which the liquid becomes neutral. Since the point of neutralization of an acid by a base or vice versa, is not indicated by any visible phenomena, a minute quantity of some substance is added whose color is altered by an excess of the neutralizing liquid. Such substances are litmus (blue in alkaline and red in acid solutions), phenolphthalem (red in alkaline, color- less in acid solutions), methyl orange (yellow in alkaline, red in acid solutions), and many others. Therefore, on gradually adding an a a 1 1 ne solution to an ^^. solution in the presence of one of acid alkaline 352 INORGANIC CHEMISTRY. [240^ these substances a change of color will be noticed when the point of neutrality is just passed. Coloring-matters like the above are termed indicators. The change of color is due in many cases to a transformation of the substance into a salt whose free acid is very unstable and passes over almost immediately into an isomer hav^ ing a different color from the free acid or the salt. 241. From the standpoint of the ionic theory the following theory of indicators has been advanced: If a couple of drops of the indicator are introduced into an acid solution, the ionization of the indicator, which is only very slight, is reduced by the great excess of acid to practically zero. If a base is then added, the H-ions of the acid to be titrated are removed by the OH-ions. However, if the acid is very strong, enough H-ions remain in the liquid up to the last to prevent anything like an extensive ioniza- tion of the coloring-substance; not until the first excessive drop of alkali is added do the anions of the coloring-substance come into existence, the alkali compound of the latter being strongly disso- ciated. The change of color is therefore sharply defined, for it is due to this difference in color of the non-ionized molecule and the anion. On the other hand, if the acid is a weak one, there will not be enough H-ions present when the end of the titration is nearly reached to prevent a slight ionization of the coloring-substance. As a result we shall have in the solution not only the undissociated coloring-substance but its anions as well, even before the titration is completed, in other words, the change of color becomes more gradual and hence the end reaction more indefinite. The effect will be the same if the alkali employed contains carbonate. In that case near the end of the titration the solution will only contain carbonic acid, which is very weak; consequently the color change will not be sudden. It is for this reason that in titrating soda solu- tions (see 240) the carbonic acid must be expelled by boiling. If a weak acid is to be titrated, it is necessary, according to the above, to select an indicator which is much less ionized even than the acid itself and whose alkali salts are sufficiently ionized to produce a distinct change of color. A very suitable one for this purpose is phenolphthalem. Acetic acid, for example, can be sat- isfactorily titrated with it, if a strong base is employed, for the reasons set forth above. On the other hand, in case a weak base is to be titrated, phenolphthalein is not so satisfactory. Ammonia 242.] COPPER. 353 does not color a phenolphthale'in solution till a considerable excess is added, because at the great dilution in which the ammonium- phenolphthalein compound exists in a titration it is almost com- pletely split up by hydrolysis (66). If a weak base is to be titrated, an indicator must be selected which is a relatively strong acid, for then the salt of the coloring- substance will be hydrolyzed only to a limited extent, even near the termination of the titration (i.e. when the concentration of the base has become weak), and the color of its ions will therefore still predominate. For such a titration a strong acid (e.g. hydro- chloric or sulphuric acid) must be used, in order that the first drop after the point of neutralization is reached may diminish the elec- trolytic dissociation of the coloring-substance and so give the solu- tion the color of the non-ionized molecules. Methyl orange is an indicator that answers these requirements; it serves very well in the titration of ammonia. All other indicators are intermediate to these two extremes (phenolphthalem and methyl orange) as re- gards their ionization, and their applicability is determined accord- ingly. COPPER. 242. This metal occurs native in America, China and Japan, forming regular crystals. Other copper minerals are cuprite (Cu 2 0), malachite and azurite (both basic carbonates), chalcocite (Cu2S) and particularly chalcopyrite, or copper pyrites (CuFeS2). The United States furnishes about 60% of the world's copper supply. The extraction of the metal from non-sulphurous ores is very simple. They are smelted with coal and thus reduced to the metallic state. If the copper ore contains sulphur, the metal- lurgical process is much more complicated and has numerous modi- fications. The ore is broken up and " calcined " so as to convert some of the copper sulphide into copper oxide. Thereupon it is fused with sand and other siliceous fluxes (as well as coal for reducing copper sulphate), and the iron, but not the copper, is converted into silicate. The object of the flux, here as with other metals, is to lower the fusing temperature of the ore and collect the impurities (iron in this case) into a " slag " consisting of fused 354 INORGANIC CHEMISTRY, silicates, etc. The slag floats and can be run off. The fusion process is repeated until all the iron is eliminated. The resulting mixture of impure copper sulphide and copper oxide is called matte (also regulus and coarse metal). By repeated roasting and fusing, crude metallic copper is obtained: 2Cu 2 O + Cu 2 S = 6Cu + S0 2 . Finally it is fused with coal to reduce any copper oxide remaining. Refining. The copper thus obtained often contains small quantities of other metals. Since these impurities lower its conductivity, a better grade is demanded for electrical purposes. Crude copper is now refined by an electrolytic process which yields chemically pure copper. If an impure copper solution is electrolyzed, it is possible under suitable con- ditions to precipitate pure copper in a compact mass, while the impuri- ties remain in solution or are deposited as powder. From this powder ("slimes ") a considerable amount of gold and silver is obtained. The usual arrangement is to suspend plates ("anodes") of crude copper and thin sheets of pure copper alternately in a copper vitriol solution acidified with sulphuric acid. If the crude plates are then connected with the positive pole and the thin sheets with the negative pole of the dynamo current, pure copper is deposited on the sheets, while an equivalent amount of the crude copper dissolves to take its place. Physical Properties. Copper has a bright red color. It is rather hard but very extensible and flexible; it can be drawn out into very fine wire and beaten into extremely thin sheets (imitation gold-leaf), which are green in transmitted light. Sp. g. = 8.94; melting point =1083; boiling-point, 2310. Chemical Properties. In dry air copper is permanent at ordi- nary temperatures, but in moist air it becomes covered with a thin coating of basic copper carbonate, which protects it from further rusting. On being heated in the air it turns to copper oxide, CuO. It is readily attacked by nitric acid ( 120), but not by dilute hydro- chloric acid. Sulphuric acid has no effect on it at ordinary tem- peratures, but at higher temperatures a reaction takes place in which sulphur dioxide is given off ( 78). Ammonia and oxygen dissolve it to form a blue liquid/ copper oxide ammonia. Copper is deposited from solutions of its salts by iron, magnesium and other metals. 243.] COMPOUNDS OF COPPER. 355 Uses and alloys. Copper finds extensive use in the arts, both as such and in alloys. The well-known yellow b r a s s is an alloy of 1 part zinc and 2 parts copper and is harder than copper itself. German silver consists of about 50% Cu, 25% Ni, and 25% Zn; its electrical conductivity is affected very little by changes of temperature, which makes it valuable for resistance coils, etc. For bronzes see 199. Copper is employed in large quantities inelectrotyping. A cast is first constructed of plaster of Paris and made a conductor by being coated with graphite, whereupon it is suspended by the wire of a battery into a copper sulphate solution; a plate of pure copper serves as the anode. If the potential difference at the electrodes is properly regulated, the copper is deposited on the plaster cast in compact form, so that all the details of the original are reproduced with the greatest fidelity. Compounds of Copper. 243. Copper forms two sets of salts, which are derived from the oxides Cu2O, cuprous oxide, and CuO, cupric oxide. CUPROUS COMPOUNDS. Cuprous oxide, Cu2O, can be obtained from cupric salts in various ways, e.g. by reducing them in alkaline solution with grape sugar, hydroxylamine, arsenious acid, or the like. It forms a reddish-yellow crystalline powder, which is unaffected by the air at ordinary temperatures. When cuprous oxide is heated, it breaks up into cuprous oxide and oxygen; 2CuO^Cu2O + O. At 1025 the dissociation tension of the cupric oxide reaches 150 mm. ; consequently at this temperature in the air it passes over completely into cuprous oxide, since the partial pressure of the oxygen of the air is = 152 mm. Cuprous oxide dis- 5 solves in ammonia; this solution rapidly turns blue because of the absorption of oxygen, the cuprous oxide going over into cupric oxide. Cuprous oxide is transformed by sulphuric acid into copper sulphate and copper: Cu 2 + H 2 SO 4 = CuSO 4 + Cu + H 2 0. 356 INORGANIC CHEMISTRY. [ 243. It is possible that cuprous sulphate is first formed and that the cuprous ions of this solution are forthwith changed into cupric ions and non-ionized copper: Of the cuprous salts the sulphate and halides are known. Cu 2 Cl 2 , Cu 2 Br 2 , and Cu 2 I 2 are all " insoluble " (cf. 235) ; their solubility decreases with increasing atomic weight of the halogen. Cuprous chloride, Cu 2 Cl2 (the vapor density indicates this doubled formula), separates out when a solution of cupric chloride is boiled with copper, or when a mixed solution of copper sulphate and sodium chloride is saturated with sulphur dioxide gas and the resulting liquid poured into water. It is a white crystalline sub- stance, which must be kept under water, for it absorbs oxygen rapidly when moist and turns green because of the formation of basic copper chloride, CuCl-OH. It melts at 430 and distils at about 1000. It is soluble in concentrated hydrochloric acid and in ammonia. These solutions are at first colorless but very soon become blue because of the absorption of oxygen (formation of cupric compounds). They also have the power of absorbing car- bon monoxide, forming an unstable compound, Cu 2 Cl2 CO 2H 2 O which crystallizes in colorless laminae. Use is made of this prop- erty in gas analysis. Cuprous iodide, Cu 2 I 2 , is formed when a solution of copper sulphate is treated with potassium iodide, half of the iodine being liberated: 2CuS0 4 +4KI =2K 2 S0 4 +Cu 2 I 2 +I 2 . It may be supposed that cupric iodide is first formed and that it then breaks up into cuprous iodide and iodine, or, rather, that the ions of cupric iodide interact thus: Cu"+2I'=CuI+I, the cuprous iodide being unionized because "insoluble." According to OSTWALD, however, an equilibrium is formed here, for, though the cuprous iodide is but slightly soluble, the reaction does not complete itself and some cupric ions still remain in solution. The 243.] CUPROUS COMPOUNDS. 357 reversibility of this reaction is evident from the fact that cuprous iodide is dissolved by an alcoholic iodine solution, so that we have Cu" + 2I'<=CuI+I. Therefore, in order to make the precipitation more complete, a sub- stance (S0 2 ) is added, which will remove the iodine, one of the reaction products. This treatment is especially effective because the iodine is thereby converted into ions and this raises the concentration of one of the components on the left side of the equilibrium equation. Cuprous cyanide, Cu2(CN)2, can be obtained in a manner analo- gous to that described for cuprous iodide, viz., by mixing solutions of copper sulphate and potassium cyanide. Half of the cyanoge^ escapes as gas: 2CuS0 4 +4KCN=2K 2 S0 4 +Cu 2 (CN) 2 + (CN) 2 . Cuprous cyanide dissolves very rapidly in an excess of potas- sium cyanide, forming a salt, 2KCN-Cu2(CN)2, which contains a complex anion [Cu 2 (CN) 4 ]". Practically all of the copper ions go to form these complex ions on the addition of potassium cyanide, for the solution gives none of the ordinary reactions for copper, not even that with hydrogen sulphide, although copper sulphide is precipitated by this reagent even when the concentration of the copper ions is very slight ( 73). Cuprous sulphate, Cu 2 S04, is formed by the action of methyl or ethyl sulphate on cuprous oxide in the absence of water at 160: Cu 2 + Me 2 SO 4 = Cu 2 SO 4 + Me 2 O. Water decomposes it rapidly with evolution of heat : Cu 2 SO 4 (solid) + Aqua = CuS0 4 (dissolved) + Cu (solid) +21 cals. f This explains ( 101) why many attempts to prepare the salt resulted in failure and it was for a long time thought incapable of preparation. 358 INORGANIC CHEMISTRY. [244- CUPRIC COMPOUNDS. 244. Cupric oxide, CuO, is a dense, black powder, obtained by heating copper in the presence of oxygen at a high temperature. It can also be prepared by heating the nitrate to redness or igniting the hydroxide or the carbonate. When finely divided it occludes on its surface large amounts of steam. It finds extensive use in organic analysis. ^Cupric hydroxide, CuO-nH 2 0, separates out as a flocculent, voluminous blue precipitate (hydrogel, 195) when the solution of a copper salt is treated with caustic potash or soda. On boiling the liquid it turns black, water being liberated and cupric oxide formed. Cupric chloride, CuCl 2 2H 2 0, is obtained by dissolving cupric oxide or carbonate in hydrochloric acid. It crystallizes in blue rhombic needles, which, however, often appear green because of mother-liquor adhering to them. It is readily soluble in water and alcohol. The anhydrous salt is yellow; the concentrated aqueous solution is green; the dilute solution is blue. This differ- ence can be attributed to the breaking up of the salt into its ions, for all dilute copper solutions are blue, no matter what the acid radical is. It therefore follows that the copper ion imparts a blue color to solutions. The green color of a concentrated solution seems most probably due to the formation of complex ions, such as (CuCl 3 )'. Cupric bromide is analogous to the chloride; cupric iodide is unstable, decomposing at once into iodine and cuprous iodide ( 243). Copper sulphate, CuS0 4 5H 2 0, blue vitriol, is the most familiar salt of copper. It is obtained as a by-product, chiefly from gold and silver refineries, and is also manufactured by dissolving copper in sulphuric acid. It crystallizes in large blue triclinic crystals, which lose four molecules of water at 100; the fifth is liberated at 200. The anhydrous copper sulphate is a white powder, which absorbs water greedily, turning blue again. At 20 100 parts H 2 dissolve 42.31 parts of the crystallized sulphate. Blue vitriol is employed in large quantities in electroplating, etc. ( 242). 245-] SILVER. 359 Copper nitrate, Cu(N0 3 ) 2 , can crystallize with three or six molecules of water and is dark blue. Copper carbonate. The normal salt is unknown, but basic salts have been prepared. Copper arsenite, CuHAs0 3 , is used as a pigment under the name of Scheele's green. Schweinfurth green, or Paris green, is a double compound of copper arsenite and copper acetate. Since both are very poisonous, their use in dyeing textile fabrics, wall- paper, etc. ( 157) is being restricted. Copper sulphide, CuS, is formed as a black precipitate by pass- ing hydrogen sulphide into a copper solution. When moist it oxidizes slowly in the air to copper sulphate. On being heated in a current of hydrogen it yields cuprous sulphide, Cu 2 S, and hydrogen sulphide. Copper salts and ammonia. On mixing a solution of ammonia with a copper salt, a precipitate of copper hydroxide is first formed, if not too much ammonia is used; this precipitate is dissolved by an excess of ammonia to a dark blue solution. If the latter is evaporated or treated with alcohol, ammoniacal compounds crys- tallize out; a typical one is CuSO4-4NH 3 -H 2 O, which is trans- formed into CuSO4-2NH 3 on being heated to 150. The aqueous solutions of these substances are to be assumed to contain complex ions of copper and ammonia, since they do not give some of the ordinary copper reactions, e.g. precipitation with caustic potash. The fact that certain other reactions of copper do however appear, e.g. precipitation with hydrogen sulphide, proves that free copper ions are still present in the liquid, although only to a small extent. SILVER. 245. This metal occurs native; nuggets weighing 100 kilos are not unknown. The important silver ores are argentite, Ag 2 S, stromeyerite, Cu 2 S-Ag 2 S, pyrargyrite, 3Ag 2 S-Sb 2 S 3 , and stephanite, Ag 5 S4Sb. It is also found in smaller amounts in cerargyrite, or horn silver, AgCl. Traces of silver compounds are known to exist in sea-water. Many lead ores, e.g. galenite, contain a small per- centage of silver and in some cases it is extracted. The chief silver-producing countries are the United States (Colorado and neighboring States), Mexico, Australia and Bolivia. 360 INORGANIC CHEMISTRY. [ 245. The present annual world's production of silver is about 55,000,000 troy ounces (1,710,000 kg.). Silver is now generally obtained from its ores by the cyanide process. In this process the pulverized ore is allowed to stand for some time in a weak solution (0.1-0.4%) of sodium cyanide. The silver sulphide dissolves as the double cyanide according to the equation: Ag 2 S + 4NaCN= 2AgNa(CN) 2 + Na 2 S. As soon as a certain amount of silver has gone into solution, an equilibrium is formed, because the Na 2 S tends to react back- ward with the formation of silver sulphide. This is avoided by blowing air into the solution, and thereby oxidizing the sodium sulphide. Metallic silver is also taken up by the sodium cyanide solution ( 248) ; so is the chloride, or horn silver. The recovery of the silver from the double cyanide solution is accomplished by precipitation with zinc, or by electrolytic deposition. Lead ores usually contain some silver. In the smelting of lead the silver all goes into the lead and is recovered from it in the following way: The argentiferous lead is fused and then cooled slowly till it begins to congeal. Just as pure ice crystallizes out of a dilute salt solution on cooling, so the lead separates out here in crystals free from silver. These are removed and this process called Pattinsonizing after its inventor is kept up till the percent- age of silver reaches about 1%. This rich lead is then subjected to the cupellation process, i.e. the lead is fused in a rever- beratory furnace, whose hearth consists of a porous mass (cupel, or test). The lead is oxidized to the easily fusible oxide PbO (litharge) which is partly driven off by a blast from time to time through the channel provided for its escape, and partly absorbed by the porous material (bone-ash, or clay and limestone) of the cupel. Towards the end of the process the film of litharge remain- ing becomes so thin that the silver beneath reflects the light, pro- ducing a beautiful iridescence. Here and there the film breaks, disclosing the brilliant surface of the metal (" brightening " of the silver). The silver is finally left in the metallic state. Another method (PARKES') involves the principle of distri- bution between two slightly miscible solvents (ORG. CHEM., 24). Molten zinc and lead are only slightly soluble in each other. Silver is several hundred times more soluble in molten 245.] SILVER. 361 zinc than in molten lead; thus the silver can be very fully extracted from the lead by fusing with zinc. The process is as follows: To the fused argentiferous lead some zinc (containing 0.5% Al) is added and the mixture is stirred. The zinc takes up most of the silver from the lead and floats on the molten mass. It is skimmed off and cast into plates, which serve as the anodes in the subsequent electrolysis. At the cathode nearly pure zinc is deposited, while silver powder (70-80% Ag, the rest Pb) sinks to the bottom and is removed to the cupel. The electrolytic refining of silver is now carried on exten- sively in America. A great deal is recovered from the copper slimes ( 242). The pure silver of commerce usually contains a little copper and other metals; STAS obtained it chemically pure by dissolving the product of the smelter in nitric acid and precipitating it with hydrochloric acid as the chloride; this was then reduced by boiling with dilute caustic potash and milk sugar and finally distilled with the aid of an oxy hydrogen flame in an apparatus made of lime. Physical Properties. Silver crystallizes in regular octahe- drons. It has a white color and a high lustre. It is the best conductor of heat and electricity of all the metals and it is very malleable and ductile. Sp. g. = 10.5; m.-pt. = 961; boiling- point =1955. It volatilizes in the form of a blue vapor (STAS). Molten silver absorbs oxygen, but allows it to escape on becoming solid ( 9). Chemical Properties. Silver is one of the precious metals; this term is applied chemically to those metals which do not com- bine with oxygen directly (under ordinary pressure) either at ordinary or higher temperatures. If, however, the pressure is raised, silver combines with oxygen directly at an elevated tem- perature: 2Ag+OAg 2 0. Nitric acid attacks it readily at ordinary temperatures, sulphuric acid only at higher temperatures; hydrochloric acid has very little effect on it. Uses; alloys. Pure silver is seldom made use of practically, but its alloys are employed in the manufacture of silverware and coins. For these purposes an alloy with copper is used. Silvei 362 INORGANIC CHEMISTRY. [245- plate and jewelry usually contain 75 or more per cent of silver; the silver coins of the United States and continental countries consist of 90% silver and 10% copper; the English shillings ("sterling" silver) contain 92.5% silver. The admixture of copper makes the metal harder. Considerable silver is consumed in silvering objects of copper or other metals (silver-plating). At present this is usually done by electrolysis ( 242). The object to be plated is made the cathode and a silver plate the anode; the bath consists of silver cyanide dissolved in an excess of potassium cyanide. The market price of silver, on a gold basis, was almost constant from 1650 to 1870, the market ratio of the two metals at London remaining very close to 1:15.5, but in the last few decades the value of silver has decreased relatively very greatly. The reason for this shrinkage of value is to be found not only in the chemical improvements in the metallurgy of silver, but much rather in the discovery of large deposits of high-percentage silver ores, especially in the Western Hemisphere. In 1902 the ratio was 1:39.15; since then it has improved somewhat, being in 1911 1:38.74. Compounds of Silver. 246. The known oxides are: Ag40, silver suboxide (very un- stable); Ag20, silver oxide, from which the salts of silver can be derived; and AgO, silver peroxide (formed from silver and ozone). Silver oxide, Ag20 ; is deposited as a dark brown amorphous precipitate when the solution of a silver salt is treated with caustic soda or baryta-water free from carbonic acid. It is somewhat sol- uble in water (2.16X10~ 4 mole per liter at 25) ; the solution prob- ably contains the silver hydroxide, for it reacts alkaline and must therefore contain hydroxyl ions. In its saturated aqueous solu- tion 70% of the molecules are found to be ionized. Silver hydrox- ide is thus not so strong a base as the alkalies, but considerably stronger than ammonia. Moist silver oxide (AgOH) absorbs car- bon dioxide from the air and the silver salts react neutral, while the salts of most of the other heavy metals give an acid reaction be- cause of a slight hydrolytic dissociation in aqueous solution. By heating to 250 silver oxide is broken up into its elements., It is reduced by hydrogen at as low a temperature at 100. Ammonia 246-] COMPOUNDS OF SILVER. 363 water dissolves it readily because of the formation of a complex ion, Ag(NH 3 ) 2 * . Silver chloride, AgCl, is obtained by precipitating a silver solu- tion with hydrochloric acid or a soluble chloride like sodium chlo- ride; it forms a characteristic " curdy" precipitate. It is almost insoluble in water, 1- part in 715,800 H 2 at 13.8. When a silver solution is added very carefully to a sodium chloride solution (or to another chloride), a point can be found when the liquid gives a cloudiness (due to AgCl) with either solution. This must be attributed to the fact that the liquid is saturated with silver chloride and contains no other silver salt nor any other chloride. In view of the very high dilution of such a silver chloride solution (see above) it may be assumed that the dissolved part is completely ionized. If silver or chlorine ions are now introduced into the liquid, the ionization of the silver chloride is diminished and AgCl molecules are formed, but these cannot remain in solution, since the solution is already saturated with them. Silver chloride dissolves readily in ammonia, potassium cyanide and sodium thiosulphate, forming complex ions. If a solution of silver chloride and ammonia is allowed to evap- orate in the dark at room temperature, silver chloride crystallizes out in finely developed octahedrons. Silver bromide, AgBr, is less soluble than the chloride and has a yellowish color. It dissolves with difficulty in ammonia but easily in thiosulphate. Silver iodide, Agl, is even less soluble than silver bromide at ordinary temperatures. It is insoluble in ammo- nia. It is yellow. At high temperatures these halides melt and on cooling form a horny mass, which can be cut with the knife (" horn-silver, " cf. 245). Silver fluoride, AgF, is much more solu- ble in water than the three preceding halogen compounds. Potassium silver cyanide, KAg(CN)2, obtained on adding potas- sium cyanide to a silver solution, dissolves readily in water and is used in large quantities in electro-plating. When a current passes through it, potassium is deposited (primarily) at the cathode, while the anion Ag(CNV wanders to the anode; however, potassium pre- cipitates silver from potassium silver cyanide: K+KAg(CN) 2 =2KCN+Ag. 364 INORGANIC CHEMISTRY. [ 246- Thus silver is deposited on the cathode while the anion Ag(CN}/ takes up an atom of silver at the silver anode to form silver cyanide and again unites with potassium cyanide to form the double salt; if the anode is of platinum, cyanogen gas is set free from the anion Ag(CN) 2 ', and the anode becomes covered with silver cyanide, which soon interrupts the current. All the silver salts, particularly the chloride, bromide and iodide, are sensitive to light, i.e. they are decomposed by light, espe- cially by the violet and ultra-violet rays of the spectrum; as a result, the halogen passes off and the color of the salt becomes first violet and then black. A blackened preparation of this sort can be re- whitened by chlorine-or bromine-water. The sensitiveness to light depends in large measure on the manner in which the silver halide is precipitated. 247. Photography. The property of silver chloride and silver bromide just mentioned forms the basis of photography. The process is essentially as follows: A glass plate is coated with a ""sensitive film," i.e. a thin layer of silver chloride or bromide is spread over it. Formerly this was usually prepared by the pho- tographers themselves from collodion (see ORG. CHEM., 231) which contained a halogen salt, e.g. CdI 2 , in solution. After the evaporation of the solvent a halide coating remained, and by dipping the plates so prepared into a solution of silver nitrate, the silver halide was formed on them. These were the "wet plates"; now they are almost entirely superseded by the "dry plates." The latter are prepared commercially on a large scale. They consist of a film of silver bromide in gelatine (less frequently in collodion) on a glass plate. A sensitive plate of this sort is placed in the photographic apparatus, which is essentially a camera obscura, and the plate is there "exposed" to a light-image, which affects the silver halide chemically. It is very probable that by the action of the light a subhalide is formed; the liberated bromine enters into combination with the gelatine or -the collodion and is therefore unable to trans- form the subhg^Jfc into halide. As yet no picture can be seen on the plate; it must first be "developed." For the latter purpose the plate is immersed in a liquid containing a reducing substance. A typical developer is a solution of ferrous oxalate in an excess of potassium oxalate; various other organic compounds (amido- 247.] PHOTOGRAPHY. 365 phenols, etc.) are at present frequently used. At those places on the plate where the light has struck, more or less silver (according to the intensity of the action of the light) is set free in the metallic form as a very thin coating, while the remaining silver halide is not affected by the developer. This halide must next be removed, else it would be decomposed by the light and more silver liberated; therefore it is immersed in a solution of sodium thiosulphate ("hypo")- This operation is called "fixing" the image. Up to this time the plate must be kept from the light. After the fixing we have a so-called negative, i. e. there remains on the glass plate a picture which is black in those places which were illuminated in the object and clear on those places which were dark. From this a positive impression is prepared by laying the negative on a paper coated with a sensitive film and exposing the whole to direct sunlight. Those places on the nega- tive where silver was deposited let no light or very little through (according to their thickness), so that a positive image is now pro- duced. Finally the positive image is also fixed, for which purpose a bath containing thiosulphate and a little gold chloride is used. The latter improves the color- 1 o n e of the photograph. The photographic process in its various stages is very interesting also from a theoretical standpoint and deserves a little more detailed study. 1. Preparation of the Plates. A mixture is made of solutions of silver nitrate and ammonium bromide containing enough gelatine to make them congeal at room temperature. No separation of silver bromide is observed immediately on mixing, as is the case when the corresponding aqueous solutions are mixed. It may be assumed that the gelatine acts as a protective colloid toward the silver bromide, which of itself is unable to form a hydrosol (cf. 196). That silver bromide is really formed can be demonstrated by measuring the electrical conductivity. If, instead of the silver bromide, its ions, Ag' and Br', were present, the conductance would have to be much greater than that corresponding to the ammonium nitrate which results from the mixing (AgNO 3 + NH 4 Br = AgBr + NH 4 NO 3 ). The observed con- ductance is, however, very nearly equal to that of a gelatinous solution of ammonium nitrate having the same concentration. This freshly pre- pared cofloidal silver bromide in gelatine is relatively not very sensitive to light. In order to increase its sensitiveness the mixture is allowed to "ripen" by standing in the warm for a considerable length of time. It then loses its transparency and becomes yellowish white. The resulting increase in sensitiveness must be accounted for by supposing that the light is not sufficiently absorbed by the transparent colloidal silver bromide to exert its full action, and that this is only accomplished when in the process of ripening the colloid is slowly coagulated, the finer particles of silver bromide 366 INORGANIC CHEMISTRY. [247- having collected to form larger ones, which render the mass opaque and therefore increase its absorptive power. The ripened silver bromide gela- tine is then spread upon the plates. 2. The Latent Image. When the plates are exposed to light there is formed on all places that the light has affected, a " photohaloid," i.e., a mixture of silver bromide with some ultra-microscopic particles of silver. When AgBr is exposed to the light ordinarily (apart from gelatine), free bromine is formed during the exposure ; if a closed apparatus is used and it is afterward placed in the dark, silver bromide is formed again. More- over, not all the silver bromide is decomposed, but an equilibrium is established : AgBr?=Ag+Br, which is displaced farther to the right, the stronger the illumination. Light thus plays the same role in this dissociation as heat in other dis- sociations. If a gelatine plate is used the latent image remains for months unaltered because the free bromine is taken up by the gelatine. 3. Developing. This process is explained by some as follows: By the reducing action of the developer silver is immediately set free from silver subbromide but not from silver bromide, notwithstanding that the latter is capable of being reduced. The system silver bromide plus developer can be compared to a supersaturated solution, which only deposits solid salt when it comes in contact with a crystalline nucleus of salt (cf. 237). The nuclei of metallic silver are furnished by the silver separated out of the subbromide. The deposition of additional silver molecules takes place only upon those molecules already there and not on spots where there was no subbromide originally, i.e. silver is deposited only where the light acted on the plate. According to this nucleus theory the developing process would be comparable to the following experiment: If a few letters are written on a glass plate with a piece of alum and the plate is laid in a supersaturated solution of this salt, the letters become visible, because alum is deposited on them. Silver sulphate, Ag 2 S0 4 , is obtained by dissolving silver in hot concentrated sulphuric acid. It is scarcely soluble in cold water. Silver nitrate, AgN0 3 , prepared by dissolving silver in nitric acid, crystaUizes isomorphous with saltpetre in beautiful rhombic crystals. It is very soluble in water (1 part in 0.5 part at room temperature) and melts at 218. In medicine it is frequently employed, especially as a caustic; it goes under the name of "lunar caustic." Indelible inks are also prepared from it. 248.] GOLD. 367 Silver nitrite, AgN(>2, is formed as a yellowish precipitate on mixing an aqueous alkali nitrite solution and silver nitrite ; it dis- solves in boiling water and crystallizes on cooling in beautiful needles. GOLD. 248. This metal generally occurs native, being found in beds of quartz and alluvial deposits resulting from the decay of quartz rocks. Traces of gold have been detected in sea-water. It occurs in Hungary, Transylvania, the Ural and particularly in Australia, in Transvaal and in the western part of the United States and Canada. Recently large quantities have been discovered in Alaska (Klondike region) . In Colorado considerable gold is obtained from tellurides (sylvanite, etc.). Inasmuch as the amount of gold contained in a cubic meter of ore or rock in the most profitable instances is only very small, it becomes the task of metallurgy to extract it from proportionately large quantities of rock. In the Transvaal this is accomplished as follows: The gold occurs there in so-called reefs, which are vertical veins in the quartz. These reefs are seldom more than one meter thick, but extend for miles east and west; their depth is unknown. They are mined by blasting with dyna- mite; the large pieces are reduced to about the size of an egg in a heavy iron apparatus and then sent to the stamps, that move in a large trough through which plenty of water is kept running. The water carries off the fine auriferous slime, which is made to flow over amalgamated copper plates that are somewhat inclined. The gold is retained by the mercury. After some time the plates are scraped off and the mercury removed by distillation, leaving the gold. The extracted slime ("tailings ") is treated again for gold, for which purpose the cyanide process of SIEMENS is employed. By this process the tailings are allowed to stand for from one day to three weeks in contact with a 0.1 to 0.01% potassium cyanide solution. Under the influence of the oxygen of the air the gold dissolves in it, forming a double cyanide, KAu(CN) 2 : 2Au +4KCN+ 2H 2 O+ O 2 = 2KAu(CN) 2 + 2KOH + H 2 2 . Hydrogen peroxide is also formed and serves to bring further amounts of gold into solution : 2Au+4KCN+HA = 2KAu(CN) 2 +2KOH. 368 INORGANIC CHEMISTRY. [ 248- From this solution the gold is obtained by electrolysis between steel anodes and lead cathodes. At the anode Prussian blue ( 308) is formed, which is treated for potassium cyanide; the gold is deposited at the cathode ( 246). This gold is separated from the lead it contains by cupellation. [Particularly in the United States two processes (chlorination and cyanide) are in general use for extracting gold from its ores without amalgamation. Both processes are especially applicable to low-grade and sulphurous ores, e.g. the tellurides of Colorado. In the chlorination process the ore is crushed and roasted and then treated in revolving barrels with chlorine, prepared either chemically or electrolytically, after which the gold is precipitated with hydrogen sulphide and roasted. The cyanide process is much similar to that described above for treating the tailings, but zinc generally serves as the precipitant instead of elec- trolysis. Placer and hydraulic mining find application in newly discovered deposits but are much less common than vein mining. For the present status of the metallurgy of gold as well as other metals the student should consult a mining annual. TR.] 249. Physical Properties. When pure, gold is reddish yellow, very soft, (much like lead) and extremely malleable and ductile. The thinnest gold-leaf appears green in transmitted light. Sp. g. = 19.265 at 13. It is a very good conductor of heat and elec- tricity. At 1063 it melts to a greenish liquid. Chemical Properties. Gold is the typical representative of the precious metals; it is not attacked by acids and is dissolved only by chlorine-water, aqua regia and potassium cyanide solution (see above). Its compounds are all very unstable; on warming they decompose, leaving the metal. Uses. About one-half the world's production of gold is used for industrial purposes. For these purposes the pure metal is too soft, however, and must be alloyed with copper or silver. The proportion of gold in the alloy is ordinarily expressed in carats; the pure metal is 24 carats; gold jewelry, etc., usually 14-18 carats i.e., 24 parts of the alloy contain 14-18 parts of gold. v The gold coins of the United States contain 1 part copper to 9 parts gold, those of England 1 part copper to 11 parts gold. For purposes of gold-plating the same electrolytic processes are employed as for silver-plating. 249.1 TESTING OF GOLD AND SILVER. 369 Testing of Gold and Silver. The oldest method of testing is by means of the touchstone, or ' ' Lydian stone/' a black basalt. This stone must be dull black, unaffected by aqua regia and somewhat rough. The sample is rubbed on the surface of the stone so as to leave a bright streak of particles of the metal. This streak is then compared with that of a series of touchneedles of known composition. Silver streaks are compared merely as to color. A skilled observer can usually estimate the proportion of silver to within 2.0-1.5%. In the case of gold objects it must be known whether the metal con- tains copper, silver, or both. Therefore the color of the streak is com- pared with that of touchneedles of the presumably corresponding alloy. The streaks are then moistened with a little acid consisting of 1 part HC1, SO HX0 3 and 100 H 2 0. Alloys with 75% or more gold are not attacked by this mixture at ordinary temperatures. If the percentage is less, it is possible to detect differences of 1%. This method is decidedly crude and is usually employed only in confirming a supposed percentage. Where the metal is rich it is very deceptive; but gold of a quality such as is generally used for ornaments, etc. (ca, I 5 o 8 o 3 s by no means abso- lutely established. (See 266-7) . Although we now ascribe to every metal fixed, unalterable properties, it might well have seemed possible to the alchemists, with their more limited knowledge, that the properties of the metal? could vary. None of the metals except gold occur pure in nature; they have to be extracted from oxides or sulphides, which frequently contain various impurities. The metals thus obtained had no definite properties; distinction was made between various sorts of lead, copper, etc. The mutability of the metals may be said to have been the first principle which observation taught; indeed, when a piece of metal is fused with small amounts of various other substances, its properties (color, etc.) really do change. Moreover, at the time of the alchemists the present concept "element " was not yet established; this was first introduced by BOYJ-E (1627-1691). Before then, the doctrine of ARISTOTLE was very generally accepted, according to which all substances are made up of air, fire, earth and water. In order to produce gold it therefore seemed only necessary to deprive the baser metals of certain properties and substitute others. As to the metals themselves the idea was prevalent in alchemistic circles that mercury was the primordial substance and that it had undergone various changes. Before gold could be made from it it must be made refractory and of a yellow color. Not a few alchemists were convinced, moreover, that the success of the "great work " depended on the ?ooper- ation of a higher power. SUMMARY OF THE GROUP. 252. The* metals copper, silver and gold form a bridge from the difficultly fusible metals, Ni, Pd, Pt (Group VIII), tc the easily fusible, Zn, Cd, Hg (Group II); their melting-points are between those of the two groups. The following brief table sum- marizes the physical constants of these metals as well as those* of the related elements, lithium and sodium: 372 INORGANIC CHEMISTRY, [ 252- Li Na Cu Ag Au Atomic weight 7.00 23.00 63.57 107.88 197.2 Specific gravity 59 0.97 8.94 10 5 19.33 Melting-point 180 97.6 1083 961 1063 Color white white red white red The analogy in the chemical properties is chiefly apparent in the -ous compounds. These have the type R 2 for the oxygen compounds and RX for the halides. The -ous halides of Cu, Ag and Au are all white and insoluble in water; they are isomorphous with sodium chloride. Moreover, there are certain analogies in solubility. Lithium carbonate and hydroxide are less soluble in water than the corre- sponding sodium compounds; copper carbonate and hydroxide are insoluble, while the corresponding silver compounds dissolve to some extent. The sulphate of sodium (third horizontal series) crystallizes preferably with 10H 2 0, that of copper (fifth series) with 5H 2 O, while silver sulphate (seventh series) is anhydrous. The oxygen compounds exhibit a gradual decrease in stability. Li 2 O and Na 2 O are unaffected by high temperatures, but CuO is transformed into Cu 2 0, and the oxides of silver and gold break up even at comparatively low temperatures into their elements. However, it must be admitted that the analogy between these elements is not so great as in other groups. Their difference in valence is especially striking and, moreover, there is little simi- larity in the properties of the higher stages of oxidation. This is one of the weak parts of the periodic system. BERYLLIUM AND MAGNESIUM. I. Beryllium (Glucinum). 253. This is one of the rarer elements. It occurs in the mineral beryl, Al 2 O 3 -3Si0 2 + 3(BeO-SiO 2 ); that variety of beryl which is colored green by traces of a chromium compound is the gem called emerald, or smaragd. Chrysoberyl has the composition BeO-Al 2 3 . Almost all the beryllium compounds are made from beryl. This is disintegrated by fusing with potassium carbonate. The fused mass, after cooling, is treated with sulphuric acid to precipitate the silica. Most of the aluminium is then removed by crystallization in the form of alum, 253.] BERYLLIUM AND MAGNESIUM. 373 as this is sparingly soluble in cold water, while beryllium sulphate remains in the mother liquor. The latter is then mixed with a hot solution of ammonium carbonate to precipitate aluminium and iron, beryllium still remaining in solution. After acidifying with hydrochloric acid, the beryllium is precipitated as the hydroxide by ammonia The metal was obtained by heating the double fluoride BeF 2 -2KF with sodium. It is a malleable solid with the specific gravity 1.64. It does not decompose water, even at 100. At ordinary temperatures it is permanent in the air. Hydrochloric and sulphuric acids dissolve it readily with the evolution of hydrogen; dilute nitric acid does not attack it so readily. Beryllium is also dissolved easily by caustic potash and soda with the evolution of hydrogen and the formation of salts having the formula Be(OR) 2 . The hydroxide thus behaves as a weak acid towards strong bases. These properties correspond to those of aluminium; in 215 attention was already called to the analogy between these two elements. This analogy also characterizes their compounds, e.g., beryllium carbide yields pure methane with water, just like alu- minium carbide (178). Only one oxide of beryllium is known, BeO ( 215). It is a white powder, which after ignition is difficultly soluble in acids (like A1 2 3 ). It is obtained by heating the hydroxide, Be(OH) 2 , which is precipi- tated from solutions of the salts as a white gelatinous mass. When freshly precipitated, it is easily soluble in alkalies, ammonium carbonate and dilute acids. On being heated with water, dilute ammonia solution or dilute alkali solution, or on being ignited, or even on standing for some time, it "grows old" and loses these properties. Heating with ten-fold normal solution of an alkali hydroxide " rejuvenates" even the " oldest" beryllium hydroxides, which are dissolved only slowly by warm concen- trated hydrochloric acid. Beryllium hydroxide is distinguished from aluminium hydroxide in two respects: it dissolves in ammonium carbonate (see above) and is precipitated from the solution in caustic soda or caustic potash by prolonged boiling. Beryllium sulphate, BeSO 4 crystallizes with four or seven molecules of water, in the latter case being isomorphous with MgSO 4 -7H 2 O. The double salt BeSO 4 -K 2 SO 4 -3H 2 O is (like alum) sparingly soluble in cold water. Beryllium chloride, BeCl 2 , must be prepared from the oxide by heating with charcoal in a current of chlorine. Its vapor density corresponds to the formula BeCl 2 . It crystallizes with 4H 2 O. Beryllium carbonate is soluble in water. It loses carbon dioxide very easily. The beryllium salts taste sweet, hence the name g 1 u c i n u m (or glycin- ium), which is common in France and America. 374 INORGANIC CHEMISTRY. [ 254- II. Magnesium. 254. This element occurs as carbonate, silicate, and chlo- ride in considerable quantities. Magnesite is MgC0 3 , dolomite MgCa(C0 3 ) 2 . Among the silicates containing magnesium we have talc and soapstone, H 2 Mg3Si 4 Oi2; serpentine (asbestos), H 4 Mg 3 Si 2 09; meerscliaum, H 4 Mg 2 Si30i . It is found in smaller amounts in many other silicates, e.g. hornblende (asbestos) , augite, tourmaline. Other salts found in nature are carnallite, MgCl 2 KC1 6H 2 0, kie- serite, MgS0 4 -H 2 O, and kainite, MgS0 4 -KCl-3H 2 O (Stassfurt Abraum salts). Upon the weathering of the silicates the mag- nesium goes into the soil, whence it is absorbed by the plants (to which this element is invaluable) and finally taken into the animal body. The metal is manufactured on a large scale, since it is employed for illumination in photography, pyrotechnics, etc., on account of the intense light (flash-light) produced by its combustion. At present it is prepared mainly by the electrolysis of fused mag- nesium chloride or carnallite in a cast-steel crucible, which serves as cathode; gas carbon is used for the anode. It is also obtained by heating the double chloride MgCl 2 -NaCl with sodium. It is silvery-white and has a high lustre. Sp. g. =1.75. It is malleable and ductile and comes on the market in the form of wire or ribbon as well as powder, but the ribbon frequently contains zinc. It melts at 651 and boils at 1120. It is quite permanent in the air, since it soons becomes coated with a thin cohesive film of the oxide; at an elevated temperature it burns to magnesia, MgO. When it is heated red-hot in a limited supply of air, a large part is converted into the nitrite, Mg 3 N 2 , a yellowish-green substance. Boiling water decomposes it slowly with the evolution of hydrogen. It dissolves readily in acids but is unaffected by alkalies. It is a powerful reducing-agent, reducing silica ( 190), for example; moreover, when ignited, it burns in water vapor. Magnesium oxide, MgO, magnesia, is the only oxide of mag- nesium known. It results from the combustion of the metal or from heating the hydroxide or carbonate. It is a white, very light powder, which is employed in medicine under the name mag- nesia usta. With water it forms the hydroxide Mg(OH) 2 . Magnesium hydroxide, Mg(OH) 2 , is precipitated from solutions of magnesium salts by alkalies. It is slightly soluble in water and 255.] MAGNESIUM SALTS. 375 turns red litmus blue; however in an excess of alkali its ionization is so diminished that it becomes practically insoluble. It is a weak base, but is strong enough to absorb carbon dioxide from the air. It dissolves readily in an aqueous solution containing ammonium salts. According to OSTWALD, this is to be explained as follows: The solution of an ammonium salt contains a large quantity of NH 4 -ions. When a substance is introduced into the solution, which gives off OH-ions ; as does magnesium hydroxide, these NH 4 -ions unite with OH-ions to form NH 4 OH, or rather NH 3 +H 2 (cf. 234). As a result of this reaction OH-ions dis- appear. In order to restore the equilibrium between the undis- solvecl magnesium hydroxide and the solution, more of this hy- droxide must go in solution, but again the freshly formed OH-ions are taken up by the NH 4 -ions. If sufficient of the latter are present, this process will go on till all the magnesium hydroxide has entered into solution. It now becomes clear why, on the other hand, the solution of a magnesium salt is not precipitated by ammonia in the presence of a sufficient quantity of ammonium salt. MAGNESIUM SALTS. 255. Magnesium chloride, MgCl2, crystallizes with six mole- cules of water and is very hygroscopic. The deliquescence of common salt is due to the magnesium salt it usually contains. OTT On evaporating the aqueous solution the basic chloride, Mg nl , 01 and hydrochloric acid are formed; sea-water cannot be used in boilers because of the magnesium salt it contains, for the hydro- chloric acid set free attacks the iron. Many double salts of mag- nesium chloride are known. It can be obtained anhydrous by heating the double chloride MgCl2 NH 4 C1 6H 2 O, when it forms a laminar-crystalline mass, which melts at 708 and distils without decomposition at bright red heat. Careful study of the decomposition of magnesium chloride by oxygen and by steam has shown that a reversible reaction is involved in each case: 2MgCl 2 + 2 z=2MgO +2C1 2 ; MgCl 2 + H 2 O<=MgO +2HC1. In the former reaction a rise of temperature displaces the equilibrium toward the right, although below 500 the velocity is still very small. 376 INORGANIC CHEMISTRY. [ 255- In the second process the composition of the gaseous equilibrium mix- ture at 700 has been found to be 90% HC1 + 10% H 2 0. Magnesium sulphate, MgS0 4 7H 2 0, Epsom salt, finds use in medicine. It is very soluble in water. It loses 6 mols. H 2 O at 150, and the seventh above 200. In this respect it behaves like other sulphates, e.g. ZnS0 4 -7H 2 0, FeS0 4 -7H 2 0, and those of nickel and cobalt, which are, moreover, isomorphous with it. A further analogy between these sulphates appears in the fact that with sulphate of potassium or ammonium they form double salts of the same type, K 2 S0 4 -MgSO 4 -6H 2 0, which are also isomorphous. Magnesium ammonium phosphates, MgNH 4 P0 4 6H 2 0, serves for the precipitation of magnesium as well as of phosphoric acid. It is not wholly insoluble in water, but does not dissolve in ammonia, the reason for which is again to be found in the reduc- tion of the ionization. Completely analogous to this compound is the corresponding arsenate, MgNH 4 As0 4 6H 2 O. Magnesium carbonate. From solutions of magnesium salts soda precipitates a basic carbonate, Mg(OH) 2 -4MgC03-4H 2 0. The carbon dioxide liberated holds part of the magnesium in solu- tion as acid carbonate. This precipitate is known as magnesia alba. The neutral carbonate can be prepared from it by sus- pending magnesia alba in water, passing in carbon dioxide and allowing to stand; in time the salt MgCO3-3H 2 O crystallizes out, which is, however, readily split up hydrolytically by water, form- ing basic carbonate again. CALCIUM, STRONTIUM AND BARIUM. I. Calcium. 256. This element is one of the ten principal constituents of the earth's crust (8). Particularly the carbonate is found in large quantities in nature, limestone, calcite, aragonite, marble and chalk, all being forms of it. An earthy deposit containing a certain amount of calcium carbonate is termed marl. Calcium silicates and especially calcium double salts constitute the major portion of the siliceous rocks. There are also extensive beds of calcium phosphate, phosphorite, apatite, etc., particularly in Spain and Florida. Calcium occurs as sulphate in the form 'of gypsum and alabaster. Moreover, in the animal kingdom large quantities 257.] OXIDES AND HYDROXIDES OF CALCIUM. 377 of this element are found. The skeletons of vertebrates are chiefly phosphate and carbonate of calcium; the shells of mollusks con- sist of calcium carbonate, as do also eggshells. As for the plants, lime is one of their indispensable inorganic constituents. Metallic calcium can be obtained by electrolysis of a fused mixture of calcium chloride and calcium fluoride. Such a mixture melts much lower than the single salts ( 237). The lower temperature makes the separation of the metal easier and prevents its combustion. Calcium is a silvery-white metal, which melts at 800; it is soft enough to cut and is malleable, but less so than potassium and sodium; it has a crystalline frac- ture. Sp. g.= 1.52. It is relatively little affected by oxy- gen, chlorine, bromine, and iodine, all of which react with the metal only at a higher temperature than the ordinary one. In a current of air calcium unites with both oxygen and nitrogen ( 110). With hydrogen it forms a compound CaH 2 , which is also prepared commercially by passing hydrogen into molten calcium. The calcium hydride reacts with water most vigorously Since 1 kilo of the hydride evolves about 1 cubic meter of hydrogen, it constitutes a very suitable material for generating hydrogen for aeronautic purposes, especially in out-of-the-way places. OXIDES AND HYDROXIDES OF CALCIUM. 257. Calcium oxide, CaO, (quick-lime, unslaked lime) is pre- pared commercially by " burning " limestone or mollusk shells. The limestone is mixed with coal and the latter is set on fire; the heat of the burning coal decomposes the carbonate of lime into calcium oxide and carbon dioxide. The kilns are usually con- structed in such a way that the burned lime can be drawn out at the bottom while the mixture of fuel and limestone is fed in at the top, so that the process is continuous. In the United States " long-flame " periodic kilns are generally used because they are simpler and fuel is inexpensive. Calcium oxide is a white amorphous powder, which requires the temperature of t^e electric r r fl i mace for fusion ( 176). On 373 INORGANIC CHEMISTRY. [257- being heated strongly with an oxy-hydrogen flame it emits an intense white light ( 13). It absorbs water and carbon dioxide from the air; as a result the chunks of lime, which are hard and solid when they come from the kiln, gradually crumble to fine powder. Calcium hydroxide, Ca(OH) 2 , (slaked lime) is obtained by " slaking " quick-lime with water. Its formation is attended by the evolution of much heat. It is only sparingly soluble in water (forming lime-water), but more soluble in cold water than in warm. The solubility, is however, sufficient to make the precipitation of this hydroxide by ammonium hydroxide impossible, for the con- centration of the hydroxyl ions of the latter is too small together with that of the calcium ions present to reach the value of the solubility product of calcium hydroxide. At red-heat it is recon- verted into the oxide. Mortar. Calcium hydroxide is used in masonry. For this purpose quicklime is mixed with water and sand so as to form a thick paste, called mortar, which is thrown in between the stones. After some time the mass becomes as hard as stone; this is due to the conversion of the hydroxide into the carbonate by the action of the carbon dioxide of the air. The sand makes the mass porous, so that the process of hardening extends inward; the older the wall the harder the mortar. The formation of cal- cium silicate appears to play only a minor role in this process. If the lime contains more or less magnesia it is difficult to slake; it is therefore less adapted to masonry purposes and is called " poor," or " lean," in contrast with tr^e pure, easily slaked " fat " lime. Cement contains, besides lime (50-60%), principally silica (ca. 24%) and alumina (ca. 8%). It is made by burning a mixture of limestone, clay and sand. In some places, e.g. Brohlthal in the Rhine region, such a mixture occurs as " tuffstone," which yields cement directly on burning. Cement after being mixed with water sets very firmly in a short time; this is due, in all probability, to the fact that on treating it with water calcium aluminate is dissolved and the solution slowly deposits a hydrous aluminate, which is much less soluble and causes the setting of the cement. At the same time insoluble calcium aluminium silicates are formed. 258.] SALTS OF CALCIUM. 379 Calcium peroxide, CaO2-SH 2 O, is deposited when lime-water is treated with hydrogen peroxide solution. It gives up oxygen on heating. SALTS OF CALCIUM. 258. Calcium chloride, CaCl 2 , is obtained by dissolving the hydroxide or carbonate in hydrochloric acid. It can crystallize with various amounts of water. The hydrate CaCl 2 -6H 2 forms large crystals. Calcium chloride is very hygroscopic and is there- fore frequently used to dry gases or to absorb water dissolved in organic liquids (ether, carbon disulphide, etc.). It melts at 719. It unites with ammonia to form CaCl2-8NH 3 ; hence it cannot be used 'to dry this gas. When crystallized calcium chloride is mixed with ice the temperature falls considerably, even reaching 48.5. Such a mixture is called a cooling- or freezing-mixture and is often employed for producing low temperatures. Besides calcium chloride and ice, many other such mixtures are known; the one most frequently used is that of common salt and ice, with which a temperature of 21 can be obtained. Ice is not absolutely necessary; for instance, if solid ammonium nitrate is added to its own weight of water, a temperature of 15.5 can be produced. In order to understand why such mixtures become so cold we must recall 237. Suppose that ice is introduced into a saturated salt solu- tion of 0, solid s$lt being present at the bottom so that the liquid remains saturated. The system solution + ice is not in. a state of equi- librium at 0, for the salt solution has a freezing-point much lower than 0. It cannot therefore continue in this state, but, if it is to be in equilibrium with ice as solid phase, the temperature must sink, and this is only possible as the ice melts, by which process heat is changed into the latent condition. If enough ice is present, it can, by melting, continue to withdraw free heat from the system till the cryohydric point is reached; for only at or below that point can ice and salt exist perma- nently side by side. It follows, therefore, that the cryohydric tempera- ture is the lowest that can be reached by the mixture. In 237 it was shown, further, that there is no essential difference between the two components of a solution; this is also seen on considering cooling- mixtures containing no ice. For instance, when ammonium nitrate is added to water, the solution has a freezing-point much lower than Here it is the great absorption of heat in dissolving the salt, that causes the fall of temperature necessary to establish the equilibrium. If this 380 INORGANIC CHEMISTRY. [ 258. fall is to be considerable, the solubility of the salt must of course be great. In this case also the cryohydric point is the lowest tempera- ture that can be reached by the mixture. Chloride of lime is a name given to a product obtained by saturating slaked lime with chlorine at ordinary temperatures. Just what compound is formed here is not yet definitely known although the matter has been frequently investigated. There is, OC1 however, much evidence in favor of the formula Ca < , accord- ing to which it is a mixed salt of hydrochloric and hypochloious acids. At any rate this is more probable than the supposition that chloride of lime is a mixture of calcium hypochlorite and calcium chloride ( 56), for it is not possible to extract any chlo- ride of calcium from it with alcohol, although this salt is very solu- ble in alcohol, and almost all the chlorine is expelled by a current of carbon dioxide. Chloride of lime is employed in large quantities for bleaching and disinfecting (bleaching-powder) . It is an incoherent white powder with the odor of chlorine (on account of decomposition by the carbon dioxide of the air.) When treated with hydro- chloric or other acids it yields chlorine: Ca C1 +2HCl =CaCl 2 +H 2 O+C1 2 ; + H 2 S0 4 = CaS0 4 + H 2 + C1 2 . A solution of chloride of lime, when mixed with a cobalt salt and warmed, evolves oxygen. This reaction can be regarded as primarily an oxidation of CoO to Co 2 O 3 , the latter then yielding oxygen with chloride of lime and forming CoO anew. The cobaltous oxide would thus act as a catalyzer. Calcium fluoride, CaF 2 , occurs in nature as fluor spar or fluorite, forming cubes, which are often fluorescent. It is insoluble in water. It fuses at red-heat and is frequently employed as a flux in metallurgical processes. It can be obtained artificially by treat- ing a solution of calcium chloride with sodium fluoride, NaF. Calcium sulphide, CaS, is made by heating gypsum with char- coal. On treating the mass with water calcium hydrosulphide, Ca(SH) 2 , is formed, whose aqueous solution loses hydrogen sul- 258.] SALTS OF CALCIUM. 381 phide on boiling. Calcium sulphide (like the sulphides of barium and strontium) has the property of emitting light in the dark after it has been exposed to sunlight, but seems only to show this phenomenon when it contains traces of other elements, such as vanadium or bismuth. A boiled mixture of lime-water and sulphur is coming into extensive use as an insecticide under the name of " lime-sulphur solution." Calcium sulphate, CaSO 4 2H 2 O, occurs in nature as gypsum ( 256) . It is only slightly soluble in water. We also find calcium sulphate in nature as anhydrite, which has no water of crystal- lization and is very difficultly soluble in water. Gypsum passes over into this anhydrous modification on being ignited. However, the reverse transformation, recombination with water, does not take place, or at least proceeds very slowly, so that ignited gypsum is said to be " dead-burnt." If the dehydration is carried out at a lower temperature, an anhydrous gypsum is obtained which is comparatively easily soluble in water ("soluble anhydrite") and absorbs water very rapidly. In addition to these varieties there is also a "half-hydrate", 2CaS0 4 -H 2 0. This is the chief constit- uent of "plaster of Paris." On being stirred with water it takes up the latter rapidly and, like the soluble anhydrite, forms the dihydrate, CaSC>4-2H 2 0, whereupon the mass becomes hard. This is the basis of the application of gypsum in the manufacture of casts, etc. The "setting" depends upon the relatively high solubility (about 1%) of this half-hydrate, on account of which it forms a solution supersaturated as to gypsum (CaSO/i 2H 2 ; solubility about 0.2%) and gypsum is deposited. Another very essential fac- tor in the setting is the filamentary character of the precipitated gypsum, a property which is entirely lacking in the case of calcium hydroxide, for which reason slaked lime does not hold together. The credit of having explained the conditions governing the exist- ence of the above-mentioned modifications as well as of having deter- mined the positions of their transition points is due to VAN'T HOFF. The investigation was rendered the more difficult because of the retard- ation phenomena which obscure the true situation. It was found that the half-hydrate is to be regarded as a metastable modification, because, for one reason, the temperature at which it goes over into soluble anhy- drite is lower than that at which the dihydrate loses all its water, while in general the loss of water by hydrates proceeds step by step with rising 382 INORGANIC CHEMISTRY. [258- temperature. Moreover the greater solubility of the half -hydrate, as compared with that of the dihydrate, is an additional reason. There is thus the same relationship here as between the metastable crystals Na 2 SO 4 -7H 2 O and the salts Na 2 SO 4 -10H 2 O and Na 2 S0 4 , except that in this latter case the transformation from metastable to stable modifi- cation takes place very easily on touching the heptahydrate with a crystal of the decahydrate, while the half-hydrate of calcium sulphate, even in contact with the dihydrate, retains its identity indefinitely. Calcium nitrate, Ca(N0 3 ) 2 , results from the decay of nitrog- enous organic substances in the presence of lime. It crystallizes with four molecules of water. The anhydrous salt deliquesces in the air and dissolves readily in alcohol. It is converted into salt- petre by potash or potassium chloride ( 229). Calcium phosphates. The tertiary salt, Ca 3 (PO 4 ) 2 , is insoluble in water, as is also the secondary salt, Ca 2 H 2 (P0 4 ) 2 , The primary salt, CaH 4 (P0 4 ) 2 , however, is. readily soluble; it is employed in large quantities as an artificial fertilizer, under the name of " super- phosphate." This superphosphate is manufactured by thoroughly mixing ground phosphorite (or bone meal) in a cast-iron mixer with chamber acid according to the proportions of the equation Ca3(P0 4 ) 2 + 2H 2 SO 4 = CaH 4 (P0 4 ) 2 + 2CaS0 4 . The mass, which is at first semi-solid, soon becomes solid, since the calcium sulphate that is formed takes up the water contained in the chamber acid to form crystals. When superphosphate is mixed with soil the primary calcium sulphate goes into solution and, since every soil contains lime, it is forthwith reconverted into insoluble secondary or tertiary phosphate. Appar- ently nothing has been gained toward "making the phosphoric acid soluble." However the phosphate is now diffused widely in the soil and is therefore much more accessible to the roots of the plants than if the soil had been mixed with tertiary phosphate only. 259. Calcium carbonate, CaC0 3 , is dimorphous, occurring rhombohedral as calcite and rhombic as aragonite. When the solution of a calcium salt is treated with soda, calcium carbonate is at first precipitated in an amorphous, very voluminous and more soluble form; after a short time, however, it turns to a finely crystalline powder. It is very slightly soluble in water, but more 259.] SALTS OF CALCIUM. 383 extensively so in water containing carbonic acid, since the acid calcium carbonate is then formed. The latter decomposes when the solution is boiled, carbon dioxide escaping and crystalline neutral carbonate being deposited. Hardness of Water. Almost every river- or spring-water holds more or less lime in solution. The lime is present as sulphate or as acid carbonate. Such a water forms but little, if any, lather with soap; the fatty acids of the soap form white insoluble salts with the lime, so that water containing much lime is not good for washing. Such a water is termed hard in contrast with a water that is free or nearly free from lime, which is called soft. If the hardness is due to acid carbonate (also called "bicarbonate" of lime), it disappears on protracted boiling, calcium carbonate being precipitated. In such a case we speak of temporary hardness. In metallic boilers and similar vessels the carbonate, of lime that is deposited adheres firmly to the sides (" boiler- scale"). If the hardness of a water is due to gypsum, which is only partially removed by boiling ( 236), it is spoken of as permanent hardness. When heated, calcium carbonate breaks up into lime and carbon dioxide. We have here a case of complete heterogeneous equilibrium ( 71), for the susbtances are CaO and C02 and the phases CaO, CaCOs and C02. This is confirmed by experiments, which show that the concentration of the gaseous phase (the dis- sociation tension) at a definite temperature is constant and there- fore independent of the amount of each phase. Complete decom- position into lime and carbon dioxide can only occur, therefore, when the gaseous phase is removed (as in lime-burning, 257) or when its tension is kept below the dissociation tension. On the other hand, if the tension of the carbon dioxide is greater than the dissociation tension, 'calcium carbonate cannot decom- pose. Under these circumstances it is possible to fuse calcium carbonate; on solidification it assumes a crystalline structure and becomes marble. In the adjoining Fig. pn let AB repre- sent the dissociation curve of calcium car- "bonate in a coordinate system Pt. Only along this curve are the three phases in FIG 60 equilibrium with each other; under any other conditions one of the phases disappears 384 INORGANIC CHEMISTRY. [ 259- and we enter either the region of the phases CaO + C02 or that of CaC0 3 +C0 2 . GLASS. 260. Calcium silicate is chiefly important because it is a con- stituent of almost all sorts of glass. Glass is a mixture of silicates of the alkalies with calcium silicate or lead silicate. The alkali silicates are soluble in water, amorphous and easily fusible. The calcium silicates, however, are insoluble, very hard to fuse and frequently crystallized. By fusing both together an. insoluble amorphous transparent mass of moderate fusibility is obtained, which is glass. It is prepared by fusing a mixture of clean sand, lime and soda in refractory crucibles. The properties of glass depend primarily on the quality of the materials and secondarily on the proportions used. By varying these two conditions it is easy to obtain grades of glass varying widely in fusibility, hardness, lustre, refractive power, etc. There are very many different sorts in use. Some of the most important are the following: Soda glass (window-glass) is a soda-lime silicate. It is readily fusible and is used for most purposes of the household. Potash glass (crown glass, Bohemian glass) consists of a silicate of potassium and calcium. It is very difficult to fuse and is therefore extensively used for chemical purposes (combustion tubes, etc.). Leadglass (flint glass) is a silicate of potassium and lead. It is softer, more easily fusible and highly refractive and takes on a beautiful lustre when polished. It is therefore used for optical instruments and fancy glassware ("cut glass"). Besides the substances mentioned many others are used in glass factories to impart particular properties to the glass. The addi- tion of boric acid or the partial replacement of lead with thallium gives lead glass a still higher refractive index. An admixture of alumina, A1 2 3 , prevents or hinders chemical utensils of glass from becoming brittle and allows the replacement of part of the alkali by lime. Certain metallic oxides form colored silicates and are therefore mixed in with the furnace charge to color the glass (cobalt, blue; chromium or copper, green; uranium, yellow- green fluorescent, etc.). The addition of bone-ash, Ca 3 (PO 4 )2, 260.] GLASS. 385 or tin oxide gives a milky-white opaque glass. The following table shows the percentage composition of various kinds of glass, as determined by analysis: S1O2 K 2 O NaaO CaO PbO AhOs and Fe2Os Window-glass .... 70 15 13 2 Bottle-glass Crown glass 64 74 2 19 4 21 7 9 Flint glass 55 14 31 Plate glass 72 17 6 5 Water has in general very little effect on glass; nevertheless it attacks it somewhat. Old window-panes have a peculiar irides- cence, due to surface weathering'. As it is very important in exact analyses to know how much glass can be dissolved from the utensils, careful investigations have been carried out, the results indicating the following: When the glass is new a rela- tively large amount goes into solution; this amount 'gradually decreases in the course of a few weeks to a minimum. At the first the alkali in particular is dissolved from the surface and the resulting solution then acts as a solvent for the silicic acid. To prepare glass vessels so that they are almost wholly unaffected by water they are subjected to a jet of steam for a quarter of an hour or left for several weeks full of water, the water being renewed occasionally. Thus there is formed on the surface a thin layer, rich in silica and lime, which protects the inner portion from the action of the water. The dissolving action of water on the alkali of glass can be readily shown by agitating finely powdered glass in water. The liquid at once turns phenolphthalein bright red. Glass is a typical amorphous substance. Such substances are often defined as liquids with a very high internal friction and the behavior of molten glass on cooling is an excellent illustration of this definition. At high temperatures molten glass is a thin liquid ; if the temperature is allowed to sink, the consistency .of the glass becomes tougher, so that between the wholly liquid 386 INORGANIC CHEMISTRY. [260- and the wholly " solid " states, there is a continuous series of half-liquid states. As it is thus impossible to find a temperature limit to the applicability to glass of the laws of solutions, e.g. the law of diffusion, it seems rational to consider the " solid " amorphous state as liquid, in contradistinction to the crystalline state, which latter is truly solid, having very different properties from liquids. Solid solution. This term was introduced by VAN'T HOFF to apply to a solid homogeneous mixture. The best example is to be found in mixed crystals, including isomorphous mixtures ( 210). Thus, for instance, when a molten mixture of silver and gold solidifies, the components do not separate, but solidify together in homogeneous crystals of the same composition as the melt. (See Fig. 68, III.) The term " solid solution " is applied to this and somewhat similar solid mixtures, because they exhibit some of the properties of liquid solutions, e.g., in mis- cibility relationships. Glass represents an amorphous type of solid solutions, of which the constituent silicates are the integral components, but, as intimated in the preceding paragraph, many are inclined to regard the amorphous solid solutions as pseudo- solid solutions, i.e., really undercooled liquid solutions. II. Strontium. 261. This is one of the very widely diffused elements. CLARKE showed that in most of the rocks containing calcium this latter metal is accompanied by small quantities of strontium and barium. The principal strontium minerals are strontianite, SrCO 3 , and celestite, SrS0 4 . Its compounds are very analogous to those of calcium. The metal has been obtained by the electrolysis of fused strontium chloride. Its specific gravity is 2.5. In its properties it corresponds to calcium throughout. Strontium oxide, SrO, is formed on igniting the hydroxide or carbonate. The temperature required for the complete dissocia- tion of the latter is higher than that for the corresponding calcium compound. The hydroxide, Sr(OH) 2 -SH 2 O, is more soluble in water than calcium hydroxide. The chloride, SrCl 2 -6H 2 O, is hygroscopic, like that of calcium. It is soluble in alcohol and 262.] BARIUM. 387 can, with the aid of the latter, be easily separated from barium chloride, which is insoluble in alcohol. Strontium sulphate is much less soluble than calcium sulphate; at 16.1 1 part SrS04 dissolves in 10070 parts H 2 O (CaSO 4 , 1 part in 543 at 15.2). In a mixture of alcohol and water it dissolves to an extremely small extent . Strontium nitrate, Sr(NOs)2, is insoluble in alcohol; this forms the basis of separating it from calcium nitrate, which dissolves in alcohol. Strontium salts are used in pyrotechnics because of the beauti- ful crimson color they impart to a flame. III. Barium. 262. This element occurs combined as barite, or heavy spar, BaSC>4, and as witherite, BaCOs, in considerable quantities. In preparing the other barium salts it is merely necessary to dissolve the latter mineral in the proper acid. Barite, however, must first be reduced by ignition with charcoal. This can be accomplished in the electric furnace: (1) 4BaS0 4 +4C =BaS+3BaS0 4 +4CO; (2) 3BuSO 4 +BaS=4BaO+4SO 2 . The metal is, in this case also, obtained by the electrolysis of the fused chloride. Another method is to heat the oxide with magnesium. Barium decomposes water vigorously even at ordinary temperatures. Sp. g. = 3.75. Barium oxide, BaO, is obtained by igniting the nitrate or hydroxide at a high temperature. It unites very readily with water to form the hydroxide, Ba(OH) 2 , which is rather soluble in water (yielding baryta-water), and crystallizes from the hot solu- tion on cooling in pretty lamince, which contain eight molecules of water. Barium peroxide, BaC>2, forms on heating the oxide in a cur- rent of oxygen or air. When it is introduced into dilute sulphuric acid, barium sulphate is precipitated and hydrogen peroxide left in solution. If baryta-water is again added, the hydrate BaC>2 8H 2 O crystallizes out. Barium chloride, BaCl2-2H 2 0, is not hygroscopic like the chlorides of strontium and calcium. The nitrate crystallizes anhydrous. 388 INORGANIC CHEMISTRY. Barium sulphate, BaS0 4 , is characterized by an exceedingly small solubility in water and acids; at 18.4 1 part dissolves in 429,700 parts H 2 O. It is used as a filler and as a pigment under the name of " permanent white/' or blanc fixe. Barium carbonate yields carbon dioxide only at very high temperatures, prolonged heating at 1450 being required for complete decomposition. SUMMARY OF THE GROUP OF THE ALKALINE EARTHS. The following small table summarizes the physical properties of the elements of this group: Be Mg Ca Sr Ba Atomic weight 9.1 24.32 40.09 87.62 137.37 Specific gravity \tomic volume 1.64 5.6 1.75 13.8 1.58 25.2 2.5 34.9 3.75 36.5 Color white white . white white white As to the specific gravity we observe that only in the cases of Ca, Sr and Ba is a steady increase noticeable. In respect to the chemical properties, it has already been remarked that these elements act only as bivalent; all compounds of the group therefore have the same formula type. In the solu- bility of the sulphates a gradual decrease is to be observed with rising atomic weight. Just as in the first group three elements K, Rb, Cs, exhibit a particular kinship, so here calcium, strontium and barium are closely related in their properties, while the two other members of the group are unlike them in many respects. Beryllium displays analogy with aluminium in certain points just as lithium does with magnesium. 263-] SPECTROSCOPY. 389 SPECTROSCOPY. 263. If the light from an ordinary gas flame or the WELSBACH incandescent light is broken up by a prism, there is projected a continuous series of perfectly blended colors from red through yellow, green, and blue to violet. This phenomenon is called a spectrum, and since it is unbroken, a continuous spectrum. We have previously remarked that the luminosity of a gas-flame is due to incandescent solid particles of carbon. It has been found to be a general rule that incandescent solids give a continuous spectrum. With incandescent gases it is different. If, for instance, we split up the light from a Bunsen flame, in which salts of sodium, calcium or other metals are volatilized, we see only a few narrow bands of light in certain places, the rest of the spectrum being dark. This is termed a line spectrum. Every element has its own peculiar spectrum lines. If the spectrum of the incandescent vapors of a mixture of elements is carefully examined, it is found to contain all the characteristic lines of each element. Since it is only necessary to volatilize extremely small amounts of sub- stances in order to show their lines, it is readily seen how important the spectrum-analytical methods introduced by BUNSEN and KIRCHHOFF must be. For the examination of spectra a number of instruments have been constructed, varying according to the particular object in view. For chemical analysis the apparatus of VOGEL or that of JOHN BROWNING is now very generally used. It is a small direct- vision spectroscope which gives a very bright spectrum and has a sufficient dispersion. At the end B (see Fig. 61) is the slit which can be made narrower or wider by turning the rim D. The small mirror m serves to throw light through the hole P on to an auxil- iary prism, in order to compare the spectrum of the light which is to be analyzed with that of a known source. At the left end 390 INORGANIC CHEMISTRY. [263- is the ocular through which the spectrum is seen. For further information text-books on physics should be consulted. In order to examine the spectra of metals it is necessary to convert the latter into the form of vapor at a high temperature. There are different ways of doing this. One is to introduce salts of the metals into a colorless flame by means of a platinum wire. The heat dissociates halogen salts and in the case of oxysalts converts them into oxides, which are reduced to the metallic condition by the hot gases of the flame. This methcd is very satisfactory for some elements, e.g. those of the alkali and alkaline earth groups, when there is plenty of material. In other cases a flame spectrum of this sort is not so good as a spark or an arc spec- trum, for with the latter it is possible to detect with accuracy extremely small amounts of a substance. Other advantages of the latter spectra are their greater light intensity, the greater con- venience in execution, and the like. Moreover, at the high tem- perature here prevailing most elements exhibit spectra which cannot be obtained with the gas-flame. A spark spectrum can be obtained in a very simple manner, thus: Into the bottom of a little glass cup (n, Fig. 62), about 15 mm. wide is fused a platinum wire, which ends in a tube g containing mercury and is thus connected with the negative pole of an induction coil; it is incased in a conical capillary tube x, beyond which the wire projects about 0.5 mm. At the opposite end is the positive electrode in the form of a platinum wire, which, with the excep- \J\J} ri tion. of the short end d, is fused into a glass tube ; the latter is fitted into the cork a. If some of the salt solution is poured into the cup about half way up the negative electrode, the liquid is drawn up to the end of x by capillarity and every spark volatilizes a tiny portion. In this way there is no loss of material and the sparks are very uni- form, so that the observation of the spectrum can be continued at length. For the study of the spectra of substances which are gaseous at ordinary temperatures the P L i) c K E R-H i T T o R p (GEISSLER) tubes are used (Fig. 63). The gases are sealed up in them in a very dilute condition. On connecting one of these FIG. 62. 264.] SPECTROSCOPY. 391 with the poles of an induction coil, the whole tube is illuminated most intensely in the narrow portion. This part is placed ver- tically in front of the slit of the spectroscope. Some substances have the property of absorbing certain colors and transmitting others. If the solution of such a substance is placed before the slit of a spectroscope and the light of a contin- uous spectrum allowed to pass through it, dark bands or lines are observed in the spectrum. A number of substances have very characteristic absorption spectra. 264, The spectroscope is one of the most delicate means we have of detecting many substances. This is readily seen on con- sidering how small an amount of the substance under examina- tion is volatilized by the sparks. We arrive at numbers like FIG. 63 0.3X10" 6 mg. sodium, for instance, as the least amount that can be detected. It has thus been possible to discover elements which occur only in company with large amounts of others and would therefore have been very difficult to find in the ordinary way. BUNSEN and KIRCHHOFF themselves found caBsium and rubidium in this way in Durkheim mineral water, In order to obtain these elements from it in the form of chlorides, it was necessary to evaporate 44,000 kg. water, which yielded 16.5 g. of a mixture of the chlorides. Other very rare elements which were discovered by spectrum analysis are thallium, indium, gallium, ytterbium and scandium. The spectra of the elements differ greatly in appearance, as may be seen at once from Table II (Frontispiece). The numbers indicate wave lengths of light expressed in hundredth microns (10~ 5 mm.). Certain metals, such as sodium, thallium and indium, exhibit only one distinct line when their flame spectra are examined with a spectroscope like the one described above. If a sparking current or an electric arc is employed for the vola- tilization of the substance and the spectroscope is one giving strong dispersion, many more lines become visible. It is further found on photographing spectra that there are still more lines in the infra-red and ultra-violet portions, which are invisible to the eye. Present-day spectroscopic studies deal, therefore, almost exclusively with carefully prepared photographs. 392 INORGANIC CHEMISTRY. [264- The number of spectral lines increases rapidly as we proceed to elements of the higher groups of the periodic system. While lithium, sodium and potassium give 20, 35, and 41 lines, re- spectively, the spectrum of barium contains 163 lines and that of iron more than 5000 lines. Among these lines there are certain ones which, in virtue of their position (color) and intensity, are specially characteristic of an element, like that of the yellow line in the case of sodium, the green line of thallium and the blue lines of indium. For purposes of identification of such elements these prominent lines are generally observed directly in the spectral apparatus. Nitrogen is an example of a substance that gives a band spec- trum when it is examined in the manner described in 263. 265. The position of the spectrum lines was formerly indi- cated numerically according to an arbitrary scale. Now it is expressed with the aid of the wave length/}, 10 ~ 7 being taken as a unit and the unit being called the ANGSTROM unit after the physicist who introduced it. The wave length of the sodium DI line was found to be expressed by 5896.16 such units. The visible part of the spectrum comprises the wave lengths of about 7500-4000 A.U. 'Thanks to the researches of ROWLAND, MICHELSON, KAYSER and RUNGE, and others, the wave lengths of a very large number of spectrum lines have been determined with great exactness, so that one is encouraged to attack the question whether in the apparently very promiscuous distribution of lines in the spectra there is such a thing as order. BALMER was the first to show that this is the case in the 779 hydrogen spectrum. The formula X=A 2 -, where A is a constant (3646.13) expresses the wave lengths X of the lines of the spectrum of the element with very close approximation, provided ra is substituted by consecutive whole numbers begin- ning with 3. The spectra of other elements have been examined for similar regularities, chiefly by RYDBERG and by KAYSER and RUNGE, and it has been found that the regularities are in all of the cases more complex than for the hydrogen spectrum. It would lead us too far to enter upon a discussion of these questions, which 265.) SPECTROSCOPY. 393 properly belong to Physics, but a few of the interesting results are worth mentioning here. RYDBERG, who has devoted particular attention to the spectra of the alkalies, introduced into his formulae the reciprocal of the wave length, the oscillation frequency n, which represents the number of wave lengths per centimeter. In the spectra of the alkalies he found three series of lines, whose oscillation fre- quencies can be expressed by the formula In this formula NQ is a constant having the same value for all these metals and all the series; UQ and ,, however, are two constants that have different values for each series of each metal. For m we substitute again consecutive integers, as in BALMER'S formula. This last formula is, moreover, a special case of that of RYDBERG, since BALMER'S formula can be trans- formed into 1 M/ A m* (where n ' = A and N Q ' = 4A), into which RYDBERG 's formula is also transformed when j = 0. The values of the constants of these different series were found by RYDBERG to have still further definite relationships. The formula of KAYSER and RUNGE is T = A + Bm~ 2 + Cm~ 4 , A in which A, B and C are constants and m consecutive whole num- bers. It represents the wave lerjgth of the lines in many cases more faithfully than does the formula of RYDBERG; however, there is no relation between the constants A, B and C of the different series. The spectral lines of the alkalies also exhibit the peculiarity of consisting of double lines (doublets) or triple lines (triplets), the wave-length differences being constant for each series. Similar series of lines, whose oscillation frequency can be expressed by one of the above formula?, are found in the spectra of some of the elements. In the spectra of many others, how- 394 INORGANIC CHEMISTRY. [ 265- ever, they are lacking. In their place we find in the spectra of lead, tin, arsenic, bismuth and others a constant difference between the oscillation frequencies of a considerable number of their lines. Such investigations as these are prompted by the notion that a knowledge of the laws which govern the distribution of the spectral lines of one and the same substance on the one hand, and the variation in the distribution from substance to sub- stance on the other hand, would throw some light on the nature and kinetic condition of the atoms. With the aid of spectroscopy it has been possible to deter- mine what elements are present in the heavenly bodies. When light from the latter is passed through a prism, line spectra are obtained and these lines correspond in position to those of terres- trial elements. The composition of sunlight has been especially the object of a most extensive study. The spectrum of that body contains numerous black lines, known as FRAUNHOFER lines. The theory of this phenomenon is explained in Optics. By comparing the FRAUNHOFER lines with the spectra of ter- restrial substances it has been found that the sun's atmosphere contains chiefly Fe, Na, Mg, Ca, Cr, Ni, Ba, Cu, Zn and H (the latter in enormous quantity). Moreover, for 450 lines of the iron spectrum there are found to be corresponding dark lines in the sun's spectrum. On the other hand, the solar spectrum displays countless lines which are not yet identified in terrestrial spectra. We are led to presume that many of the elements to which these lines are due will also be revealed on the earth by more careful research, especially when we consider what a small part of the earth is known (see footnote, p. 8). This presumption has been strongly confirmed by the discovery of helium ( 111). The principal line of the latter, D s ~ so termed because of its proximity to the double D-line (D^) of sodium was observed in the spectra of many fixed stars as well as in that of the sun before the element itself was identified on the earth. Helium was thus discovered in the stars before it was found on the earth. It is a striking fact that it occurs in exceedingly large quantities in the fixed stars (according to spectrometric observations) while there is apparently only a very small amount of it on the earth. 266. THE UNITY OF MATTER.] 395 THE UNITY OF MATTER. 266. The notion that all substances are derived from a single original substance and that the variety that we observe in the material world is merely the result of a difference in arrangement and form of the smallest particles has long been prevalent. Even the old Greek philosophers had a fondness for it. However, the rise of experimental investigation was not very conducive to the idea. BOYLE (1626-1691) introduced the concept element in its present form. According to this concept all substances are to be regarded as elements, which, with the means at our command, cannot be further resolved into dissimilar components. The sub- sequent development of chemistry has shown that the number of these elements is rather large. Notwithstanding that the idea of a primordial substance lacked substantiation and was more or less discredited, it was by no means rejected, for we have been expressly reminded again and again that the substances which chemistry regards as the simple substances are only classed as elements conditionally; the pos- sibility always remains that a so-called element may be found capable of division into dissimilar components, as has often ac- tually been the case. Impossible as it was to deny the existence of a primordial substance, the researches of the 18th, and a large part of those of the 19th, century brought to light nothing to support the idea. Not until the discovery of the Periodic System did the question again demand serious attention. This discovery was the first to supply an experimental basis for the assumption of a primordial substance ( 220). The striking dependence of the properties of the elements on the periodic functions of their atomic weights, which finds its expression in this system, leads of itself to the thought of a fundamental substance, of which the simple sub- stances called elements may be said to be polymers, incapable of resolution by the means at our command. Another argument for the divisibility of the elemental atoms is contributed by spectroscopy. In order to explain the line spectra exhibited by many elements, we assume that the move- ments of the atoms give rise to light vibrations of definite wave 396 INORGANIC CHEMISTRY. [ 266- length, which are perceived by us in the ppectral lines. However, since the spectrum of a single element is extremely complicated, we should have to assume that the atoms engender very complex movements. The simplified hypothesis was then offered that it is not the entire atom but smaller particles of which the atoms are composed, that give rise by their vibrations to the different? spectral lines. The physical investigations of the last decade have furnished substantial reasons for believing that the chemical atoms are not in reality the ultimate particles of matter but that they are divi- sible into particles approximately 2000 times smaller than the hydrogen atom. These particles, carrying with them, as they do, very strong electrical charges, are called " electrons." We cannot in this book do more than indicate the main observations arid inferences. The hypothesis of electrons is the result of the study of cathode rays in connection with the below-mentioned investiga- tions of radio-active elements. Cathode rays are generated when the discharges of an induction coil are sent through a rarefied gas. It is assumed that from the cathode there are projected particles with strong* negative charges, electrons, which are propagated with a velocity of several thousand kilometers per second. Ac- cordingly, the cathode rays consist of a stream of these electrons. Measurements of the mass of an electron have shown that it is about iniSnr f that f a hydrogen atom. The electrons are the same, whatever gas is contained in the apparatus and whatever electrodes are used, so that we are evidently dealing with a decom- position of the atoms into their ultimate components. The anions, according to the electron theory, consist of atoms and one or more electrons; the chlorine ion, for example, con- sists of chlorine and an electron, which latter is represented by 6. The cations are formed from the atoms by the release of one or more electrons. We may thus write: C1' and K-0=K>. The ionization of potassium chloride in water can be represented by the equation: KCl aq = (K-0) Certain physicists presume to be able to go a step further. Some very remarkable investigations of ROWLAND have shown 267.] RADIO-ACTIVE ELEMENTS, 397 that a, moving electrically charged conductor exerts the same effect as an electric current, so that the latter may be regarded as^consisting of very swiftly propagated, discrete electrical par- ticles. Electricity would thus have an " atomistic" structure. FARADAY'S law points in the same direction. According to this law the charges on the ions are either equal to or a multiple of the charge on a hydrogen ion. Fractions of the charge are not found. This is a very significant fact. Just as for the explana- tion of the analogous laws of DALTON we have assumed that the elements consists of atoms, so we cannot avoid inferring from FARADAY'S law that electricity is divided into discrete elemental particles, which are to be regarded as "atoms of electricity." Furthermore, it has been shown that induction and other pheno- mena proceed just as if electricity had mass, for electricity has the same properties of inertia as ponderable matter. The conse- quence is that matter is identified with electricity and the elec- trons are no longer to be regarded as electrically charged mass par- ticles but as electrical charges themselves, without a material body. These hypotheses would not only unify matter but would also dispel the time-honored notion that energy and matter are distinct. 9 RADIO-ACTIVE ELEMENTS. 267. BECQUEREL discovered that uraninium emits a peculiar sort of rays which are propagated in a straight line and act o-n a photographic plate, but are not reflected, refracted, or polar- ized. When gases are traversed by them the gases become electrical conductors. Now when uraninite (or pitchblende, a uranium-bearing mineral of very complicated composition) was investigated as to its radiation the strange fact was brought out that the radiation of the mineral is 4.5 times as powerful as that of its constituent metal, uranium, although only 50% of the mineral is uranium. Uraninite must therefore contain one or more substances having a stronger radiating power than uranium. We are indebted principally to the gifted couple, M. and MME. CURIE, for the discovery that the emission of these special rays, or the radio-activity, is due to the presence of very small amounts of elements, hitherto unknown and of very sur- prising properties. 398 INORGANIC CHEMISTRY. [ 267- The only means of control in the separation of these elements from the other compounds in uraninite after the removal of uranium was to measure the radio-activity of the product ob- tained in each operation. This was accomplished by measuring the conductivity of a layer of air that was exposed to the rays. Thus after numerous chemical operations the active substance was concentrated more and more. This method is comparable to that employed by BUNSEN and KIRCHHOFF in isolating rubid- ium and caesium from the Diirkheimer mineral water, where the spectroscope ( 232) indicated the progress of the concentration of these elements. However, the measurement of radio-activity is many thousand times more sensitive than a spectroscopic examination. Were it not for this fact, the discovery of the radio-active elements would have been impossible, because they occur in such extremely small quantities. For example, 2000 kg. uraninite residues from Joachimsthal yield only about 0.2 g. radium chloride. Radium is the best known of these elements. It is the only one that has been isolated and whose compounds have been prepared in the pure state. Its spark spectrum has three very bright lines in the blue and violet and accordingly the Bunsen flame color is carmine. In its chemical behavior it shows close analogy to barium; it is separated from the latter element by fractional crystallization of the bromides, radium bromide being more difficultly soluble than the corresponding barium salt (this is true for all the respective salts of the two elements). With the aid of the spectroscope it can be determined whether the salt is entirely free from barium bromide. The atomic weight of the radium thus purified was found to be 226.4, which could not be raised by further fractional crystal- lization. With this atomic weight radium fits exactly into the second group of the periodic system. All radium salts are lumi- nous and excite a large number of substances, such as barium platinocyanide, BaPt(CN) 4 , uranyl sulphate, precious stones, and the like, to powerful fluorescence. It similarly affects the dia- mond. Genuine diamonds can thus be distinguished from imi- tations. The radio-activity of the pure bromide is about a million times that of uraninite. MME. CURIE and DEBIERNE succeeded in 1910 in isolating the 297.] RADIO-ACTIVE ELEMENTS. 399 element itself. They electrolyzed a solution, using a mercury cathode, and obtained a radium amalgam, from which the mer- cury was distilled off in a current of hydrogen. Radium is a white metal, melting at 700. Even as low as this temperature it volatilizes appreciably. It is attacked by the air and decom- poses water vigorously. Besides uranium and thorium the most important radio- active elements are polonium, actinium, ionium, and radio- thorium. Polonium is precipitated in a number of reactions with bismuth; by hydrogen sulphide, as well as when the basic salts of bismuth are precipitated by water; stannous chloride precipitates it in the same way as mercury and tellurium. It is also deposited on a rod of silver or bismuth when one of these is immersed in a solution containing polonium. The radio- activity of polonium is about a thousand-fold as great as that of radium. From 15 tons of pitchblende MARCKWALD could only obtain 3 mg. polonium salt, still somewhat impure; so that polo- nium also surpasses radium considerably in scarcity. Actinium occurs with the rare earth metals, particularly lanthanum, and can be partially, though unsatisfactorily, separated from them by fractional crystallization of the manganese double nitrate. For ionium see below; for radiothorium see under thorium. The rays emitted by radium preparations are of three sorts and are distinguished as a- ft- and ^-rays. Quantitatively the first are predominant. All of them have the above-mentioned properties in common; they differ, however, in their penetrating power and in their behavior in the magnetic field. The a-rays are not very penetrating and are only slightly deflected in a strong magnetic field. A sheet of aluminium foil 0.1 mm. thick almost entirely stops their passage. Moreover they are com- pletely absorbed by a layer of air a few centimeters in thickness. The /?-rays are strongly deflected in a magnetic field and consist of rays of various but greater penetrating power; some kinds of /?-rays can even pass through an aluminium plate 1 cm. thick. The f-rays are scarcely deflected at all and go through obstruc- tions with ease, several centimeters of lead being insufficient to stop them; they form only a small part of the total radiation. The interesting thing is that these rays are analogous to those generated by electric discharges in highly rarefied gases. The 400 INORGANIC CHEMISTRY. 266- /3-rays are to be regarded as cathode rays of great velocity. They consist of negative electrons which are propagated with very great velocity, some almost with the velocity of light (300,000 km. per sec.). From the deflection which they undergo in an electrical field and a magnetic field of known intensity their mass is calculated to be (as in the case of the cathode rays) about 2 - V <) of that of a hydrogen atom. The velocity of these electrons can also be calculated from the same data. Their enormous velocity explains the great penetrating power of /?-rays. The a-rays resemble a sort of radiation which is also obtained by discharging electricity in a rarefied gas, viz., the canal rays of rrn FIG. 64. EFFECT OF A MAGNETIC FIELD ON THE a-, /?-, AND J--RAYS. GOLDSTEIN. They behave as positively charged projectiles hurled at a great velocity (about -fa that of light). Their mass is about equal to that of a hydrogen atom, or much greater than the mass of the projectiles formed by the /?-rays and the cathode rays. Their greater size and relatively small velocity explains their slight penetrative power. In the light of the more recent investigations they appear to consist of helium atoms bearing two positive charges each, or, more specifically, having lost two electrons. The ?--rays are analogous to the X- or RoENTGEN-rays. These proceed from a metal plate which is placed in the path of cathode rays; they do not consist of a stream of electrically charged particles, but are regarded as a form of wave motion of the ether, which originates when electrons are projected with great velocity against a solid body. The manner of detecting the various sorts of rays follows 267.] RADIO-ACTIVE ELEMENTS. 401 readily from the above description of their properties. Use can be made, for example, of their dissimilar penetrative power. Their separation in a magnetic field is diagrammed in Fig. 64. While the f-rays suffer no deflection, the a-rays are deflected to one side, the #-rays very much to the opposite side. According to what has been recited in the preceding sections, we are to look upon these forms of radiation as evidencing a spon- taneous decomposition of the atoms of the radio-active elements. The decomposition is accompanied by a very considerable evolu- tion of heat. One gram of radium gives off about 118 g.-cal. per hour; for this reason radium salts have a higher temperature than their surroundings. Even cooling with liquid hydrogen (253) does not stop this evolution of heat. The magnitude of the heat effect is more apparent upon comparison with other caloric effects attending chemical reactions. We now assume that the heat evolution in the decomposition of 1 g. radium is about 10 9 g.-cal. On the other hand, the formation of 1 g. water from its elements evolves 4X10 3 cal., so that the first-mentioned process gives off 250,000 times more heat than the second. The spontaneous decomposition of radio-active substances is accompanied by other phenomena. Every substance that is brought into proximity with a radium salt acquires a temporary, or induced, radio-activity, i.e., it emits the same rays as radium itself. This induced radio-activity is best observed on putting a radium salt in an enclosed space. The enclosing walls, as well as all bodies within the space, become active. It is not the radium rays that cause this effect, for a radium salt in a sealed tube emits rays without exciting any radio-activity. RUTHERFORD discovered the cause of this phenomenon by the observation that there is a constant outflow from radio-active substances, which outflow he calls emanation. Since bodies with induced radio-activity give out rays that are identical with those of radium itself, these rays can be regarded as transformation products of the emanation of radium. Emanation behaves in many respects as a gas; it diffuses from one vessel into another, follows the law of BOYLE in its compres- sion, can be condensed by cooling with liquid air and volatilized .again if the temperature is allowed to rise. Neither physical nor 402 INORGANIC CHEMISTRY. [ 267. chemical agencies are able to alter emanation. It is indifferent to temperature variation between -180 and 500, is not absorbed by concentrated acids or alkalies, and can be conducted without change over hot copper oxide. It has the properties of a gas of the argon group. Ramsay has lately succeeded in preparing radium emanation in somewhat larger quantities. He calls this element niton. It is found to be a water-clear liquid with a specific gravity of about 5 and a boiling-point of 62 under atmospheric pressure. In glass vessels it is very highly fluores- cent. Emanation and induced radio-activity must be considered as intermediate stages in the complete disintegration of the radium atom into the above-mentioned radiations. How- ever, other substances, part of which have not been further studied, are formed simultaneously. One of them is pretty well known, viz., helium. RAMSAY and SODDY have demon- strated that helium is formed in the spontaneous decom- position of radium emanation. The maximum quantity of emanation that could be obtained from 50 mg. radium bromide was conducted by them with the help of an oxygen current into a U-tube cooled by liquid air and the U-tube was then evacuated with a pump. A vacuum tube which was fused on to the U-tube showed no traces of helium after removal of the liquid air. The spectrum appeared to be that of an unknown element presumably emanation. After the apparatus stood four days the helium spectrum appeared. This phenomenon explains the mysterious persistent occurrence of helium in radium-bearing minerals. The law governing the rise and decay of radio-active sub- stances is the same as that for a unimolecular reaction. As we have seen in 50, the velocity S of such a reaction can be repre- sented by the equation: if we let C stand for the concentration and K for a constant. With the aid pf higher mathematics this equation can be trans- formed into: C=C n e~ Kt . 267.] RADIO-ACTIVE ELEMENTS. 403 where CQ is the initial concentration and e the base of natural logarithms. The logarithmic form of the equation is : l~=-Kt. This same equation holds, as above stated, for the velocity of decomposition of a radio-active substance; in that case, however, we understand by C the intensity of radiation. This magnitude can be determined electrometrically. If a radio-active substance changed into only one new sub- stance, the phenomenon would be very easy to represent graph- ically; for upon plotting the time on the abscissa axis and the logarithm of the activity on the ordinate axis the phenomenon would be represented by a straight line. But when the substance A is converted into another active substance B, and this again into a new active substance and so on, the situation becomes much more complicated. A graphic representation with the same co-ordinates as before would no longer yield a straight line, but a rather complicated curve. Nevertheless, it has been found possible to resolve these experimental curves and to calculate with certainty the number of active substances which participate in the transformations, as well as their constant K. This is not the only method of ascertaining the number and kind of the intermediate products. We can often distinguish the individual substances involved, by a study of the kind of radiation given off, certain of the substances emitting only ct-rays, others only /?-rays, and still others a mixture of all three rays; indeed there are some of the substances which emit no rays at all. In some instances these active substances have been actually separated by physical or chemical means. Certain of the sub- stances are found to be gaseous; others form a deposit on solid bodies. The gaseous substances can be condensed by cooling. The best way to characterize the various radio-active sub- stances is by the exponent K of the above equation; this constant is to a large degree independent of temperature and pressure; which is not true of ordinary reactions. Frequently, however, 404 INORGANIC CHEMISTRY. [267. use is made of another magnitude, related to K, viz., the period of half decay. If in the equation i., ~K we take 4, usually crystallizes out of its aqueous solution as 3CdS04-8H 2 0. There is also a salt CdS04-7H 2 O, which is analogous in composition to the sulphates of magnesium, zinc, iron, etc. MERCURY (Quicksilver). 271. Mercury is the only metal that is liquid under ordinary conditions. It occurs in nature in cinnabar, HgS, and also native. The chief localities are Almaden in Spain, Idria in Illyria, Mexico, Peru, California, China and Japan. To obtain mercury from cinnabar the latter is roasted in furnaces, sulphur dioxide and mercury being formed. The mercury vapor is condensed either > large chambers or in peculiarly shaped earthen retorts, or pipes, called aludels. It is brought on the market in 75-lb. iron flasks. 271.] MERCURY. 411 The commercial product is not pure, containing more or less of other metals in solution (e.g., lead, copper, etc.). Such impurities can be readily detected by the fact that they make the mercury adhere to a glass vessel. A suitable process of purification consists in letting it fall in fine drops through a long column of nitric acid (sp. g., 1.1), as in Fig. 67. The foreign metals are thus completely dissolved, while almost no mercury is FIG. 67. PURIFICATION OF MERCURY. lost by solution, because these foreign metals precipitate mercury from its salt solutions. After being washed with water the metal is dried and, if absolute purity is desired, it is then distilled in vacuo. But a vacuum distillation of itself is insufficient, for some lead goes over with it. Physical Properties. Mercury solidifies at 39.4 and boils at 360. Even at ordinary temperatures it is somewhat volatile, especially under reduced pressure; when gold leaf is suspended in a bottle over mercury, for instance, it eventually becomes white. 412 INORGANIC CHEM IS TR Y. 271- The metal has a specific gravity of 13.595 at 0. The vapor density is 99.36 for H = 1 ; hence the molecule weighs 198.72. This number also represents the atomic weight, as has been found from molecular weight determinations of many volatile mercury compounds. Amalgams. Many metals have the property of dissolving in mercury or forming compounds with it. These metal solutions or compounds are called " amalgams." Besides by the direct contact of the two metals they can sometimes also be obtained by allowing mercury to act on the solutions of metal salts, e.g. silver amalgam can thus be prepared. Some metals, such as tin, dissolve in mercury with heat absorption; others like potassium and sodium with great heat evolution and vigorous action. If a great excess of mercury is used, the amalgams are liquid, other- wise solid. Sodium amalgam is exceedingly firm when it contains more than three per cent, of sodium. Opinions have been divided as to whether amalgams, and metallic alloys in general, are mixtures or compounds. In order to solve the question a study has been made of the freezing-point curves (cf . 237) of different pairs FIG. 68. TYPICAL FREEZING-POINT CURVES OF PAIRS OF METALS. of metals. Additional information, often of a decisive nature, has been afforded by the microscopical examination of the surface of the alloy after having been etched with dilute acid and polished, the individual crystals being generally distinguishable in this way. As a result of these investiga- tions it has been found that the freezing-point curves for pairs of metals (binary alloys) may be of three different types (Fig. 68). In some cases there is no association of the metals and the curve (I) takes the same form as the frtezing-point curve of a salt solution, having a eutectic point. In others a 272] MERCUROUS COMPOUNDS. 413 compound may be formed (II), in which two eutectic points are observed, one for each of the components with the compound. The compound has the composition corresponding to the abscissa for the maximum of the curve between the eutectic points, the freezing-point of the pure compound being lowered by the addition of one of the components, just as is the freezing-point of a pure component by the addition of the other component. In still other typical cases the components may form mixed crystals, or solid solutions ( 260). The form of the freezing-point curve (III) is very instructive in this case. When mixed crystals can be formed in all proportions the curve has no eutectic point ; every liquid phase gives crystals of a definite composi- tion corresponding to the composition of the liquid phase. The freezing-point curve is uninterrupted in its course. In addition to the above three types we may have various combinations of them. Investigations of the above sort have shown that amalgams of potassium and sodium form compounds, such as Hg 6 Na and Hg 2 Na. In the amalgams of zinc there are neither compounds nor mixed crystals. Mixed crystals are formed in the amalgams of tin, lead, and cadmium. The study of the solid products of the cooling of molten metallic mixtures seems at first somewhat complicated, because we may have not only the solids that result from the slow cooling in accordance with the freezing-point curves, but we may also have solid metallic mixtures formed by the sudden chilling of a hot mixture. This possibility is rather advantageous, however, since we are thus enabled to fix for study at room temperature the relation- ships prevailing at a higher temperature. Chemical Properties. At ordinary temperatures the metal is not affected by the air; at higher temperatures it takes up oxygen to form the oxide HgO, which, however, splits up again into its elements on further heating. Dilute hydrochloric and sulphuric acids do not attack it at ordinary temperatures and dilute nitric acid acts only in the presence of nitrogen dioxide (see 127). Mercury unites instantaneously with the halogens and sulphur. Mercury forms two sets of salts, ous and ic, the former being derived from mercurous oxide, Hg 2 O, and the latter from mercuric oxide, HgO. Mercurous Compounds. 272. Mercurous oxide, Hg 2 O, is dark brown. It is precipitated from the solution of a mercurous salt by caustic soda. It decom- poses at as low a temperature as 100 or in the light, yielding mercuric oxide, HgO, and mercury. Mercurous chloride, Hg 2 Cl 2 , calomel, can be prepared in the wet way by precipitating a dissolved mercurous compound with a chloride, or in the dry way by subliming a mixture of mercuric chloride and mercury. It is a white powder, insoluble in water, 414 INORGANIC CHEMISTRY. [272- but turns dark in the light on account of the separation of metal- lic mercury. Ammonia blackens it by forming a mixture of mercuric ammonium chloride, NH 2 HgCl, and finely divided mer- cury: 2HgCl + 2NH 3 = H 2 NHgCl + Hg + NH 4 C1. Calomel is frequently used as a medicament. The vapor density of calomel has been found to be 117.6 (H=l), which corresponds to the molecular formula HgCl. When calomel evaporates, however, a dissociation into HgCl 2 and Hg occurs; these products unite again on cooling, but they can be previously separated by diffusion. It is for the above reason that the vapor density was found to be half the amount calculated for Hg 2 Cl 2 ; hence the correct formula of calomel is Hg^C^. Here also BAKER noted the influence of traces of water (cf, pp. 333, 334). According to his investigations thoroughly dried mercurous chloride does not dissociate on volatilizing and gives a vapor density which corresponds to the formula Hg 2 Cl 2 . Mercurous bromide and iodide are even less soluble than the chloride. The solubility decreases, as in the case of silver, with an increase in the atomic weight of the halogen. Mercurous nitrate, HgNOs, is formed when cold dilute nitric acid acts on an excess of mercury. It is hydrolyzed by water, OTT a yellow basic salt Hg 2 <^rQ being deposited. It therefore dissolves without decomposition only in dilute nitric acid. The mercurous ion is evidently only very feebly basic. A solution of mercurous nitrate is slowly oxidized by the oxygen of the air to the mercuric salt, but the addition of a little mercury reconverts it into the lower form. Mercuric Compounds. 273. Mercuric oxide, HgO, is red and crystallized when pre- pared by heating mercury or mercury nitrate, but yellow and amorphous when precipitated from solutions by a hydroxide of potassium or sodium. The difference between these lorms seems to be due only to a difference in the coarseness of their grains. Mercuric oxide turns black on heating and red on cooling. Mercuric chloride, HgCl 2 , corrosive sublimate, is manufactured on a large scale by heating a mixture of common salt and mercuric 274.] MERCURIC COMPOUNDS. 415 sulphate; it sublimes over, whence its name. At room tempera- ture 1 part HgCl 2 dissolves in 15 parts H 2 O. It is more soluble in alcohol. The acid reaction of its aqueous solution indicates hydrolytic dissociation; if sodium chloride or potassium chloride is added to the liquid, the reaction becomes neutral because of the formation of a double salt HgCl 2 KC1 H 2 O. This is more soluble in water than sublimate itself. Mercuric iodide, HgI 2 , is yellow when it is first precipitated from the solution of a mercuric salt by potassium iodide, but it soon becomes red. If this modification is heated, it passes over into a yellow form at 150, the original red color returning on cooling, however. There is evidently a transition point here. A similar change of color (red to brown) is observed in the double salt, Cu 2 I 2 2HgI 2 , even at a rather low temperature. On cooling, the red color promptly reappears. This is an excellent example of a sub- stance whose modifications interchange quickly on passing the tran- sition point. Usually the transition occurs slowly. Mercuric iodide dissolves readily in potassium iodide solution. NESSLER'S solution, a very valuable reagent in testing for ammonia, is made by mixing the above mercuric iodide solution with caustic potash. It should be noted, however, that many organic nitrogen compounds give much the same coloration as ammonia with NESSLER'S solution. Mercuric cyanide, Hg(CN) 2 , is obtained by boiling Prussian blue with mercuric oxide. It crystallizes in 'fine large colorless crystals. 274. The mercuric halides, in contrast to the other salts of the mer- curic ion, are only slightly ionized in aqueous solution. For this reason they exhibit some peculiar reactions. On mixing a mercuric solution with one of a chloride, for instance, considerable heat is given off because, undissociated HgCl 2 molecules are formed, while the mixture of solu- tions ordinarily obeys the law of thermoneutrality (238, 2). Again, if mercuric oxide is shaken with a solution of chloride, bromide, or iodide of potassium, the liquid becomes strongly alkaline because of the liberation of potassium hydroxide. This is due partly to the slight ionization of the mercury halides and partly to the combination of the latter with the excess of alkali halide to form very stable alkali 416 INORGANIC CHEMISTRY. [ 274- mercuric halides. The stability of these complex compounds increases with rising atomic weight of the halogen. The same cause explains the reverse fact, viz., that the halogen compounds of mercury are only with difficulty decomposed by alkalies. In order to precipi- tate all the mercury from mercuric chloride a large excess of potassium hydroxide must be employed; mercuric iodide and mercuric cyanide cannot be decomposed by potassium hydroxide alone. Mercuric cyanide is so little ionized that its conductivity can hardly be measured; hence it does not give any of the ordinary mercury reactions, except the forma- tion of the sulphide (since the latter is so very insoluble) . This cyanide can be regarded as a type of compounds rendered inactive because of non-ionization. This low ionization also explains the formation of mercuric cyanide according to the method mentioned above. When mercuric ions and cyanide ions are brought together, even in extremely dilute solution, they must unite to form Hg(CN) 2 molecules. The union of these ions necessitates the sending of more of them into solution by the mercuric oxide and Prussian blue, and so the process goes on until the formation of mercuric cyanide and ferric oxide, Fe 2 3 , is complete. The mercuric halides (especially corrosive sublimate) are very strong antiseptics. It is an interesting fact that in this respect also, they become more effective as their ionization increases. The chloride is a more powerful antiseptic than the cyanide. The addition of metal chlorides diminishes the ionization of sublimate and at the same time reduces its disinfecting ability. The reason why the mercuric chloride for use in sublimate tablets is nevertheless mixed with an excess of common salt is partly that the sub- limate is thus dissolved more rapidly and also because such solutions keep longer than those of the pure sublimate, especially when prepared with well-water. Mercuric nitrate, Hg(NO 3 ) 2 , forms basic salts very readily; on diluting its solution in nitric acid .with water there is deposited a compound Hg(N0 3 ) 2 -2HgO-H 2 O, which is converted into pure mercuric oxide by boiling with water. This shows that the bivalent mercuric ion, also is very feebly basic. Mercuric sulphate is not soluble in water but is converted by the latter into a basic salt. In the presence of much water the yellow compound, HgSO 4 2HgO, is formed. With the sulphates of the alkalies it forms double salts, e.g. HgS0 4 K 2 SO 4 - 6H 2 O, which are isomorphous with the corresponding double salts of magnesium ( 255), iron, etc. 275.] SUMMARY OF THE GROUP. 417 Mercuric sulphide, HgS, is black when precipitated from solu- tion; on being heated in the absence of air it sublimes in dark-red crystals, which are similar to natural cinnabar and are used as a pigment (vermilion). This transformation to the red modification also occurs when black amorphous mercuric sulphide is left in contact with a solution of alkali sulphide. The black form is more easily soluble than the red. After some time red dots are seen in the black mass and they gradually grow till the whole mass is red. SUMMARY OF THE GROUP, 275. Here again a gradual change in the physical properties is to be seen as the atomic weight rises. The following small table presents a few of the constants: Be Mg Zn Cd Hg 9 1 24 32 65 37 112 40 200 1 64 1 75 6 9 8 6 13 6 Mcltinsr-point >900 >651 418 322 39 4 Boiling-point >Zn 920 778 360 1 1 1 In respect to chemical properties it should be noted that all of these elements are bivalent, except that mercury can be con- sidered as univalent in its ous-compounds. Their sulphates unite with those of the alkalies to form double salts of the same type, RS0 4 -R 2 'SO4-6H 2 O (R' = K, Na, NH 4 ); the beryllium double salt alone crystallizes with 3H 2 O. The hydroxides of this group are soluble in ammonia with the formation of complex ions, or else they yield insoluble metal-ammonia compounds (Hg). The neutral salts have a tendency to go over into basic salts. This is especially marked in mercury; in the case of cadmium it is, strange to say, very weak. With the halogen compounds of the three related metals Zn, Cd and Hg the electrolytic dissociation is small; it decreases as the atomic weight of the metal rises and is very slight in the case of mercury. 418 INORGANIC CHEMISTRY. [ 276- ELECTROCHEMISTRY. 276. As early as the beginning of the nineteenth century, when DAVY isolated the alkali metals by means of the electric current ( 223 and 227), there was known to be an intimate relation between electrical and chemical phenomena. BERZELIUS even went so far as to suppose that affinity could be perfectly explained by assuming that the atoms are electrically charged and that these charges are the attractive or repellent forces. The galvanic element has been for a long time a familiar means of converting chemical energy into electrical energy. However, it was not until 1889 that a theoretical explanation of the connection between chemical and electrical phenomena was offered; this explanation by NERNST is not only a very satisfactory one, but it also affords an insight into numerous chemical phenomena. The key to the explanation is the concept of "electrolytic solution tension/ 7 which has already been referred to in a few instances ( 203 and 268). When a metal comes in contact with the aqueous solution of one of its salts a difference in potential develops between the two. This phenomenon is explained by NERNST as follows: Just as a liquid continues to evaporate at its surface until the pressure of the vapor becomes equal to the vapor tension of the liquid, so a salt must continue to dissolve in water (evaporation and solution being analogous processes) until the osmotic pressure of its solu- tion balances the solution tension of the salt. Now, according to NERNST, every metal also has a certain tendency, dependent only on its chemical nature, to force its atoms into solution as ions. This force, called the electrolytic solution tension, comes into action when the metal is immersed in an electrolyte and its strength is the less, the more cations of the metal are already in the solution. The amount of cations sent into the solution is very small, as experi- ment shows, so much so that it cannot be determined by the usual chemical means. The cause of this is not that the solution tension is low, on the contrary, the latter is often very large but that an equilibrium is very soon reached, because, notwith- standing the low concentration of the ions, they carry a very high electrical charge and the negatively charged metal soon attracts its positive ions in the solution with such force that just as many ions are precipitated on the metal as are sent out into the solution. If P represents the solution tension of a metal and p 276.] ELECTROCHEMISTRY. 419 the osmotic pressure of the cations in the solution, there are three possibilities to be distinguished: (1) P>p. The metal then behaves like a salt in contact with its own unsaturated solution. It forces cations into the solu- tion of the electrolyte, so that the solution becomes positively charged and the metal has to take on a negative charge. An equilibrium is soon established. However, if the free positive and negative electricities acquired by the electrolyte and the metal are conducted away by a connecting wire the metal will again send cations into the solution, and this action will continue till p reaches the value of P. (2) P p. There can be no potential difference. (3) P4. The former alone is used in preparing molybdenum and its compounds. It is roasted and so converted into the trioxide, MoOs. The element itself is obtained from its oxides or chlorides by heating them red-hot in a current of hydrogen. The product is a steel-gray powder which fuses with great difficulty to a silvery metallic mass. Sp. g. =8.6. Heating in the air converts it into the trioxide. It is not attacked by hydrochloric or dilute sul- phuric acid, but is readily dissolved by nitric and concentrated sulphuric acid. Molybdenum also has recently found a metallur- gical use in varying the properties of steel. This element is noted for the great variety of its compounds; some of the more important ones may be mentioned here. In addition to the oxides Mo 2 O 3 (weakly basic) and Mo0 2 (in- different) there is molybdenum trioxide, MoOa, which, like CrOs, is an acid anhydride. It is a white powder which turns yellow on heating. It is very sparingly soluble in water. With alkalies it forms molybdates. It has a tendency to form p o I y-molybdates even stronger than the similar tendency of chromic anhydride; 456 INORGANIC CHEMISTRY. [ 296- ammonium heptamolybdate, (NH 4 ) 6 Mo70 2 4-4H 2 (derivable from the acid 7H 2 MoC>4 4H 2 O), commonly known as "ammonium molybdate," is a typical example. The addition of a strong acid to a molybdate solution precipitates white, glistening crystal^lam- inae of molybdic acid, H 2 MoO4, which dissolve in an excess of acid. A solution thus prepared from ammonium molybdate and an excess of nitric acid serves as a test-reagent for phosphoric acid, with which it forms a yellow precipitate of about the composition (NH 4 ) 3 PO4-14MoO3 + 4H 2 O on warming (cf. 146 and 162). Of the chlorides the compounds Mods, MoCU, and MoCl 5 are known. In the oxychlorides MoOCU and MoO 2 Cl 2 molybdenum can be regarded as sexivalent. The chloride MoCl 2 does not exist according to MUTHMANN (neither does MoO) , but a chloride Mo 3 Cl 6 is known. A very characteristic test for molybdic acid (the most common molybdenum compound) is the following: the substance is mixed with zinc and sulphuric acid ; at first a blue coloration (a molybdate of molybdic oxide) appears but it soon turns green and then brown. This brown coloration is due to a salt of the oxide Mo 2 03. TUNGSTEN. 297. The minerals in which this element chiefly occurs are scheelite, CaWO 4 , wolframite, or wolfram, (Fe, Mn)WO 4 , and hubnerite, MnW0 4 . The metal is obtained pure by the method of GOLDSCHMIDT ( 284), i.e., by the reduction of pure tungstic acid with aluminium filings. The metal so obtained is very pure; sp. g., 18.73; melting-point somewhat above 2800. It is malleable and scratches glass. In combination with, carbon it is very much harder. It is very permanent in the air. Sul- phuric acid, hydrochloric acid, aqua regia, and hydrofluoric acid attack it very slowly, but it rapidly dissolves in a mixture of hydrofluoric and nitric acids. Fused caustic potash dissolves it slowly with the evolution of hydrogen. Tungsten is employed in the iron industry, since a small percentage of tungsten increases the hardness of steel in a marked degree (tungsten, or wolfram, steel) . Extremely fine wires of the metal are made use of in some of the newer incandescent electric lights. Tungsten, like chromium and molybdenum, is also characterized by an abundance of compound-types. The chlorides WC1 2 , WC1 4 , WC1 5 , and WC1 6 are known to exist. The lower ones are prepared from the 298.] URANIUM. 457 hexachloride by heating in a current of hydrogen or carbon dioxide. The hexachloride itself is formed by direct synthesis; it is a violet-black crystalline substance; water converts it into the anhydride, WO 3 . Tufogstic anhydride, WO 3 , is obtained by precipitating the hot solution of a tungstate with nitric acid. It is insoluble in water and acids but soluble in alkalies. The addition of an acid to the cold solution of a tungstate precipitates tungstic acid, WO(OH) 4 [=W(OH) 6 -H 2 0]. The latter forms polyacids, like chromic and molybdic acids. Like molybdic acid also it has the property of uniting with phosphoric and arsenic acids to form complex phospho-tungstates and arseni-tung- states . The following is a very characteristic test for tungstates: If stannous chloride is added to a tungstate solution, a yellow precipitate (WO 3 ) is produced. On the addition of hydrochloric acid and warming, a beautiful blue solution (W 2 5 ) is obtained. URANIUM. 298. The principal uranium mineral is uraninite, which usually con- tains some iron. The m e t a 1 is obtained by heating the chloride with sodium or by the electrolysis of the chloride or by the reduction of the oxide with carbon in the electric furnace. It is silvery-white and has a specific gravity of 18.7. It is much more volatile than iron in the electric furnace. When it is in the form of a fine powder it burns in a current of oxygen as low as 170. In the same state it decomposes water slowly at room temperature. When nitrogen is passed over uranium the two elements combine readily at 1000 to form a yellow nitride. Another interesting compound is the carbide C 3 U 2 (obtained from uranium oxide and charcoal in the electric furnace), inasmuch as the addition of water yields not only methane but liquid and solid hydrocarbons. Uranium forms two sets of compounds; in the ous compounds it is quadrivalent (UX 4 ), in the ic compounds sexivalent (UX 6 ). The former pass readily into the latter. The oxide UO 2 has an exclusively basic char- acter; it is obtained by igniting the other oxides in a current of hydro- gen. It was at one time regarded as the metal itself. Uranic oxide, UO 3 , is a yellow powder, prepared by heating the nitrate. The corresponding hydroxide, U(OH) 6 , is not known, but salts of the compound U(OH) 6 -2H 2 0=UO 2 (OH) 2 with acids have been prepared. Since the U0 2 group acts here as a bivalent radical it is called uranyl and its salts uranyl salts, e.g. U0 2 (N0 3 ) 2 , uranyl nitrate, crystallizing with 6H 2 O in beautiful greenish-yellow prisms. Uranium trioxide also has somewhat the character of an acid anhydride; if caustic potash and soda are added to uranyl salt solutions yellow uranates 458 INORGANIC CHEMISTRY. [ 298- (K 2 U 2 O 7 and Na^O?) are precipitated, which are soluble in acids. Uraninite can be regarded as the uranate of uranous oxide, U 3 O 8 =2UO 3 -UO 2 . Both oxides are converted into this U 3 O 8 oxide by heat- ing in the air. Uranium salts are used to impart to glass a beautiful greenish-yellow fluorescence. The detection of uranyl salts is accomplished with the aid of the pre- cipitate, soluble in excess, which they give with ammonium carbonate and by the reddish-brown precipitate with potassium ferrocyanide. For the radioactive properties of uranium see 267. SUMMARY OF THE GROUP. 299. The elements chromium, molybdenum, tungsten, and ura- nium, in connection with sulphur, constitute a natural group in the periodic system. Particularly in the higher oxides there is con- siderable analogy with the behavior of this metalloid. Their acids, for example, all have the formula H 2 RO4. Moreover sulphur also has the ability to form polyacids (pyrosulphuric acid) although it is not so prominent as in the first-named four elements. Several of their salts are isomorphous. The strength of the acids decreases, as in other groups, with rising atomic weight. Another character- istic of all the elements of this group is the great abundance of formula types; it is also very noticeable in the case of sulphur, whose acids are remarkably numerous. The physical properties of these elements have not yet been fully determined, but a few of them are given in the following table: Cr Mo w U Atomic weight 52 96 184 238 5 Specific gravity . . 6 7 8 6 16 6 18 7 Color white white white white Melting-point > 2800 MANGANESE. 300. This element is widely diffused in nature. Its most im- portant minerals are pyrolusite, MnC>2, hausmannite, MnsO^ and rhodochrosite, MnCOs. The metal is of minor importance. It is best prepared by the GOLDSCHMIDT method, i.e. by reducing pyrolusite with aluminium powder, when it is obtained as a regulus of brilliant 300.] MANGANESE. 459 lustre. Sp. g. = 7.2-8.0; m.-pt.= 1245; b.-pt.= 1900. It under- goes surface oxidation readily in moist air, which gives the regulus an iridescence, and whea finely divided decomposes boiling water. It dissolves in acids to form manganous salts. Manganese forms several series of compounds: the manganov* compounds of the type MnX 2 ; the manganic compounds, MnX 3 ; manganic acid, H 2 MnO4, which can be derived from an anhydride MnOs ; permanganic acid, HMnO4 derivable from the oxide Mn 2 O7. Most of the familiar salts of this element are derived from man- ganous oxide, MnO. This oxide, which is prepared by heating the carbonate in the absence of air, is an amorphous green powder, that oxidizes readily in the air to the higher oxide Mn 3 4 . Manganous hydroxide, Mn(OH) 2 , is white when freshly precipitated from solu- tions by an alkali but soon turns brown in the air because of the formation of manganic hydroxide, Mn 2 (OH)6- The solutions of manganous salts are pink (color of the Mn*- ion). The chloride, MnCl 2 , crystallizes with four molecules of water. It can be obtained anhydrous by heating the double salt MnCl 2 -2NH4Cl + H 2 O, since the hydrochloric acid set free hinders the hydrolytic dissociation of the chloride. The sulphate, MnSC>4, crystallizes below 6 with 7H 2 0, above this temperature with 5H 2 O. It forms double salts, such as K 2 SO4-MnS04+6H 2 0, simi- lar to those of magnesium and iron; they are moreover isomor- phous with the latter. Manganous sulphide, MnS, has a pinkish-white color, which distinguishes it from all other sulphides. If ammonium chloride is added to the solution of a manganese salt, no hydroxide is precipitated by ammonia; this is analogous to what is observed with magnesium ( 254). The solution is, however, readily oxidized by the oxygen of the air and brown manganic hydroxide is deposited. The manganic ion Mn"* is only weakly basic. Its salts are almost completely hydrolyzed in aqueous solution. The sulphate gives alums with cesium and rubidium sulphates, which are also very unstable. Manganic oxide, Mn 2 C>3, is obtained from any of the other oxides by heating in an oxygen current. Since dilute sulphuric acid reacts with it, giving manganous sulphate and manganese dioxide, the oxide Mn 2 C>3 is often considered as MnO-Mn0 2 . The 460 INORGANIC CHEMISTRY. [ 300- corresponding hydroxide is soluble in cold hydrochloric acid to a dark-brown solution. It is not certain whether this solution con- tains Mn 2 Cl 6 or MnCl 2 and MnCl 4 ; on being warmed it gives off chlorine and is then known to contain the manganous chloride. Mangano-manganic oxide, Mn 3 04 or MnO-Mn 2 O 3 , is obtained on strongly igniting the other oxides in the air. It is a brownish- red powder. When heated with hydrochloric acid it yields chlorine. Manganese di- (or per-) oxide, MnO 2 , the best- known man- ganese mineral (pyrolusite) , is commercially of great importance in the production of chlorine. In the cold it dissolves in hydro- chloric acid to a very dark liquid, probably containing the tetra- chloride, and gives off no chlorine; when warmed it decomposes into chlorine and manganous chloride ( 25). Since pyrolusite is comparatively expensive; various methods have been devised for reconverting the manganous chloride into the peroxide. One which is of practical importance is the WELDON process. An excess of milk of lime is added to the chloride solution, whereupon air is forced through the warmed liquid. 'The manganous hydroxide which is precipitated undergoes oxidation and is con- verted into calcium manganite, CaMn0 3 ( = CaO-MnO 2 ), which settles down as a black slimy mass: MnCl 2 + 2CaO + = CaMn0 3 + CaCl 2 . The calcium chloride solution is run off and the manganite is used for generating chlorine, since it acts towards hydrochloric acid like a mixture of lime and manganese dioxide. The value of the peroxide depends on the amount of chlorine it can produce with hydrochloric acid. In order to determine this, the mineral, finely pulverized, is warmed with hydrochloric acid and the evolved chlorine passed into potassium iodide solution, whereupon an equivalent amount of iodine is liberated. This iodine can be titrated with thiosulphate ( 93). Manganic acid and Permanganic acid* 301. When manganese compounds are fused with potassium hydroxide in the air or, better, in the presence of an oxidizing-agent (potassium nitrate or chlorate) a green mass results, which is dis- solved by cold water, forming a dark-green solution. On evaporat- ing this solution in a vacuum dark-green rhombic prisms of potas- 301.] MANGANIC AND PERMANGANIC ACIDS. 461 slum manganate, K 2 Mn0 4 , crystallize out, which have a metallic lustre and are isomorphous with potassium chromate. They dissolve in potassium or sodium hydroxide solutions without change, but are decomposed by water with the separation of man- ganese dioxide and the formation of potassium permanganate, KMn04, the latter giving the solution a deep violet color: 3K 2 MnO 4 + 3HoO = 2KMn0 4 + Mn0 2 . H 2 + 4KOH. On account of these changes of color the manganate solution received the name chamceleon minerale, from the early chemists. Both in the solution of a manganate and in that of a perman- ganate we have the anion Mn0 4 ; in the former, however, it is bivalent, in the latter univalent. This causes the difference in the properties of the two ions ; the univalent ion Mn0 4 ' is deep red and resembles the perchloric acid ion in behavior, while the bivalent MnO 4 " is deep green and displays analogy to the SO 4 " ion of sulphuric acid. The bivalent ion MnO 4 " is only stable in alkaline liquids; it is converted by water (more easily by acids) into the univalent ion: 3K 2 Mn0 4 + 4HN0 3 - 2KMn0 4 +Mn0 2 + 4KN0 3 + 2H 2 0, or, written in ions: 6K* +3MnO 4 " + 4H" + 4NO 3 ' - 2K" +2MnO 4 '+Mn0 2 +4K' + 4N0 3 '+2H 2 0. The reaction obviously amounts to a formation of water by the four hydrogen ions and two oxygen atoms which they extract from a bivalent anion Mn0 4 ", the latter being reduced to Mn0 2 . Of the four negative charges which are required to neutralize the four positive charges of the hydrogen ions two are taken from this MnO 4 anion, which is reduced to Mn0 2 , and the remaining two from two other bivalent anions Mn0 4 ", which thus become univalent. The transformation of potassium manganate into the permanganate is effected commercially by passing ozone into its concentrated solution: 2K 2 Mn0 4 +0 3 =2KMnO 4 +K 2 + 2 . The permanganate crystallizes out of the solution and the resulting mother-liquor can at once be used with a fresh quantity of pyrolusite to prepare more manganate. Potassium permanganate, KMn0 4 , crystallizes in beautiful glistening greenish-black prisms of the rhombic system, which 462 INORGANIC CHEMISTRY. [ 301- dissolve readily in water, forming a deep-violet liquid. This salt is isomorphous with potassium perchlorate. All solutions of per- manganates display the same absorption spectrum, viz., five dark bands in the yellow and green, no matter what the base is. It is thus evident that the ion Mn0 4 ' is really the coloring-agent.' The solution of potassium permanganate acts as a powerful oxidizing-agent; in acid solutions two KMnO 4 molecules yield five oxygen atoms : 2KMn0 4 + 3H 2 S0 4 = K 2 S0 4 + 2MnSO 4 + 3H 2 + 50. The process may be regarded as a transformation of the anhydride of permanganic acid, Mn 2 07(=2HMnO 4 H 2 O), into two mole- cules of basic oxide, MnO, and five atoms of oxygen; thus: In neutral or alkaline solutions, however, two KMn0 4 molecules yield only three atoms of oxygen, manganese peroxide being deposited at the same time (transformation of Mn2O7 into 2Mn0 2 +30): 2KMnO 4 + H 2 = 2MnO 2 + 2KOH + 30. Since in oxidations with potassium permanganate in acid solu- tion the deep color of the permanganate is replaced by the very faint color of manganous sulphate, many substances can be titrated with potassium permanganate in acid solution without an indi- cator. Ferrous sulphate is oxidized to ferric sulphate; oxalic acid goes over into carbon dioxide and water; nitrous acid in very dilute solutions is converted into nitric acid ( 126) ; from hydrogen peroxide water and oxygen gas are formed. All these reactions proceed quickly and quantitatively at ordinary temperatures so that they are suitable for titration. Permanganic acid is known only in aqueous solution; however, its anhydride, Mn 2 O 7 , can be obtained. It is prepared by carefully treat- ing dry permanganate with concentrated sulphuric acid. It is a vola- tile, brownish-green, oily liquid, whose vapor explodes easily, yielding oxygen and manganese dioxide. Manganese occupies an isolated position in the periodic system. No elements are known which are related to it as the elements Mo, W and U are to chromium. Moreover, only in its highest stage of oxidation, permanganic acid, does it display analogy with 302.] IRON. 463 the corresponding chlorine compound, HC104. The salts of both acids are isomorphous and both are powerful oxidizing-agents. IRON. 302. Iron is the most useful metal, and is therefore prepared commercially on an enormous scale (approximately 50,000,000 metric tons a year). It occurs only rarely native , e.g. in meteoric rocks. In the form of oxides, sulphides and silicates it is widely diffused in nature and is found in very large quantities. The most important minerals for the iron industry are magnetite T FeaQ4, hematite, FeaOs, and .s?Vfcr?!fe T Ff>f!O 3 . The pyrites (FeS 2; etc.) are worked up into iron after they have been roasted in the sulphuric acid factories. The metallurgy of iron is theoretically very simple; it is based on the ability of carbon to reduce the oxides of iron to the metal at an elevated temperature. This process (smelting) is car- ried out in blast furnaces. The iron ore is first roasted (calcined) to remove volatile substances (H 2 O, C02, S, As, etc.) and loosen up the mineral. Then it is crushed and mixed with a slag-forming substance (flux, see 242), according to the grade of the ore. If the gangue, or earthy matrix, contains much silica or alumina, limestone or dolomite is employed as the fluxing-agent, but ores rich in lime or magnesia are mixed with quartz or aluminous ore to effect the necessary fusion and formation of slag (silicates of Al, Mg and Ca). The blast furnace, previously warmed to the proper tempera- ture or already in operation, is charged from above with alternate layers of coke and the mixture of ore and flux, both being intro- duced in "rounds/ ' or "charges," of definite weight. (Sometimes charcoal or anthracite is used as fuel.) The modern furnaces (Fig. 73) are built of fire-brick encased in iron and are of much lighter construction than those formerly used. They vary greatly in size but consist mainly of a long shaft tapering towards both ends. In order to utilize the escaping hot gases (CO, etc.) an apparatus ("cup and cone") is fitted on the top to conduct them off and also allow the introduction of the charge. The air necessary for the process is forced in, hot, through pipes (twyers) at the bottom 464 INORGANIC CHEMISTRY. [ 302- of the furnace. The burning coke produces carbon monoxide, which is the principal factor in the reduction of the ore: Fe 2 O 3 + SCO = 2Fe + 3C0 2 . The reduced iron sinks downward and comes in contact with carbon at a high temperature; as a result some of the carbon is FIG. 73. BLAST FURNACE. dissolved by it and its melting-point considerably depressed. When a definite stage is reached the fused iron is drawn off below. It is protected from atmospheric oxidation by the slag floating on it. 303.] IRON, 465 The attempt to extract iron from its ores by electrical heating has met with success. STASSANO calculates on the basis of the analysis of the ore the additions which will be necessary to yield a slag as nearly as possible of the composition Si02 + 4 Base, com- presses the finely powdered material to briquetts with the aid of tar in hydraulic presses, and smelts it in a specially constructed arc furnace. 303. It was stated above that the waste furnace-gases contain a con- siderable quantity of carbon monoxide ; therefore a large amount of heat is lost, which could be utilized by burning the monoxide to dioxide. Supposing that this incomplete reaction was due to an incomplete con- tact of carbon monoxide and ferric oxide, manufacturers increased the dimensions of the blast furnaces, particularly in England and America, a height of thirty meters being not uncommon. The ratio of carbon mon- oxide to the dioxide in the escaping gases was not affected however; it was thus demonstrated by very expensive experience that the reduction of ferric oxide by carbon monoxide has a limit. A study of the laws of chemical equilibrium would have led to this conclusion much more quickly and above all much less expensively. These laws teach us that: 1 . In the reduction of ferric oxide by carbon monoxide an equilibrium is established between this action and the oxidation of iron by carbon dioxide. Fe 2 O 3 + 3CO<=2Fe + 3C0 2 . 2. The ratio CO:CO 2 must be independent of the pressure, since no change in the volume of the gas takes place (51). 3. This ratio varies only slightly with the temperature, since very little heat is generated in the reaction. An experimental investigation conducted at a few different tempera- tures and pressures would have sufficed to determine the ratio CO:CO 2 . The result, when compared with the ratio CO:CO 2 of the waste gases, would thus have shown that little could be gained by an increase of the furnace dimensions. This illustrates in a very striking way the value of physical chemistry for industrial processes. Efforts are now being made to utilize the waste gases in other ways, such as by burning them under the boilers of steam engines or in wind heaters (for heating the blast air, or "wind"). In recent years it has been found that greater efficiency is attained by using the hot waste- gases directly in gas engines for motive-power. 466 INORGANIC CHEMISTRY. [ 304- 304. The properties of iron are influenced in great measure by the slight admixtures which it contains, particularly by the carbon. The percentage of carbon forms the ordinary basis of classification of the different grades of iron under the heads> pig iron and malleable iron; however, in the industrial world this classification is not always adhered to. Pig iron, or cast iron, contains 2.3-5.1% carbon. It fuses very easily but there is no previous softening; hence it is not malleable. It is brittle. Pig iron is the direct product of the blast furnaces and the iron is therefore mixed with small amounts of silicon, phos- phorus, sulphur, etc. The presence of manganese makes it coarsely crystalline and it is then known as spiegel-eisen. This is utilized mainly for steel. Refined iron, containing less than 2.3% carbon, is harder to fuse, but is extensible and malleable, and the more so the less the impurities. If the carbon amounts to 2.3-0.5%, the iron can be hardened; in this manner steel is obtained. If there is less than 0.5% carbon, it can no longer be hardened; this is wrought iron. It is obvious that between these main varieties there are numerous intermediate sorts, which are prepared in such a way as to suit the purpose for which they are intended. The immense commercial importance of the iron-carbon system has led to extensive investigations regarding it, notwithstanding that such investiga- tions are attended by great experimental difficulties, partly because of the very high temperatures involved. Because of these difficulties it is not yet possible to give an entirely satisfactory representation of the equilibrium conditions concerned. BAKHUIS ROOZEBOOM, CHARPY, ROBERT-AUSTEN and others have succeeded in working out the accompanying graphic representa- tion which indicates the behavior of the system in the main at least. To appreciate this diagram it is necessary in the first place to know a few general facts regarding the components that are now regarded as existing in the iron-carbon system. Distinction is made between: 1. ferrite, or chem- ically pure iron (pure wrought iron); 2. martensite (steel), a solid solution of carbon in iron. It is so regarded because microscopic studies have shown that martensite is always homogeneous in spite of its changing carbon content, which may be as high as 2%. 3. cementite (the commercial white cast iron), an iron-carbon compound of the formula Fe 3 C; and 4. perlite> carboniferous iron (0.85%), that is seen under a high-power microscope to be heterogeneous and is regarded as a eutectic mixture of ferrite and cementite. The solidification curve of a binary system ( 237) does not take a nor- mal course in the iron-carbon system. Three circumstances complicate the 304.] IRON. 467 situation. The first is that pure iron does not separate out of the molten mass, but that we obtain the solid solution martensite. The second is that changes continue to occur in the cooling mass after complete solidification; the third that other substances separate out with very slow cooling than with quick cooling. We may consider first the case of slow cooling, where the equilibria that establish themselves between solid and liquid phases are presumably stable: Let us assume that we have liquid iron with a carbon content below 4.3%. On cooling the liquid the iron begins to solidify at a definite temperature (the point G! in Fig. 74); however, it is not pure iron, but a solid solution of 1500 c (1000 G 700' " FIG. 74. IRON-CARBON SYSTEM. carbon in iron that separates out; its composition is shown in the diagram by the point d t . If the carbon content of the fused iron is a different one, we have separating out at c 2 , for example, the solid substance, whose com- position is again given by the point d 2 . Thus for every solidification point of the curve AC we can find a point d lf d 2 , etc., that gives the composition of the solid substance which begins to separate out. The curve AD is the geometrical locus of these points. If, therefore, a horizontal line is drawn through the triangle ADC, the point c gives the composition of the liquid solution which solidifies at the corresponding definite temperature (indicated by the ordinate) and the point d the composition of the solid solution which begins to separate out at that temperature. At C the eutectic point is reached. Along CB graphite separates out; at C itself a mixture of graphite and martensite, the composition being given by D. The point C is at 1130 and 4.3% of carbon. The martentite formed at this tem- perature contains 2% of carbon. 468 INORGANIC CHEMISTRY. [304- Below DC all is solid; but, as we have already explained, changes con- tinue to occur in the solid mass. For example, if martensite is heated, it breaks up with the formation of graphite. The curve DE represents the change of composition of the solid solution with falling temperature or, in other words, it represents the equilibrium between graphite and the mixed crystals (solid solution) at different temperatures. Around E, where the temperature has reached about 700, the martensite contains only about 0.85% carbon. At the point E the formation of ferrite begins. Finally, the curve EG indicates the composition of the solid solutions from which ferrite separates out. Hence, if the martensite contains less than 0.85% of carbon, ferrite is deposited along EG, exactly as ice separates out of a dilute salt solution with falling temperature. If the cooling is sudden, other phases are formed, the limits of which are represented in the figure by lines, which are readily under- stood. Instead of the eutectic point C, at which graphite and martensite separate out, we have a eutectic point at C l} very close to C, where cementite, Fe 3 C, separates out with the martensite. Further, the line C^E^ represents the equilibrium between cementite and the mixed crystals (martensite). At E l ferrite is formed together with cementite. Martensite changes over at this temperature into a eutectic mixture of these last two substances, which has the fine conglomerate structure so characteristic of eutectics and is known as " perlite." Although this whole system shown by lines is meta- stable, it can exist for an indefinite period at ordinary temperatures because of the reduction of the velocities of reactions which might restore the stable forms. It is evident from the above that with slow cooling martensite entirely disappears. If the cooling is rapid, however, as in the hardening of steel, martensite can be brought to exist at ordinary temperatures even though it is in a metastable condition ; its transformation velocity is then extremely small. If the hardened steel is reheated, it changes over partially into the soft conglomerate of ferrite and cementite; this is what takes place in the "tempering" of steel. Small admixtures of other elements have an effect on the properties of iron equally as great as that of carbon. The presence of silicon has about the same effect as that of carbon, but it is less intense. Sulphur even in a small amount renders the iron brittle when hot and, therefore, useless for forging. On this account sulphurous ores as such are unsuitable for the manu- facture of iron. Phosphorus makes the iron brittle at ordinary temperatures. It should also be mentioned that as a general rule the effect of these admixtures is strongly modified by the pres- ence of others. 305. From the crude pig iron, the direct product of the blast furnace, the other varieties of iron are prepared. For this purpose 305.] IRON. 469 it must be freed from silicon, sulphur, phosphorus, etc., as well as from a large portion of its carbon. The most important process for accomplishing this commercially is the BESSEMER process. The pig iron is fused and run into a pear-shaped apparatus, or con- verter (Fig. 75), in the bottom of which are holes through which air is blown in. Thus by the oxidation of silicon, manganese and a little iron and without the use of fuel the temperature is raised high enough to effect the burning of the carbon. The BES- SEMER process is easier controlled if the elimination of carbon is continued past the steel stage and until molten wrought iron is formed, whereupon enough carboniferous iron is added to furnish FIG. 75. CONVERTER. steel with the desired percentage of carbon. At the completion of the process the converter is emptied by tipping. In some European mills a basic converter lining containing an excess of lime and magnesia is used. The phosphorus in the ore combines with the bases to form phosphates, which enter the slag, and this so-called "Thomas-slag" is used in large quantities as a fertilizer. The only successful rival of the BESSEMER process is the SIEMENS, or open-hearth, process. By employing a special furnace and gaseous fuel a mixture of cast iron and wrought iron (together with some iron ore) in the proper proportions can be fused together 470 INORGANIC CHEMISTRY. [305. so as to produce a very good steel. A basic lining can also be used with this process. The increased demand for special steels, where physical and chemical conditions have to be regulated carefully, has given greater significance to the old crucible process, the steel being made in graphite crucibles in a laboratory manner but on about ten times the laboratory scale. Recently electric furnaces of the arc and induction type have been found very successful in pro- ducing (i crucible " steel and with much less labor than the FIG. 76. HEKOULT FURNACE. crucible process requires. The cradle-shaped HEKOULT furnace is shown in the accompanying combined end-view and vertical section (Fig. 76). M is the molten metal, S the slag, and E one of the carbon electrodes; B is brick lining and L a layer of magnesium silicate. As the resistance of the metal is small compared with that of the slag and the air, most of the heat is generated at the surface, where the chemical action goes on between the slag and the metal. The furnace is eventually emptied by rocking forward. Steel, however made, is a very complex alloy, containing 305.] IRON. 471 carbon and manganese, 0.10-1.50%; silicon 0.02-0.25%, sulphur and phosphorus 0.01-0.10%; and possibly copper, arsenic, alu- minium, oxygen, nitrogen, and cyanides, and is capable, as has been explained, of containing the iron and carbon in various combinations. Steel of the above description is " ordinary " steel. Recently a large market has developed for " special " steels, having new qualities, especially with respect to hardness and brittleness, and serving new purposes, notably in tools, military materials and materials of construction. They may be produced by (1) changing the physico-chemical character with respect to the iron-carbon system, (2) removing harmful occluded gases, (3) combining other elements chemically with iron or carbon or both, and (4) adding other elements to form isomorphous solutions with iron. Steel becomes very hard and brittle, for instance, when it is suddenly cooled from a high temperature. If, however, it is then heated for a definite period and allowed to cool slowly, it becomes more or less tempered according to the temperature, i.e., it can be made to have any desired hardness and elasticity (within certain limits). Of the special alloy steels the nickel steels, chrome-nickel steels and chrome-vanadium steels seem to be most important. The maximum hardness of steel is reached when it contains 1-2% carbon; if, however, some manganese (up to 8%) or chro- mium (up to 1%) is added, a much harder modification of steel is produced. The addition of nickel gives a tougher steel, which is especially valuable for armor plate. Tungsten (cf. 297) and molybdenum are also added for different purposes. In any case, however, a careful heat treatment is essential to develop the desired properties. The production of wrought iron from pig iron is usually accomplished by the puddling process. Pig iron is melted in a reverberatory furnace lined with iron ore (oxide) ; the carbon and also the silicon are oxidized (and so removed) partly by the action of the air, but mainly by that of the ore, which is stirred in with the metal. The violent reaction due to escaping carbon monoxide gives the process the name of "pig- boiling." The iron is then allowed to become pasty, when it is worked up into large masses (blooms'), which are removed and ham- mered and rolled. The cinder is thus squeezed out and the iron is formed into bars. 472 IXORGANIC CHEMISTRY. [ 3Q5- Chemically pure iron is obtained electrolytically and by reduc- ing the oxide or chloride in a current of hydrogen. If the reduc- tion takes place at a low temperature, the resulting iron powder is pyrophoric ( 203). It is a silvery-white, lustrous metal with a specific gravity of 7.84 and a melting-point as high as 1520. It boils at 2450. It is the most magnetic of the metals; pure iron and wrought iron can be magnetized only temporarily; steel, however, permanently. Iron is permanent in dry air or in water free from air (CO 2 ). In moist air it rusts rapidly ( 279), forming ferric hydroxide; as the rust does not form a compact film, it keeps on forming. The rusting of iron is greatly retarded by contact with water contain- ing a little alkali or salts of alkaline reaction. In a soda solution, for instance, iron remains bright. The rusting of iron in contact with water can be explained by assuming that the oxygen dissolved in water endeavors to form hydroxyl ions with the hydrogen ions. In order to compensate their negative potential the iron sends its positive ions into the solution; in a short time the solubility product of ferric hydroxide is reached and the latter is. deposited; in other words, the iron rusts. Now, if hydroxyl ions are previously introduced into the liquid by the addition of a base or a salt of alkaline reaction, the ionization of the water is diminished so much that the oxygen can find almost no hydro- gen ions with which to form hydroxyl ions; therefore the iron does not send any more ions (276 and 277) into the solution and rusting is greatly retarded. Iron dissolves readily in hydrochloric and sulphuric acids with the evolution of hydrogen. At red-heat it decomposes water, but the oxide is also reduced by hydrogen, so that an equilibrium results : 3Fe + 4H 2 0<=Fe 3 O 4 + 4H 2 . In nitric acid (not too concentrated) iron dissolves readily with the evolution of nitric oxide, NO, but if the iron is first dipped in concentrated nitric acid and then rinsed off it becomes indif- ferent to the action of nitric acid. This so-called "passivity" of iron is probably caused by a very thin coating of oxide. Iron forms two sets of salts, the ferrous and the ferric. 306.] FERROUS COMPOUNDS. 473 Ferrous Compounds. 306. In the ferrous condition iron has only basic properties. Ferrous oxide, FeO, is obtained by reducing ferric oxide with carbon monoxide. It is a black powder, which oxidizes easily on warming. Ferrous hydroxide, Fe(OH) 2 , is precipitated from ferrous salt solutions as a pale green gelatinous substance by the addition of an alkali; it oxidizes very rapidly in the air to ferric hydroxide. Ferrous chloride, FeCl2, is formed on dissolving iron in hydro- chloric acid; it crystallizes from this solution in green monoclinic prisms containing four molecules of water. The anhydrous salt is obtained as a white sublimate when iron is heated in dry hydro- chloric acid gas. With potassium chloride and ammonium chlo- ride ferrous chloride forms well crystallized double salts, e.g. FeCl 2 -2KCl + 2H 2 O. Ferrous sulphate, FeS04+7H 2 (green vitriol, copperas), is the most familiar ferrous salt. It is prepared commercially, princi- pally by dissolving up the waste metal of steel- wire factories in sulphuric acid, but also by partially roasting pyrite, whereby fer- rous sulphide, FeS, is formed; the latter is left exposed to the ah-, when it oxidizes gradually to ferrous sulphate, which can be dis- solved out. It crystallizes in large, bright green, monoclinic prisms, which effloresce slightly and at the same time become coated with a brown layer of basic ferric sulphate. The double salts such as FeS0 4 -(NH 4 ) 2 S04+6H 2 O, MOHR'S salt, are not so liable to oxi- dation; for this reason use is frequently made of MOHR'S salt to standardize permanganate solutions ( 301). Iron vitriol has numerous uses, e.g. for making ink, in dyeing, as a disinfectant (it absorbs both ammonia and sulphuretted hydrogen and is there- fore used to dispel bad odors), etc., etc. Ferrous carbonate is somewhat soluble in water containing carbonic acid and is therefore often present in natural waters ( 17). The basic carbonate which is precipitated from a ferrous solution by soda oxidizes rapidly in the air to ferric hydroxide. The latter is also deposited from chalybeate waters on standing in the air for a time. Ferrous carbonate is only known as a mineral (siderite, 302). 474 INORGANIC CHEMISTRY. [307- Ferric Compounds. 37 The ferric ion has only very slightly basic properties. Ferric salts of weak acids ; such as carbonic acid, do not exist. In aqueous solution most of the ferric salts, even those of strong acids, are partially hydrolyzed. For that reason they are brown- ish-red, since this is the color of ferric hydroxide in colloidal solu- tion. On the addition of an excess of sulphuric or nitric acid this color disappears, because there is no longer any hydrolysis. From this it appears that the ferric ion itself in aqueous solution is only slightly colored. The ferric salts are readily converted into ferrous salts by reducing-agents. Ferric oxide, Fe20s, iron sesquioxide, is formed on heating various iron compounds in the air and is manufactured by igniting green vitriol ( 79). It is a dark-red powder and finds use as a pigment (colcothar) and in polishing glass, etc. Ferric hydroxide separates out as a reddish-brown hydrogel, Fe 2 03+ftH 2 O, when a ferric salt solution is treated with an alkali. The freshly precipitated hydrogel dissolves in a solution of ferric chloride or acetate. If this solution is dialyzed, a pure colloidal solution of the hydroxide is finally obtained; from this the hydrogel is reprecipitated by a small amount of alkali or acid. Ferrous ferric oxide, Fe 3 O 4 , also called ferroso-ferric oxide or magnetic iron oxide, occurs in nature as magnetite. It is produced by heating iron in steam ( 305). Ferric chloride is obtained by passing chlorine into a solu- tion of ferrous chloride. It crystallizes at different temperatures with different amounts of water, being an example of the case described on p. 341. On heating the salt hydrochloric acid escapes with the water of crystallization. Anhydrous ferric chloride can be prepared by heating iron in a current of dry chlorine. Between 320 and 440 the vapor density is approximately that calculated for Fe 2 Cl 6 ; between 750 and 1050 it falls to half, indicating a splitting off of chlorine or a dissociation into 2FeCl3. The reddish-brown color of the aqueous solution of ferric chloride must be ascribed chiefly to un-ionized FeCls molecules, for the salt has this same color when dissolved in ether, in which no ioniza- tion occurs. In part, also, this color comes from ferric hydroxide. 308.] FERRIC COMPOUNDS. 475 which is formed by hydrolytic dissociation. This dissociation increases on warming the dilute aqueous solution, for a very dilute, almost colorless solution of ferric chloride turns reddish-brown on boiling. When cooled the liquid gradually resumes its original color. Ferric sulphate, obtained by dissolving ferric oxide in sulphuric acid, forms alums, e.g. potassium iron alum, K 2 SC>4 ^62(864) 3+ 24H 2 0. When a ferrous salt is converted into a ferric salt in aqueous solution the bivalent ferrous ion is transformed into a trivalent ferric ion. The oxygen required for the conversion serves to oxidize the hydrogen ions of the acid (which must be added) to water, whereupon these hydrogen ions surrender their charge to the iron ions: 10 +2(H' + C10 + O =2(Fe- + 3Cl / ) +H 2 O. Ferrous chloride Hydrochloric acid Ferric chloride Inversely, the reduction of ferric salts to ferrous salts can be explained by supposing that every ferric ion gives up a third of its charge to another atom and thus makes the latter an ion or neu- tralizes its charge. Salts of iron are also known which are derived from the hypothetical oxide FeO 3 . They are obtained by heating iron filings with saltpetre or passing chlorine into an alkaline suspension of the ferric oxide hydrogel. From such solutions potassium ferrate, K 2 FeO4, crystallizes out in dark- red prisms, isomorphous with the chromate and sulphate of potassium. These crystals are readily soluble in water, but their dark-red solution soon decomposes with the separation of ferric hydroxide and oxygen gas. The free ferric acid is unknown. 308. Iron unites with cyanogen to form complex and unusually stable anions, viz., the jerrocyanic ion [Fe(CN) 6 ]"" and the ferricyanic ion [Fe(CN) 6 ]'". Their best-known salts are potas- sium ferrocyanide, K4Fe(CN) 6 -3H2O, and potassium ferricyanide, K 3 Fe(CN) 6 , the yellow and red prussiates of potash, respectively. The ionization of the complex ions themselves is so slight that they give none of the ordinary reactions for iron. For the commercial manufacture of yellow prussiate of potash two processes are used: .In the first, animal refuse (e.g. blood) is charred, yielding a black, highly nitrogenous mass. This is 476 INORGANIC CHEMISTRY. [308- ignited with potash and iron filings. After cooling, hot water is added and the mixture filtered; from this filtrate the yellow prussiate crystallizes out on standing. This salt is not formed until the ignited mass is treated with water, for yellow prussiate is decomposed by heat and cannot therefore be present in the ignited mass. The latter probably contains potassium cyanide, iron and iron sulphide (animal refuse always contains sulphur compounds') . These substances can interact according to the equations: 6KCN+FeS =K 4 Fe(CN) 6 +K 2 S; Fe(CN) 2 + 4KCN = IQFe(CN). The second process is employed in illuminating-gas factories, for the unpurified gas contains a little cyanogen and prussic acid. After being freed from tar and ammonia it is passed through a washer (scrubber) containing a solution of potash in which ferrous carbonate (ferrous sulphate + potassium carbonate) is suspended. The following reactions, among others, are known to go on here: FeCO 3 + 2HCN<=Fe(CN) 2 + H 2 O + CO 2 ; K 2 CO 3 + 2HCN^2KCN + H 2 O + CO 2 . Notwithstanding that these reactions are reversible, the hydro- cyanic acid can be quantitatively fixed in this way, because the ferrous cyanide and potassium cyanide interact to form potassium ferrocyanide, which is but very slightly affected by carbon dioxide. Potassium ferrocyanide, K 4 Fe(CN) 6 -3H 2 O, forms large sulphur- colored crystals. Its three molecules of water can be expelled by gently warming, whereupon the salt is left as a white powder. It is not poisonous. With dilute sulphuric acid it produces prussic acid on warming; with concentrated sulphuric acid it yields carbon monoxide. The free ferrocyanic acid, H 4 Fe(CN)6, separates out as a white crystalline precipitate when concentrated hydrochloric acid is added to a strong solution of potassium ferrocyanide. The pre- cipitate soon turns blue in the air on account of the formation of Prussian blue (and partial decomposition as well). Various salts of this acid have characteristic colors and are insoluble; hence potassium ferrocyanide finds use in analysis. It is an interesting 309.] COBALT AND NICKEL. 477 fact that this compound of iron can serve as a distinguishing reagent for ferrous and ferric compounds. The ferrous salt of ferrocyanic acid is white, but in the presence of air it passes rapidly over into the blue ferric salt (Prussian bluea valuable pigment). The copper salt ( 40) is brownish-red, the zinc salt white, etc. Sodium nitroprusside, Na 2 Fe(CN) 5 (NO) -2H 2 O, is formed by the action of nitric acid on sodium ferrocyanide. It crystallizes in ruby-red prisms and is a delicate reagent for alkali sulphides, whose solutions it colors violet. Potassium ferricyanide, K 3 Fe(CN) 6 , red prussiate of potash, is formed from the yellow prussiate by treating a solution of the latter with chlorine or bromine : K 4 Fe(CN) 6 +C1 = KC1 +K 3 Fe(CN) 6 . It appears in dark-red crystals, which are readily soluble in water. The aqueous solution is unstable. The salt is often employed as an oxidizing-agent in alkaline solution, being itself converted into the ferrocyanide: 2K 3 Fe(CN) 6 +2KOH ==2K 4 Fe(CN) 6 +H 2 +0. Iron forms some very peculiar compounds with carbon monoxide: Fe(CO) 4 and Fe(CO) 5 . They are produced when carbon monoxide is passed over finely divided iron at 80, or at ordinary temperatures if the gas is under pressure. Iron vessels which have held compressed illuminating-gas for some time are more or less attacked by the carbon monoxide of the gas, for if gas which has been kept in such a vessel is allowed to escape through a hot glass tube an iron mirror is formed on the inside of the tube. COBALT AND NICKEL. Cobalt. 309. The two best-known minerals of this metal are smaltite, CoAs 2 , and cobaltite, or cobalt glance, CoAsS. The metal is ob- tained by calcining these minerals and reducing the resulting cobalto-cobaltic oxide, Co 3 4 , with carbon (or hydrogen). It has a pink color and a high lustre. Sp. g. 8.9;. m.-pt., 1490. It is magnetic but much less so than iron. It is indifferent to the air. Hydrochloric and sulphuric acids dissolve it very slowly but it readily forms a nitrate with nitric acid. 478 INORGANIC CHEMISTRY. [g 300- Besides the oxide, Co 3 4 , just referred to there are two others, cobaltous oxide, CoO, and cobaltic oxide, Co 2 O3. The salts are all cobaltous, corresponding to the bivalent ion Co". COBALTOUS COMPOUNDS. The solutions of the salts are red; hence this is the color of the cobalt ion. The non-ionized cobalt salts are blue, e.g., the anhydrous CoCl 2 , the silicate, etc. This difference in color enables us to tell readily whether a cobalt salt in solution is ionized or not. Thus in concentrated solutions, for instance, all those cir- cumstances which reduce the ionization cause a change of color from red to blue, e.g. when a concentrated cobalt chloride solu- tion is warmed or treated with hydrochloric acid. That the ioni- zation is diminished by warming was mentioned in connection with cupric chloride ( 244). Cobaltous chloride, CoCl 2 -6H 2 0, forms red monoclinic crys- tals, which turn blue on heating because of dehydration. Cobalt sulphate, CoSCU -7H 2 0, is obtained in dark-red monoclinic prisms and is isomorphous with FeS04-7H 2 0. It forms double salts with alkali sulphates, e.g. K 2 S0 4 -CoS0 4 + 6H 2 0. Cobalt nitrate, Co(N03) 2 -6H 2 0, appears in red hygroscopic prisms. Cobalt sili- cate is very deep blue; hence its use for coloring glass. Pulverized cobalt silicate serves as a pigment (smalt) in painting, etc. TH- NARD'S blue is a pigment, obtained by igniting cobalt salts with alumina. COBALTIC COMPOUNDS. 310. Cobaltic oxide, Co 2 03, is obtained by igniting cobalt nitrate. It is a black powder, which passes over into cobalto- cobaltic oxide, 00364, at red heat and at white heat yields cobaltous oxide. It has the character of a peroxide; for by the addition of sulphuric acid it is converted into a cobaltous salt with the evolu- tion of oxygen and it yields chlorine with hydrochloric acid. How- ever, in cold dilute hydrochloric acid it dissolves without generat- ing scarcely any chlorine. Like iroa, cobalt also forms complex ions, of which those with cyanogen are very stable. There are cobalt salts corresponding in composition to the yellow and the red prussiates of potash; the salt Iv3Co(CN)6, potassium cobalticyanide, crystallizes in colorless 3ii.[ cu/tM.r A.VD \ICKI:L. 479 rhombic prisms. A peculiar complex ion occurs in the potassium cobaltic nitrite, 6KNO 2 -Co 2 (N0 2 ) 6 +nH 2 0, or K 3 .Co(NO 2 ) 6 + wH 2 0. It is formed on treating a solution of a cobalt salt, with potassium nitrite and acetic acid. It is a yellow crystalline pre- cipitate, which is very slightly soluble when potassium ions are present, in excess in the liquid. Cobalt also forms numerous complex ions with ammonia (317). Nickel. 311. Nickel occurs in niccolitc, NiAs, and nickel glance, or f/rr.sWo/7///r, XiAsS. Ks|)cci;illy iiu| >nrl ;i nl is the nickel silic;ite, garnierite, H 2 (Ni,Mg)Si04-faq(?), which was discovered by GAB- NIER in New Caledonia, where it occurs in enormous quantities. Canada, too, has some rich nickel deposits. From this ore the nickel is obtained by a blast-furnace process similar to that for iron. The discovery of garnierite marked the beginning of a new era in the nickel industry. Much nickel is refined electro, lytically. Nickel is almost as white as silver, is very tough and has a high metallic lustre. Sp. g..= 8.8-9.1; m.-pt. - 1452. It is feebly magnetic. It dissolves sparingly in hydrochloric and sulphuric acids but freely in nitric acid. It is permanent in the air. It is employed in nickel-plating metallic objects and as a con- stituent of several alloys. German silver contains about 50% copper, 25% nickel, and 25% zinc. The nickel coins of Germany and the United States consist of 75% copper and 25% nickel. The use of nickel to vary the properties of iron has already been mentioned ( 305). The oxides of nickel, NiO and Ni 2 0a, are very similar to those of cobalt. The nickelous oxide, NiO, is the only one which forms s;i,lts. Nickel chloride, NiCl 2 -6H 2 0, yields green monoclinic prisms. When healed it turns yellow on account of loss of water. Nickel sulphate, NiSO 4 -7H 2 0, crystallizing in green rhombic prisms, is isomorphous with the corresponding ferrous and other salts and also forms analogous double salts. Nickelic oxide, Ni 2 3 , also behaves as a peroxide; when warmed with hydrochloric acid it yields chlorine gas and nickel chloride. 480 INORGANIC CHEMISTRY. [311- Nickel carbonyl, Ni(CO)4, is formed when carbon monoxide is led over finely divided nickel at ordinary temperatures. A state of equilibrium results here, viz.: Ni+4CO<=Ni(CO) 4 , which is displaced to the left with rising temperature, since the decomposition of nickel carbonyl takes place with a considerably absorption of heat ( 103). Even as low as 60 the decomposition is of an explosive nature. It follows from the above equation that an increase of pressure ( 122) must greatly increase the propor- tion of nickel carbonyl formed. Experiments confirming this showed at the same time that both the formation and the decom- position of this compound are very sensitive to traces of foreign substances. Nickel carbonyl is a colorless, highly refractive liquid, which boils at 43 and congeals (crystalline) at 23. When heated in the air it burns with a very sooty flame. This compound is of aid in extracting nickel from low-grade ores. Nickel also forms a complex ion with cyanogen. On dissolving nickel cyanide in an excess of potassium cyanide the compound K 2 Ni(CN) 4 is produced; it is, however, unstable, being decomposed by hydrochloric acid with the deposition of nickel cyanide, Ni(CN) 2 . 312. A peculiar property is exhibited by the sulphides of cobalt and nickel, CoS and NiS. Hydrogen sulphide does not precipitate these sulphides from acid solutions, but, once precipitated (by ammonium sulphide), they are apparently not redissolved by dilute acids. This is contrary to the general rule of 146 (see also 73), for the sulphide should either be precipitated by hydrogen sulphide from a feebly acid solution (e.g. CuS), which is the case when the solubility product is very small, or else, when the solubility product is larger, it should dissolve in dilute acids, as is the case with ferrous sulphide. As a matter of fact, however, no real anomaly exists here, for the rate of solubility of these sulphides is only very slow under the usual conditions of the reaction, viz., dilute acid and room temperature. It increases with the concentration of the acid, temperature of reaction and fineness of grain of the precipitate. Nickel sulphide is soluble in alkali sulphides immediately upon its formation, but when once deposited in the solid state it is insol- uble, or nearly so. This is seen when a nickel solution is treated' 313.] PLATINUM METALS. 431 with tartaric acid and then with an excess of sodium hydroxide, no nickelous hydroxide being precipitated. If hydrogen sulphide is passed into this solution a very dark-colored liquid results, from which nickel sulphide is deposited only very slowly. The same is true of cobalt in very dilute solution; in concentrated solution, however, the cobalt sulphide soon passes over into the insoluble modification and separates out. PLATINUM METALS. 313. Under this head are included the metals ruthenium, rho- dium, palladium, osmium, iridium, and platinum. They occur only in metallic form and are associated together in mixtures or combinations. The principal deposits are in the Ural and Caucasus, but smaller quantities are also found in Colombia, Brazil and Borneo. The Ural yields 95% of the total production. The most important of these metals is platinum. The platinum ores usually contain admixtures of iron, gold, etc. This group falls into two subdivisions : the light metals, ruthenium, rhodium and palladium, and the heavy metals, osmium, iridium and platinum. The two sub-groups differ con- siderably in atomic weight and specific gravity: Light. Heavy. Ru Rh Pd Os Ir Ft Atomic weight. . . . Specific gravity. . . 101.7 12.26 103.0 12.1 106.5 11.9 191 22.4 193.0 22.38 194.8 21.45 A complete separation of the platinum metals from each other is extremely difficult, in the first place because their properties are very similar, and in the second place because their behavior is con- siderably modified by their mutual presence a fact which indi- cates the existence of compounds with each other. Thus, for instance, platinum dissolves readily in aqua regia while pure iridium is insoluble in it; nevertheless when an alloy of the two metals is treated, with aqua regia, some of the iridium is carried into solu- 482 INORGANIC CHEMISTRY. [313- tion. Further, the presence of iron (which occurs in all platinum ores) is often very disturbing; for example, pure platinum solu- tions are not precipitated by soda or barium carbonate, but if iron is present more or less platinum comes down with the hydroxide of iron. In spite of these difficulties platinum, palladium, rhodium and iridium can now be purchased in a remarkably pure state. For the manufacture of platinum and the other metals of the group in the pure state the various factories employ their own secret methods. In general the procedure is about as follows: The ore is first treated with aqua regia to dissolve out the major part of the "noble" metals, leaving in the residue the alloy iridosmine, besides more or less sand. The ore thus consists not of one alloy, but of two: the crude platinum and the iridosmine. Both contain all six platinum metals, although in different relative amounts. The crude platinum contains, besides platinum, principally palladium, rhodium, and iridium, while the iridos- mine, as its name indicates, consists mainly of iridium and osmium. It is comparatively easy to separate out the osmium and ruthenium; they form volatile oxygen compounds and can therefore be removed by distillation. Platinum and iridium give difficultly soluble com- pounds with ammonium chloride, which in turn are reduced to the metal form by ignition. However, if the solution of the crude platinum is precipitated with ammonium chloride, the resulting precipitate is found to contain considerable amounts of rhodium and palladium, and an extended procedure is necessary for the isolation of the pure metals. On the other hand, it is not possible to precipitate the platinum and iridium completely in this way ; the filtrate from the ammonium chloride contains palladium and rhodium, with smaller amounts of iridium and platinum, and, in order to work it up, a further. complicated procedure is necessary. Ruthenium. 314. This steel-gray metal occurs only in very small quantities; it is hard, very brittle, and very difficult to fuse, a temperature of at least 1800 being necessary. Even when finely divided it is but very sparingly soluble in aqua regia, forming Ru 2 Cl 6 , but when alloyed with platinum, it dissolves readily. The compound RuCl 4 is known only in double salts. As a powder the metal oxidizes in the air to RuO and Ru 2 3 . Ruthe- nium also forms characteristic salts, in which it plays the part of an acid. Potassium ruthenate, K 2 RuO 4 , results from fusing ruthenium with caustic potash and saltpetre. It crystallizes with !H 2 Oin black prisms of 314.] PLATINUM METALS. 483 a greenish lustre. With water it forms a dark orange-colored solution. Its conduct reminds one of potassium manganate, for under the influence of dilute acids it is converted into potassium perruthenate, KRu0 4 , with the simultaneous precipitation of a black oxide, Rh 2 O 5 (or RuO 2 ?). It crystallizes in black octahedrons of metallic lustre, which dissolve in water to a dark-green solution. A peculiar compound is the tetroxide, RuO 4 , which volatilizes when chlorine is passed into the concentrated solution of potassium ruthenate. It can be solidified by cooling, when it forms a golden crystalline mass, fusible at 25.5. There is no acid corresponding to this oxide. RuO 4 is used for the preparation of pure ruthenium. Osmium. This metal is very analogous to ruthenium; it melts as high as 2500. The chlorides OsCl 2 and OsCl 4 and the oxides OsO, Os 2 O 3 and OsO 2 are known. The great similarity to ruthenium is especially noticeable in the highest oxides. Thus fusion with caustic potash and saltpetre pro- duces potassium osmiate, K 2 OsO 4 , which crystallizes from aqueous solu- tion in dark-violet octahedrons containing two molecules of water. The characteristic osmium compound is the tetroxide, OsO 4 , formed by igniting finely powdered osmium in the air or by the action of chlo- rine on the metal in the presence of water. The aqueous solution of Os0 4 reacts neutral, but is often (wrongly) called osmic acid. It is employed in microscopy since organic substances (i.e. reducing-agents) reduce it to black osmium. No salts derived from OsO 4 are known. This compound is used in preparing pure osmium. Rhodium. The metal in the fused state has the appearance of aluminium and is just as extensible (malleable and ductile) as silver. It is prepared pure in the arts by way of the chloro-purpureo rhodium chloride, Rh(NH 3 ) 5 Cl 3 (cf. 317). Neither acids nor aqua regia affect it. When heated in the air it is oxidized to the rhodious oxide, RhO. It is able to absorb a considerable amount of hydrogen. The rhodic oxide, Rh 2 O 3 , yields salts with acids. Of the chlorides only Rh 2 Cl e is known; this is obtained by direct synthesis as a reddish-brown substance; it forms sol- uble double salts with the alkali chlorides. The most satisfactory thermocouple for measuring high temperatures is ma'de of pure platinum and an alloy of platinum and rhoplium. 484 INORGANIC CHEMISTRY. [314- Iridium. This very refractory metal is obtained from iridosmine by heating in a current of oxygen, when the osmium volatilizes as tetroxide. In the form of a platinum alloy it is employed in the manufacture of "platinum" crucibles, dishes, distilling vessels for the concentration of sulphuric acid ( 86), etc. The prototype of the meter at Paris is made of an alloy of 90% platinum and 10% iridium. The admixture of iridium makes the platinum more indifferent to chemical agents, although at high tem- peratures the volatility of the iridium is often troublesome. When pure, iridium is not attacked by aqua regia. Iridium forms two chlorides, Ir 2 Cl 6 and IrCl 4 . Both of them give double salts with the alkali chlorides; e.g. Ir 2 Cl 6 6KC1 + 6H 2 O and IrCl 4 -2KCl. The former dissolves in water readily, the latter with difficulty. The tetrachloride appears as a black substance forming with water an intensely red solution. For this reason a platinum chlo- ride solution which contains iridium has a much deeper color than a pure solution. Palladium. 315. The silvery-white metal fuses at 1549, i.e. more easily than platinum. When finely divided it dissolves in boiling concentrated hydrochloric, sulphuric and nitric acids. On ignition in the air it is at first oxidized, thus losing its lustre, but at a higher temperature the metallic lustre reappears. The most peculiar characteristic of the metal is its ability to absorb hydrogen in large quantities (occlusion). Freshly ignited palladium foil absorbs 370 times its own volume of hydrogen at room temperature. By making palladium foil the cathode in a water electrolysis apparatus the metal can be made to take up even 960 times its own volume. This absorption does not alter its metallic appearance. The absorbed hydrogen can all be expelled by heating in a vacuum. Palladium charged with hydrogen is a strong reducing-agent ; chlo- rine and iodine are reduced by it (see 200) to hydrogen chloride and hydrogen iodide, respectively, and ferric salts are reduced to ferrous salts. Palladium forms two series of compounds, the -ous PdX 2 , and the -ic, PdX 4 . A characteristic compound of the first series is palladious iodide, PdI 2 , which is precipitated by potassium iodide from solutions of -ous salts as a black insoluble substance. This reaction is occasionally used to separate iodine from the other halogens, since their palladium compounds are readily soluble. Palladia chloride, PdCl 4 , is produced by dissolving the metal in aqua regia. With KC1 or NH 4 C1 it forms a difficultly soluble double chloride, K 2 PdCl 6 or (NH 4 ) 2 PdCl 6 . On the evaporation of its solution PdCl 4 dissociates into PdCl 2 . and C1 2 . 316.] PLATINUM METALS. 485 Platinum. 316. This metal is the principal constituent of the platinum ores. It fuses at about 1760 and is extremely malleable and ductile, hence it can be made into very fine wire and very thin foil. Since it becomes soft at red heat it can be easily worked. Platinum is used in the greatest variety of ways, among others the manufacture of utensils for the chemical laboratory, in the distillation of sulphuric acid, in electrical apparatus, in jewelry for settings of gems, in electric furnaces, in incandesecent lights and dentistry. The last named use consumes about one-third of the total production. When finely divided it absorbs oxygen (partially combining with it), a property to which is attributed the phenomenon that numerous oxidations proceed with unusual ease in the presence of platinum. Its use in gas-lighters depends on this property. When the metal is precipitated from its solutions by reducing-agents, it is frequently obtained as an extremely fine velvet-black powder, platinum black. When the double chloride (NH4) 2 PtCl 6 is ignited, the metal is left as a porous mass platinum sponge. At red heat a platinum partition allows hydrogen to pass through, while other gases are held back. This is due to the forma- tion of a compound or to the solubility cf hydrogen in platinum. Various substances attack platinum at elevated temperatures, e.g. the hydroxides, cyanides and sulphides of the alkalies; hence these substances should not be fused in platinum vessels. This also applies to lead and other heavy metals, for they form low-melting alloys with platinum. There are two sets of platinum compounds according to the gen- eral formula PtX 2 and PtX 4 . The best-known platinum compound is chlorplatinic acid, H 2 PtCl6, obtained by dissolving platinum in aqua regia. When the solution is evaporated the chlorplatinic acid is left in the form of large, reddish-brown, very hygroscopic prisms. Its aqueous solution contains the anion PtCle", for such an anion goes to the anode in an electrolysis; silver nitrate precipitates from the solution not silver chloride, which it would certainly do if free chlorine ions were present, but the compound Ag 2 PtCls. Two characteristic salts of this acid are those of potas- sium and ammonium; they are very difficultly soluble in water and insoluble in alcohol; when the aqueous solution is evaporated 486 INORGANIC CHEMISTRY. [316- the salt remains in the form of small, but well-formed octa- hedrons of a golden hue. The potassium salt is often made use of in determining potassium when sodium is also present, the sodium platinic chloride being very soluble, even in alcohol. Of the remaining platinum compounds a few may be referred to. If a solution of the above acid, H 2 PtCl6, is treated with sodium hydroxide and then with acetic acid, platinum hydroxide, Pt(OH) 4 , is precipitated. It is soluble in strong acids and also in alkalies, so that basic as well as acidic properties must be ascribed to it (platinic acid). Salts of this acid are moreover formed when platinum is fused with alkalies. Platinous chloride, PtCl2, is produced by heating chlorplatinic acid to 200 and in small amount, also, when the solution of this acid is strongly concentrated. It is a green powder, insoluble in water. With the alkali chlorides it gives soluble double salts, such as PtCl 2 -2NaCL Double cyanides of platinum with many metals are also known, e.g. K 2 Pt(CN)4* 3H 2 O, BaPt(CN) 4 -4H 2 O, etc. The latter has come into prominence because of its ability to make Rontgen rays visible. All these double salts are noted for their beautiful colors and strong dichroism. METAL-AMMONIA COMPOUNDS. WERNER'S EXTENSIONS OF THE NOTION OF VALENCE. 317. Several metals, notably those of the eighth group of the periodic system, are capable of forming complex compounds with ammonia and acid radicals. Such compounds have long been known, some having been pre- pared by the old master, BERZELIUS. The study of these substances occupied various investigators of the nineteenth century, especially JORGENSEN. In recent years, however, this field has been explored and greatly extended by the investigations of WERNER and his pupils, so that at the present time over 1700 compounds of the general type MX p (Am) q , are already known, M being a metal atom, X an acid radical and Am ammonia or an organic base (or even water). Chief credit is also due WERNER for having taken up the theoretical study of the relationships between these compounds and, as a result, generalizations of considerable importance for the structure of inorganic compounds, especially the complex salts, have been established. The whole subject deserves a little attention at this point. Concerning the methods of preparing these metal-ammonia compounds very little of a general nature can yet be stated. It is readily appreciated that the preparation of such a large number of complex compounds calls for the most diversified synthetical methods. 317.] METAL-AMMONIA COMPOUNDS. 487 In order to acquire an insight into the nature of the compounds concerned we may first examine the trinitrito triammine cobalt, Co(NH 3 ) 3 (NO 2 ) 3 , which can be obtained by mixing cold solutions of cobalt chloride, ammo- nium chloride and sodium nitrite and treating the mixture with a current of air. The compound then separates out as a difficultly soluble crystalline powder. It can be recrystallized from hot water containing a little acetic acid, without liberating nitrous acid, and is not attacked by dilute mineral acids in the cold. The electrical conductance of its aqueous solution is approximately zero. Evidently, therefore, the substance lacks the ordinary properties of a nitrite; the NO 2 -groups must be joined to the molecule differ- ently than in the nitrites. The ammonia, too, is otherwise combined than in the ammonium salts. This follows at once from the fact that the com- pound is a non-electrolyte: furthermore, from the fact that the action of even concentrated acids is insufficient to split off ammonia. The NH 3 -groups are not held by the acid radicals present in the molecule, for by energetic reactions it is possible to substitute other acid radicals for these without liberating the ammonia molecules. The supposition of an active participa- tion of the acid radicals in the linkage of the NH 3 -molecules is thus excluded. In deciding how the ammonias are connected to the molecule it is significant that the ammonia molecules can be replaced successively by other molecules. This shows that each ammonia molecule must be linked independently of the others in the complex molecule. About the only satisfactory explana- tion is that the NH 3 -molecules are attached directly to the metal atom. WERNER, however, makes the same assumption for the acid radicals in order to explain their abnormal behavior. This theory, whereby the special properties of the groups in compounds of the type MX p (NH 3 ) q are explained on the assumption that these groups are in direct combination with the metal, has proved to be of great importance for the classification of these compounds. The acceptance of this principle, however, necessitates an extension of the present notion of valence. Cobalt, for example, is at most trivalent in its salts and oxides; but, if we assume a direct linkage to the metal of the three NH 3 - and the three NO 2 -groups, the cobalt must have a valence of six. These new-appearing affinities of the metal atom cannot offhand be classed with the ordinary valence bonds. For this reason it seems appropriate to assign a special name to them. In order to distinguish them from the ordinary valences, which are termed "principal" or "primary" valences, they are called "subordinate," or "secondary," valences. Compounds of the type of the trinitrito triammine cobalt, that is, of the general formula MX 3 (NH 3 ) 3 have the property of combining with more ammonia molecules still. Thus there are compounds of the types MX 3 (NH 3 ) 4 , MX 3 (NH 3 ) 5 , and MX 3 (NH 3 )g. The addition of ammonia gives rise to a peculiar change in the function of the acid radicals, because for every addi- tional NH 3 -molecule that is taken on, one of the acid radicals enters the ionizable condition. If we compare the molecular conductivities in or -gfa normal solution at 250, to wit: 488 INORGANIC CHEMISTRY. [ 317- Co(NH 3 ) 6 X s Co(NH 3 ) 5 X 3 a 402 245 117 with those of the salts: MgCl, NaCl - a 370 249 125 ore find that the first of the above complex salts must be a quaternary, the second a ternary, the third a binary, electrolyte, and the fourth a non-elec- trolyte. The chemical properties of these compounds accord perfectly with this theory. In the compound, Co(NH 3 ) 4 Cl 3 , formerly called " praseo-cobalt chloride," only one chlorine can be directly precipitated by silver nitrate; in Co(NH 3 ) 5 Cl 3 two can be so precipitated, while in Co(NH 3 ) 6 Cl 3 , hexammine cobaltic chloride, all three chlorine atoms appear to be ionized in aqueous solution, since all three react at once with a silver salt solution. This is explained on the assumption that the added ammonia molecules displace the acid radicals from their immediate connection with the metal atom and themselves enter into this direct union with the metal. In order to distin- guish the ammonia molecules which are linked up in this way WERNER applies to them the term ammine and devises the following formulae and names to express the constitution of the above-named compounds: M(NH 3 ) 3 X 3 Tri-acido triammine compounds M(NH 3 ) 4 X 3 Di-acido tetrammine salts [M(NH 3 ) 4 X 2 ]X, M(NH 3 ) 6 X 3 Acido pentammine saltt [M(NH 3 ) 5 X]X 2 , M(NH 3 ) 6 X 3 Hexammine salts [M(NH 3 ) 6 ]X 3 , the atoms or groups within brackets being regarded as in direct union with the metal atom. The ionizable groups X outside of the brackets are in indirect union with the complex and, according to WERNER, are not joined to a definite elemental atom. If we compare the composition of the various metal-ammonias, we find that the formation of complexes is not without its limitations, but that after the addition of a definite number of NH 3 -molecules it comes to an end. Particularly sharp is the limit in respect to the number of groups which can unite directly with an atom serving as the center of a complex radical. It is a striking fact that this limit is the same for a good many elements and is commonly six (6). It seems most likely that this limiting number is characteristic of the elemental atom and of considerable importance. WER- NER calls it the coordination number and defines it as the maximum number of individual groups that can be directly united to an elemental atom. As already stated, the coordination number for most elements is 6; for a few elements, boron, carbon, and nitrogen, it is 4. 318.] METAL-AMMONIA COMPOUNDS. 489 318. The constitution deduced for the metal-ammonia compounds en- ables us to explain in a very simple way a series of isomerism phenomena that are characteristic of these compounds. Two isomeric compounds are known of the formula Pt(NH 3 ) 4 -SO 4 (OH) 2 . One behaves as a strong base, absorbing carbon dioxide from the air like caustic alkalies and precipitating metallic oxides from their salts, but gives no reaction of the sulphate ion in aqueous solution, e.g., no precipitate of barium sulphate with barium salts. The second compound is a perfectly neutral salt and acts as a normal sulphate, giving a barium sulphate precipitate at once with barium salts. We conclude from these reactions that the first compound gives only OH- ions ; the second, on the contrary, only SO 4 -ions. The cause of the isomerism is explained in the following coordination formulae; [o 2 S<^>Pt(NH 3 ) 4 ~J (OH), and [~HO/ Pt(NH3) <] S(V Sulphate tetrammine plato Dihydroxylo tetrammine hydroxide plate" sulphate The researches on metal-ammonia compounds have been of great assist- ance in clearing up the structure of many complex salts, since the latter can be prepared by the gradual replacement of NH 3 -groups by other mole- cules. An illuminating example of this is potassium cobaltic nitrite ( 310), whose relation to hexammine cobaltic chloride is shown by the following series of ammonia compounds: 1. [CoCNHP - 2. . Nitrito pentammine V cobaltic chloride [ (N0 2 ) s "| C (NH 3 ) 3 J Hoxammine cobaltic Nitrito pentammine Dinitrito tetrammine chloride V cobaltic chloride cobaltic chloride ro'ic - ro'ic (KB*!. CO (NH 3 ) K > Trinitrito triammine Potassium tetranitrito (unknown) cobalt diammine cobaltite 7. [Co(N0 2 ) 6 ]K 3 . Potass urn cobaltic nitrite In all these compounds there is no ionizable NO 2 -radical. The existence of such transition series clearly brings out the constitutional relationships which must exist between the metal-ammonias and the complex salts. The formulae for these are so constructed that the groups which do not respond to analytical tests, as well as the NH 3 -molecules displaced by them, are in direct union with the central metal atom. The progressive substitution of NH 3 by a negative group or element results in the gradual decrease of the valence of the cation. While this is 490 INORGANIC CHEMISTRY. [ 318. three in No. 1, it has become zero in No. 4, i.e., the molecule has become a non-conductor. If the substitution is carried further, as in Nos. 5, 6, and 7, the complex in brackets assumes the anion role and its negative charge 1234567 FIG. 77. increases from one to three. If the molecular conductivity for a given concentration is plotted on the ordinate axis in a diagram, we obtain for the series of compounds a curve, as is shown in Fig. 77. Numerous analo- gous transitions of metal-ammonia compounds to complex salts have been discovered, e.g.: [Pt(NH 3 ) 6 ]Cl 4 [PtCl 6 ]K 2 ; [Pt(NH 3 )JCl 2 [PtClJK 2 ; etc. Furthermore, relationships have been found to exist between metal- ammonias and hydrates. It is a significant fact that for a large number of salts of metals which give hexammine salts with ammonia, hydrates with six molecules of water predominate. This is the case with the salts of cobalt, chromium, nickel, etc. Moreover, JORGENSEN observed that com- pounds are derived from the hexammine salts by the exchange of one or two NH 3 -molecules for water, which compounds correspond in behavior to the hexammine salts. This follows from the fact that the molecular conductivity in aqueous solution is but little affected by this exchange, e.g.: [Co(NH 3 ) 6 ]Br 3 , a 402 390 380 It is also confirmed by the fact that with the removal of each water mole- cule an acid radical loses its -ionizing property, just as was the case for the | removal of each ammonia molecule from the hexammine salts: METAL-AMMONIA COMPOUNDS. 491 ,(H 2 0) " KUJ* Chloro pentammine H 3 0. Aquo pentammino Chloro pentammine chromic chloride chromic chloride The addition of water has the opposite effect. The trichloro triammine cobalt, for example, is known to form hydrates with 1, 2, and 3 mols. water. Since all of the three chlorine atoms in the trihydrate have an ionizing charac- ter, while in the diyhdrate only two have this property, in the monohydrate one and in the anhydrous compound none, we are justified in ascribing to them in the following constitution: r Cl2 i r cl i [Cl 3 Co(NH 3 ) 3 ], Co(H 2 0) Cl, Co(H 2 0) 2 C1 2 , and L (NH 3 ) 3 -I L (NH 3 ) 3 J A nearly complete transition series from a metal-ammonia to a hydrate (hydrous salt) is to be found in the case of chromium: r (H 2 o) i 3 ' Cr (NH 3 ) 5 J Cl3 ' Lacking Blue hexahydrate AH these compounds contain three chlorine atoms in ionizable linkage; accordingly silver nitrate at once precipitates all the chlorine as silver chloride, even from freshly prepared solutions of the salts. The function of water in these compounds corresponds perfectly to that of the water in the aquo-salts of the cobalt ammonias, i.e., with the release of each water mole- cule a chlorine atom sacrifices its ionizing ability. Two hexahydrates of chromic chloride, CrCl 3 -6H 2 O, are known, one blue and the other green. The blue one contains three ionizable Cl-atoms, and according to its general behavior should be regarded as [Cr(OH 2 ) 6 ]Cl 3 , hexaquo chromic chloride. Upon the loss of two water molecules it goes over into dichloro tetraquo chromic chloride [Cl 2 Cr(OH 2 )4]Cl, which contains but one ionizable Cl-atom. If this dichloro tetraquo chloride is crystallized out of water, it adds on two molecules of the water and goes over into the green hydrate, [Cl 2 Cr(OH 2 )JCl + 2H 2 O, which is thus isomeric with the blue hydrate. The green hydrate, moreover, contains only one chlorine atom that can be pre- cipitated directly with silver nitrate. It should also be mentioned that the green hydrate can be transformed into the blue one and the water mole- cules, that condition the isomerism, shifted. Since the isomerism must 492 INORGANIC CHEMISTRY. [318, depend on the difference in the manner of attachment of the water, WERNER has called it hydrate isomerism. Various cases of isomerism that are met with in the metal-ammonia compounds are best explained as cases of stereo isomerism. Two series of compounds are known of the general formula, Pt v 3 2 J tnev are called platin diammine compounds and have been known since 1870. Later in- --r B FIG. 78. FIG. 79. vestigations have shown that this kina of isomerism occurs only with com- pounds of the type M 2 . The assumption can therefore be made that the X5 4 six groups are arranged at the apexes of an octahedron around the central atom. Only with the type above described are two isomers possible, as ^ the accompanying figures (Figs. 78 and 79) show. Compounds of the M * r> type cannot, therefore, have isomers and, as a matter of fact, no one has yet been able to prepare such isomers. WERNER has also discovered metal-ammonias of more than one nucleus, i.e., where the molecules contain several metal atoms joined to one another in such a way as to undergo no separation either by ionization or by spon- taneous hydrolysis; but we cannot dwell longer upon the subject her*. INDEX. The principal references are in italics. Abraum salts of Stassfurt, 65, 322, 374, 441 Absolute weight of atoms, 49 Absorption, 70 Absorption spectra, 391 Accumulator, 421 Acetylene, 249 Acidimetry, 350 Acids, 39, 99 strength of, 99 Acid chlorides, 198 Acids, dibasic, 131 Acids, monobasic, 83 Acid salts, 131 Acid sodium carbonate, 322 Actinium, 399 Additive properties, 348 Affinity, 156 Agglutinants, 269 Air, 165 Alabaster, 376 Alchemists, 371 Aldebaranium, 446 Algaroth powder, 234 Alkalimetry, 350 Allotropism, 53, 104, 148, 201, 222, 241, 261, 274 Alloys, 232, 237, 276, 355, 361, 409, 437, 471, 479, 482, 483, 484 Aludels, 410 Alumina, 437 Aluminates, 438 Aluminium, 436 amalgam, 437 bronze, 437 chloride, 438 hydroxide,' 438 oxide, 437 silicate, 440 sulphate, 439 Alums, 43& Alum stone, 439 Alunite, 439 Amalgams, 412 Amalgam, aluminium, 437 Amalgam, ammonium, 175 j sodium, 313 tin, 276 Amblygonite, 310 Amethyst, 265 Amides, .198 Amidophosphoric acid, 221 Amido-sulphonic acid, 198 Ammine, 488 Ammonia, 174, 311 compounds of the metals of the eighth group, 486 -soda process, 320 Ammonium, 175, 331 amalgam, 175 carbonate, 334 chloride, 333 hydroxide, 176 magnesium phosphate, 216 metavanadate, 448 nitrate, 334, 379 phosphate, 334 salts in analysis, 375 sulphate, 334 sulphide, 335 Amorphous substance, 385 Amphoteric compounds, 278, 284 Amphoteric reaction, 319 Analysis, hydrogen sulphide in, 117 Analysis, volumetric. See Volumet- ric analysis. Anatase, 446 ANAXAJGORAS, 28 Anglesite, 286 Anhydride, acid, 83 mixeti, 186, 195 Anhydfrite, 381 Anions; 96 Anode, 96 Antichlor, 132, 317 Antimonic acid, 236 Antimony, 231 butter, 233 pentachloride, 234 pentasulphide, 236 493 494 INDEX. Antimony, pentoxide, 236 tetroxide, 236 trichloride, 233 trioxide, 234 trisulphide, 236 Antimonyl, 235 Antiseptic, 416 Anglesite, 286 Angstrom, 392 Apatite, 199, 376 Aqua regia, 198 Aragonite, 376, 382 Argentite, 359 Argon, 170, 300, 309 Argyrodite, 286 ARISTOTLE, 371 ARRHENIUS, 95, 101 Arsenic, 221 acid, 229 meal, 226 oxide, 228 sulphides, 229, 268 sulpho-salts, 230 trichloride, 226 white, 224, 226 Arseni-tungstates, 457 Arsenious acid, 228 oxide, 226 Arsenolite, 221 Arsenopyrite, 221 Arsine, 223 Asbestos, 374 Association, 82, 185 Atmosphefe, 165 Atomic heat, 289 theory, 27 volume, 297 weights, 29 v Determination of, 44, 47, 49, 288, 305 Table, of, Inside back cover Atoms, 28, 49 Augite, 374 Auric compounds, 370 Aurous compounds, 369 Aurum musivum, 281 AVOGADRO'S hypothesis, 44, 60, 293 Azides, 179 Azurite, 353 Bacteria, 194 BAEYER, 144 BAKER, 185, 333, 414 BALARD, 65, 84 BALMER, 392, 393 Barite, 387 Barium, 387 Barium peroxide, 53, 387 Baryta-water, 387 Bases, 39, 102 BASILIUS VALENTINUS, 231 BAUME degrees, 142 Bauxite, 436, 438 BECKER, 129 BECQUEREL, 397 BeU metal, 276 BERGMANN, 99 BERNTHSEN, 133 BERTHELOT, 50, 152, 157, 182, 255 BERTHELOT'S principle, 157 BERTHELOT-MAHLER calorimetric bomb, 153 Beryl, 372 BeryUium, 303, 372 BERZELIUS, 23, 29, 30, 148, 280, 293, 308, 418, 486 BESSEMER process, 469 Bimolecular reactions, 76, 206, 255 BlRKELAND, 191 Bismuth, 236 chloride, 237 glance, 237 hydroxides, 238 oxides, 238 oxychloride, 237 subnitrate, 239 sulphate, 239 sulphides, 239 Black ash, 319 Blanc fixe, 388 Blast furnace, 463 Bleaching, 36, 56, 126 Bleaching-powder, 85, 380^ Blende, 407 ^^^ Blooms, 471 BODENSTEIN, 15 Boiling-point, elevation of, 61, 272 BOLTWOOD, 404 BONE, 249 Boneblack, 244 Boracite, 432 Borax, 432, 435 Boric acid, 433 Boron, 432 BOUSSINGAULT, 166 BOYLE, 371, 395 BOYLE'S law, 47, 58, 160, 295 BRAND, 199 Brass, 355 Bricks, 440 Brimstone, 103 Bromine, 65 oxygen compounds of, 90 Bronze, phosphor, 202, 276 silicon, 276 INDEX. 495 Bronzes, 276 Brookite, 446 Brownian movement, 49 BROWNING, J., spectroscope, 389 BRUHL, 55 BUNSEN, 329, 389, 391, 398 burner, 258 ceU, 422 Burette, 147 Cadmium, 410 Caesium, 329 CAIN, 249 Calamine, 407 Calcining, 353, 463 Calcite, 376, 382 Calcium, 376 carbide, 171, 249 carbonate, 382 chloride, 321, 340, 379 cyanimide, 175 fluoride, 380 hydroxide, 378 hypochlorite, 380 manganite, 460 nitrate, 382 oxide, 377 peroxide, 379 phosphates, 382 plumbate, 285 silicate, 384 sulphate, 338, 381 sulphide, 291, 319, 380 Calomel, 413 Caloric effect, 152, 336, 357, 437 Calorimeter, 153 Calx, 165 Canal rays, 400 Carat, 368 Carbides, 247, 249, 457 Carbon, 241 amorphous, 244 dioxide, 252 monoxide, 249 oxysulphide, 251 tetrafluoride, 247 Carbonado, 241 Carbonic acid, 253 Carborundum, 247 Carburetting, 250 Carnallite, 322, 329, 374, 441 CARO'S acid, 144 Cassiopeium, 446 Cassiterite, 274 Cast iron, 466 CASTNER, process, 314 Catalysis, 34, 55, 73, 129, 134, 137, 181, 194, 233, 334, 380 Cations, 96 Cats-eye, 455 Caustic potash, 323 soda, 313 CAVENDISH, 17, 165 Celestite, 386 CELSIUS, 20 Cement, 378 Cementite, 466 Ceramics, 440 Cerargyrite, 359 Cerite, 443 Cerium, 445 Cerussite, 281 Chalcocite, 353 Chalcopyrite, 103, 353 Chalk, 376 Chamseleon minerale, 461 Chamber acid, 136, 319 crystals, 195 Charcoal, 245 CHARPY, 466 Chemical affinity, 36, 156 operations, 5 Chemistry, field of, 3 Chili saltpetre, 69, 189, 311, 318, 326 Chinese white, 409 Chloramine, 177 Chlor-auric acid, 370 Chlorazide, 179 Chloric acid, 88 Chloride of lime, 380 Chlorination process (gold), 368 Chlorine, 33 hydrate, 37 oxygen compounds of, 83 Chlorous acid, 88 Chlorosulphonic acid, 142 Chlorplatinic acid, 485 Chromates, 452 Chrome alum, 452 green, 450 yellow, 454 Chrom-sulphuric acids, 451 Chromic anhydride, 452 chloride, 451 oxide, 450 sulphate, 451 Chromite, 449, 450 Chromium, 449 in steel, 471 oxychloride, 455 Chromous compounds, 450' Chromyl chloride, 455 Chryaoberyl, 372 Cinnabar, 103, 410 CLARKE, 386 Clay, 436, 440 496 INDEX. Cleveite, 172 Coal, 245 Coarse metal, 354 Cobalt, 477 compounds (ammonia), 487 Cobaltite, 221, 477 COEHN, 129 COHEN, 425 Coke, 245 Colcothar, 474 Cold-hot tube of DEVILLE, 128 Colemanite, 432, 435 Collargol, 269 Colloids, 266, 268, 365, 438, 474 Columbium, 449 Combustion, 11, 165 Components, 110 Compound, 25 Condenser, 6 Conductivity, molecular, 97 Conservation of matter, 16 energy, 153 Constancy of natural phenomena, 3 Contact action. See Catalysis. process (sulphuric acid), 137 Converter, 469 Coordination number, 488 Copper, 353 ammonia compounds, 359 compounds, 355 (See "Cupric" and "Cuprous.") Copperas, 473 Corrosive sublimate, 414, 415 Corundum, 436 Counter-current principle, 19 COURTOIS, 69 Covering power, 287 CREDNER, 168 Crocoite, 281, 449 CROOKES, 442 Crucible process (steel), 470 Cryohydric point, 339, 341, 379 Cryohydrate, 339, 343 Cryolite, 311, 436 Cryoscopic method, 65 Crystallization, 6 Crystalloids, 268 Crystals, mixed, 386, 468 Cup and cone apparatus, 463 Cupellation, 360, 368, 369 Cupric arsenite, 359 bromide, 358 carbonate, 359 chloride, 358 hydroxide, 358 iodide, 358 nitrate, 359 oxide, 358 Cupric sulphate, 358 sulphide, 359 Cuprite, 353 Cuprous chloride, 356 cyanide, 357 iodide, 356 oxide, 355 sulphate, 357 CURIE, 397, 398 CURTIUS, 178 Cyanide process, 360, 368 Cyanogen, 255 DALTON, 27, 43, 293, 397 DANIELL cell, 420 DAVY, 241, 258, 311, 322, 418 DEACON process (chlorine), 34 DEBIERNE, 398 Decantation, 5 Decay of elements, 402, 404 Definite proportions, law of, 27 .Degree of freedom, 111 Deliquescence, 318 DEMOCRITES, 28 Dephlogisticated air, 166 Desiccators, 142 Detonating-gas, 15, 22 DEVILLE, STE. CLAIRE, 73, 127 DEWAR, 14, 172 Dialysis, 266, 269 Diamide, 177 Diamond, 241 Didymium, 445 Dilatometer, 108 Disinfection, 36, 416, 473 - Pisperse, phase, etc., 272 Dissociation, 70, 72, 94, 121, 124, 140, 148, 185, 193, 197, 210, 227, 233, 259, 277, 286, 333, 355, 370, 383, 414, 474 electrolytic, 94, 131, 214, 219, 254, 268, 313, 332, 346, 357, 363, 375, 409, 415, 417, 418, 472, 475, 478, 489. hydrolytic, 100, 207, 269, 277, 279, 352, 362, 376, 414, 433, 435, 439, 442, 447, 474 tension, 430 Distillation, ,19 fractional, 6 Distribution law of BERTHELOT, 325. 332, 360 Dithionic acid, 145 Divariant system, 114 DIXON, 255 DOEBEREINER, 212 Dolomite, 374 Double decomposition, 41 INDEX. 497 Double salts, 278, 286, 360, 440, 486 DUBOIS, 168 DULONG and PETIT, law of, 289, 407 DUMAS, 23, 24, 166 Dysprosium, 443 Earthenware, 440 Earth's crust, composition of, 8 Eau de Javelle, 85 Ebullioscopic method, 65 Efflorescence, 318, 321 EINSTEIN, 272 Eka-aluminium, 441 -boron, 445 -silicon, 273, 307 Electrical endosmose, 271 Electric furnace, 243, 249, 262, 265, 377, 457, 465, 470 Electrochemical series, 428 Electrochemistry, 418 Electrolysis, 23, 41, 79, 133, 143, 177, 310, 311, 313, 322, 325, 354, 360, 361, 362, 368, 386, 387, 430, 436,' 447, 454, 457, 485 Electrolytic separation of the metals, 430 Electromotive force, 159, 419, 426 Electrons, 396 Electrotyping, 355 Elements, 7, 296, 371 Emanation, 401 Emerald, 372 Emery, 436 EMPEDOCLES, 28, 156 Emulsion, 272 Endothermic compounds, 153, 161, 179, 180, 182, 184, 187, 223, 249 Energy, free, bound, and total, 157, 158 Epsom salt, 376 Equations, 31 Equilibrium, 73, 86, 98, 118, 128, 138, 157, 159, 160, 177, 186, 216, 226, 250, 259, 325, 334, 337, 355, 356, 375, 379, 383, 418, 465, 467, 472 heterogenous, 109, 383 state of, 77 Equivalent weights, 293, 305 Erbia, 445 ERDMANN, 170 Eudiometric analysis of air, 166 Europium, 443 Eutectic point, 339, 343, 468 Euxenite, 443 Exothermic reactions, 153, 155, 158, 161, 357 Explosion wave, 255 EYDE, 191 FARADAY, law of, 159, 397 Feldspar, 436 Ferric compounds, 474 Ferrite, 466 Ferrocyanic acid, 476 Ferrous compounds, 473 Feuerluft, 166 Filtration, 5 Fire damp, 248 Flame, 256 spectrum, 390 Flash light, 374 Flowers of sulphur, 103 Fluorine, 79 / Fluor spar, 79, 81, 380 Flux, 353, 463 Formula, 30 empirical, 57 Fractional distillation, 6 N FRASCH sulphur process, 103 FRAUNHOFER lines, 394 Freezing mixture, 379 -point, depression of, 61, 94, 105, 121, 272, 339 Freezing-point curves. See Melting- point curves. FR:MY, 308 Fuels, 245 Fumaroles, 433 Furnace, electric. See Electric Fur- nace. reverberatory, 360 , Gadolinite, 443 Galenite, 103, 281 Gallium, 441 Galvanic cells, 418 ^ Galvanized iron, 409, 424) Garnierite, 479 Gas carbon, 245 GATTERMANN, 71 GAY-LUSSAC, 69, 87, 192, 311) law of, 43, 47, 58 tower, 136, 196 GEBER, 198 Germanium, 273, 307 German silver, 355, 479 ^ Gersdorffite, 479 GIBBS, 109 Glacial phosphoric acid, 217 Glass, 384 etching, 82, 263 GLAUBER, 156 GLAUBER'S salt. See Sodium suit phate. GLOVER tower, 135, 196 Glucinum, 372 Gold, 269, 367 498 INDEX. Gold compounds. See Aurous com- pounds and Auric compounds. Gold-plating, 368 GOLDSCHMIDT, reduction method, 437 449, 456, 458 GOLDSTEIN, 400 Graduation salt process, 315 GRAHAM, 266, 268 Gram equivalent, 146 molecule, 32 Graphite, 244 Graphitic acid, 244 GRIESHEIM process, 313, 325 GROVE'S gas battery, 425 GUIGNET'S green, 450 Gun-metal, 276 Gunpowder, 26, 327 GUTZEIT'S test (arsenic), 222, 224 GUYE, 294 GYPSUM, 103, 376, 381 Halite, 311 HALL process (aluminium), 436 Hammer scale, 12 HAMPSON, 10, 170 Hardness of water, 383 HARGREAVE'S method, 317 Hartshorn, salt of, 334 Hausmannite, 458 Heat, atomic, 290 molecular, 173, 291 of dilution, 153 formation, 152, 155 neutralization, 153, 350 solution, 153, 161, 336 specific, 290 Heavy spar, 387 Helium, 172, 300, 394, 400, 402 HELMHOLTZ, 158 Hematite, 463 HENRY'S law, 11, 39, 95, 133, 332, 434 Hepar sulphuris, 328 HEROULT, 436 furnace, 470 HESS, law of, 153 Hexammine cobalt salts, 489 Hittorf, 202 HOFMANN, 24 Holmium, 443 Hopper crystal of salt, 317 Hornblende, 374 Horn silver, 359, 363 Hiibnerite, 456 Hydrate isomerism, 192 Hydrates, 490 Hydraulic mining, 368 Hydrazine, 177 Hydrazoic acid, 178 Hydriodic acid, 71 Hydrobromic acid, 67 Hydrocarbons, 248 Hydrochloric acid, 37 composition of, 41 Hydrofluoboric acid, 433 Hydrofluoric acid, 81 Hydrofluosilicic acid, 264 Hydrogel, 270, 358, 438, 450, 474 Hydrogen, 13 antimonide, 232 arsenide, 223 bromide, 67 chloride, 37 cyanide, 256 disulphide, 120 fluoride, 81 iodide, 71 peroxide, 53, 64, 367, 455 persulphide, 120 phosphide, 204 selenide, 149 silicide, 262 sulphide, 115 telluride, 150 trisulphide, 121 Hydrolysis. See Dissociation, hydro- lytio. Hydrosol, 270 Hydroxyl, 143 Hydroxylamine, 180 -disulphonic acid, 198 Hypo, 132, 317 Hypochlorous acid, 84 oxide, 84 Hyponitrous acid, 187 Hypophosphoric acid, 218 Hypophosphorous acid, 220 Hyposulphurous acid, 133 Hypothesis, 2 Ice-machine, 175 Ice stone, 311, 436 Illuminating gas, 476 Incandescent electric light, 456 as light (WELSBACH), 443 elible ink, 366 Indicators, 350, 352 Indium, 305, 441 Induced radio-activity, 401 Inversion, point, of, 106 Iodine, 69 chlorides, 83 oxygen compounds of, 91 lodometry, 146, 226, 228 Ionic equilibrium, 98 equation, 118, 132, 277, 282, 313, 356, 408, 422, 453, 461, 475 INDEX. 499 Ionic equilibrium, theory of, 94, 118 Ionium, 405 lonization. See Dissociation, elec- trolytic. Ions, 95, 346, 356, 363, 408, 409, 439, 451, 453, 459, 485, 487 valence of, 123 Iridium, 484 Iridosmine, 482, 484 Iron, 463 and carbon monoxide, 477 Isomerism, 492 Isomorphism, 292 ic solutions, 62 Jasper, 265 JORGENSEN, 486, 490 Kainite, 322, 374 Kaolin, 436, 440 KASSNER, process, 285 KAYSER, 392, 393 Kelp, 69 Kieserite, 374 Kinetic theory, 47, 249 KIPP generator, 115 KIRCHOFF, 329, 389, 391, 398 KNIETSCH, 137 KOHLRAUSCH, 19 KOPP, 291 KRYPTON, 172, 300 KUHNE method, 261 "Labile," 108 LADENBURG, 53 Lampblack, 245 Lanthanum, 445 Lapis lazuli, 440 LAVOISIER, 10, 13, 18, 165, 241 Law of AVOGADRO, 44, 47, 60, 293 BOYLE. See BOYLE'S law. the constancy of natural phe- nomena, 3 onstant composition (definite proportions), 27 distribution (BERTHELOT), 325 332 DULONG and PETIT, 290, 407 GAY-LUSSAC, 43, 47, 58 HENRY. See HENRY'S law. HESS, 153 MlTSCHERLICH, 292 multiple proportions, 29 chemical mass action, 75 NEUMANN, 291 octaves, 296 thermoneutrality, 350, 415 Lead, 281 carbonate, 286 chamber crystals, 195 chambers, 135 chloride, 285 chromate, 454 glass, 384 nitrate, 286 oxides, 283 peroxide, 294, 422 persulphate, 286 sulphate, 286 sulphide, 287 tree, 282 white, 286 LE BLANC soda process, 142, 319, 327 LE CHATELIER'S rule, 160. 186, 250, 338 LECLANCHE cell, 423 LECOQ DE BOISBAUDRAN, 441 Lepidolite, 310, 329 LEUCIPPUS, 28 LEYDENFROST, phenomenon, 170 Lime, 377 -sulphur solution, 381 Limestone, 241, 376 LINDE, 10, 170 Liquation, 274 Liquefaction of oxygen, 10 Litharge, 283 Lithia mica, 310, 329 Lithium, 310 LOBRY DE BRUYN, 178, 272 LOCKYER, 172 Lunar caustic, 366 LUPKE cell, 423 Lutecium, 443 Magnalium, 437 Magnesia, 374 alba, 376 mixture, 229 usta, 374 Magnesite, 374 Magnesium, 374 ammonium phosphate, 376 boride, 432 carbonate, 376 chloride, 375 hydroxide, 374 nitride, 174, 374 oxide, 374 sulphate, 376 Magnetite, 463, 474 Malachite, 353 Malleable iron, 466 MANCHOT, 55 500 INDEX. Manganese, 458 dioxide, 33, 460 in steel, 471 Manganic compounds, 459 acid, 460 Manganous compounds, 459 Marble, 376 MARCKWALD, 399 Marl, 376 Marsh gas, 248 MARSH test, 225, 233 Martensite, 466 Mass action law, 75. See also Equi- librium. Massicot, 283 MASSON, 299 Matches, 203 Matte, 354 Matter, 3 unity of, 308, 395 Meerschaum, 374 Mellitic acid, 246 Melting-point curve, 122, 342, 412, 467 MENDEL::EFF, 296, 299, 303, 307, 441, 445 MENDEL EEFF'S table (periodic sys- tem), 301 MENSCHING, 232 Mercuric ammonium chloride, 414 chloride, 414 cyanide, 268, 414 halides, alkali, 415 iodide, 415 nitrate, 416 oxide, 9, 408 sulphate, 416 sulphide, 417 Mercurous compounds, 413 Mercury, 410 Mesothorium, 448 Metal-ammonia compounds, 486 Metalloids, 8 Metallurgy of iron, 463 Metaphosphoric acid, 217 Metaphosphorous acid, 219 Metastable system, 108, 346 Metastannic acid, 280 Methane, 248 Methyl orange, 353 MEYER, LOTHAR, 41, 296, 300, 308 MEYER, VICTOR, 232 Mica, 436 MICHELSON, 392 Microcosmic salt, 334 Mineralization of organic matter, 168 Miner's safety lamp, 258 Minium, 284 Mispickel, 221 MITSCHERLICH'S law, 292 test, 203 Mixture, 26, 169 Mobile equilibrium, VAN'T HOFF'S principle of, 160, 336 MOHR'S salt, 473 MOISSAN, 16, 38, 67, 79, 133, 241, 244, 247, 446, 449 Mole, 32 Molecular depression, 65 elevation, 65 heat, 173, 291 weight, 45 determination of, 47 by boiling-point method, 61. 227 ... . by freezing-point method, 65, 98, 131, 208, 312 Molecule, 28, 44, 49 Molybdenite, 455 Molybdenum, 455 trioxide, 455 Molybdic acid, 456 Monazite sand, 443 Monobasic acid, 83 MORLEY, 29, 212, 294 Mortar, 378 Mosaic gold, 281 Muriatic acid, 38 MUTHMANN, 456 Nascent state, 37, 54 Natural gas, 248 Negative (photography), 365 Neodymium, 445 Neon, 172, 300 NERNST, 104, 184, 282, 418, 425 glower, 51, 54 NESSLER'S solution, 415 NEUMANN'S law, 291 NEWLANDS, 296 NEWTON'S metal, 237 Niccolite, 479 Nickel, 479 steel, 471 Niobium, 449 Niton, 402 Nitramide, 199 Nitric acid, 189 oxide, 183 Nitrides, 164 Nitrilosulphonates, 198 Nitrogen, 162, 170 acid derivatives of, 195 dioxide, 185 halogen compounds of, 179 hydrogen compounds of, 174 INDEX. 505 Touchneedles, 369 Touchstone, 369 Tourmaline, 374 TOWNSEND process, 314 Transition point, 106, 160, 227, 275, 317, 334, 340, 415, 425 TRAUBE, 54 TRAVERS, 172 Triads, 296 Triboluminescence, 227 Tridymite, 265 Trinitrito triammine cobalt, 487 Triple point, 113 Triphylite, 310 Tuffstone, 378 Tungsten, 456 steel, 456, 471 Twyers, 463 TYNDALL, effect, 271 Type metal, 232 Ultramarine, 440 Ultramicroscope, 272 Ultraviolet light, 18, 37, 129, 266 Unimolecular reactions, 76, 251, 402 Unity of matter, 395 Univarient system, 112 Uraninite, 397, 457 Uranium, 177, 397 and foil., 457 URBAIN, 444 Vacuum flask, 11, 170 Valence, 122, 185, 486 Valence, maximum, 123 of ions, 123 Vanadinite, 448 Vanadium, 448 VAN DER STADT, 206, 219 VAN DER WALLS, 49, 109 VAN MARUM, 49, 203 VAN'T HOFF, 58, 60, 157, 381 VAN'T HOFF'S principle of mobile equilibrium, 160, 336, 337 Vapor pressure curve, 61, 106, 112 Varec, 69 Vein mining, 367 Velocity constant, 75 of reaction, 15, 75, 182, 233. See also_Catalysis. Vermilion, 417 Vitriol, blue, 358 green, 473 oil of, 137 Vivianite, 199 VOGEL'S spectroscope, 389 Volumetric analysis, 146, 189, 350, 460, 462 Vulcanizing rubber, 121 WACKENRODER'S liquid, 145 Washing (precipitates), 5 Washing-soda, 321 Water, 17 composition of, 22 gas, 250 glass, 266, 322, 328' natural, 20, 383, 473 physical properties of, 20 purification of, 21, 283 Wavellite, 199 Welding, 437 WELDON process, 460 WELSBACH, AUER VON, 445, 446 incandescent gas light, 250, 257, 388, 443, 447 WERNER, 486 WHITNEY, 347 WINKLER, 273, 307 Witherite, 387 Wolframite, 456 WOOD'S metal, 237 WOULFF bottle, 18 Wrought iron, 466 Wulfenite, 281, 455 Xenon, 172, 300 Ytterbium, 445 Yttrium, 443 Zinc, 407 compounds, 409 dust, 407 white, 409 Zircon, 446 Zirconia, 447 Zirconium, 446^ ZSIGMONDY, 272 THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $1.OO ON THE SEVENTH DAY OVERDUE. IP MtlMt '"; 24 1936 SEP 23 1936 ff% " '/ APR 38 1937 '-'607 (fast UNIVERSITY OF CALIFORNIA LIBRARY