1911 Encyclopædia Britannica/Silver

From Wikisource
Jump to navigation Jump to search

SILVER (symbol Ag, from the Latin argentum, atomic weight 107·88 (O=16)), a metallic chemical element, known from the earliest times and of great importance as a “noble” metal for articles of value—coinage, ornamentation and jewelry. Etymologically the word “silver” probably refers to the shining appearance or brightness of the metal. The Latin argentum is cognate with the Greek ixpyvpos, silver, which in turn is derived from apyos, shining. The Hebrew Keseph is connected with a root meaning “to be pale.” The alchemists named it Luna or Diana, and denoted it by the crescent moon; the first name has survived in lunar caustic, silver nitrate. Silver is widely diffused throughout nature, occurring in minute amount in sea-water, and in the mineral kingdom as the free metal, as an amalgam with mercury and as alloys with gold, platinum, copper and other metals. Native silver is occasionally met with in metalliferous veins, where it has been formed by the alteration of silver-bearing minerals. It crystallizes in the cubic system, but the crystals are usually distorted and indistinctly developed: twisted wire-like forms are much more common. The best crystallized specimens have been obtained from Kongsberg in Norway, large masses, weighing as much as 5 cwts., having been found. It is also found in other silver mines, especially those of Mexico, Peru and Chile; in the Lake Superior copper mining region it occurs in association with native copper. The element is a constituent of many mineral sulphides, some of which are of sufficiently frequent occurrence to rank as ores of silver. Of these the more important are noticed under Metallurgy; here we may notice the rarer minerals. Silver sulphide, Ag2S, occurs naturally as the orthorhombic acanthite, and the cubic argentite; the telluride, Ag2Te, named hessite, assumes cubic forms; other tellurides containing silver are petzite, (Ag,Au)2Te, and sylvanite, AuAgTe4. In association with antimonious and arsenious sulphides, silver sulphide forms many important minerals, which sometimes present dimorphous forms, reflecting the dimorphism of silver sulphide; moreover, the corresponding arsenious and antimonious compounds are frequently isomorphous. This is illustrated by the hexagonal pyrargyrite 3Ag2S·Sb2S3, and proustite, 3Ag2S·As2S3, and the monoclinic pyrostilpnite, isomeric with pyrargyrite, and xanthoconite, isomeric with proustite. Other pairs of isomorphous argentiferous minerals are: the cubic polybasite, 9Ag2S·Sb2S3, and pearceite, 9Ag2S·As2S3; and the germanium minerals argyrodite, 4Ag2S·GeS2, and canfieldite, 4Ag2S·(Sn,Ge)S2.

Physical Properties.—In appearance silver presents a pure white colour with a perfect metallic lustre. It is the most malleable and ductile of all metals with the exception of gold: one gramme can be drawn out into a wire 180 metres long, and the leaf can be beaten out to a thickness of 0·00025 mm.; traces of arsenic, antimony, bismuth and lead, however, make it brittle. In hardness it is superior to gold, but inferior to copper. Its specific gravity, according to G. Rose, lies between 10·514 and 10·619 at 14°; an average value is 10·57. Its specific heat is 0·05701 (Regnault) or 0·0559 (Bunsen); its coefficient of linear expansion is 0·00001921. Its thermal conductivity is, according to Wiedemann and Franz, superior to that of other metals, being in the ratio of 100 : 74 as compared with copper and 100 : 54 with gold; it is the most perfect conductor of electricity, standing to copper in the ratio 100:75, and to gold 100:73. Silver melts at about 1000°C; recent determinations give 960·7° (Heycock and Neville) and 962 (Becquerel); at higher temperatures it volatilizes with the formation of a pale blue vapour (Stas). Its vapour density has been determined at 2000°, and corresponds to a monatomic molecule. When molten, silver occludes the oxygen of the atmosphere, absorbing 20 times its own volume of the gas; the oxygen, however, is not permanently retained, for on cooling it is expelled with great violence; this phenomenon is known as the “spitting” of silver. It is prevented by preserving the molten metal from contact with air by covering the surface with non-oxidizing agents, or by traces of copper, bismuth or zinc.

Chemical Properties.—Silver is not oxidized by oxygen, but resembles mercury in being oxidized by ozone. It has no action on water. It is readily soluble in dilute nitric acid, nitric oxide and silver nitrate being formed; it also dissolves in hot, strong sulphuric acid, sulphur dioxide being evolved. Hydrochloric acid forms a surface film of silver chloride; hydriodic acid readily dissolves it, while hydrofluoric acid is without action. Sulphuretted hydrogen is decomposed with the formation of a black coating of silver sulphide; this is the explanation of the black tarnish seen when silver is exposed to the fumes of coal gas, and other sulphuretted compounds, such as occur in eggs. The so-called “oxidized” silver is a copper-silver alloy coated superficially with a layer of the sulphides by immersion in sodium sulphide or otherwise. Silver combines with the free halogens on heating and also with sulphur.

Molecular silver is a grey powder obtained by leaving metallic zinc in contact with silver chloride which has been precipitated in the cold and washed till nearly free from acid. The powder is separated from the zinc, washed with hydrochloric acid, dried in the air, and then gently heated to 150°. It assumes a metallic lustre on burnishing or heating to redness. It receives application in synthetic organic chemistry by virtue of its power to remove the halogen atoms from alkyl haloids, and so effect the combination of the two alkyl residues.

Colloidal silver is the name given by Carey Lea to the precipitates obtained by adding reducing solutions, such as ferrous sulphate, tartrates, citrates, tannin, &c., or to silver solutions. They dissolve in water to form solutions, which do not penetrate parchment membranes, hence the name colloidal. Many other methods of preparing these substances are known. Bredig’s process consists in passing an electric arc between silver electrodes under water, when a brown solution is obtained.

Production.—The economic questions which attend the production of silver and the influence which gold and silver exercise on prices are treated in the articles Money and Bimetallism; the reader is referred to the former article for the history of silver production and to the topographical headings for the production of specific countries. Since the middle of the 19th century the annual production has increased: the following table gives the average annual production in 1000 oz. over certain periods:—

1841–1850. 1851–1860. 1861–1865. 1866–1870. 1871–1875. 1876–1880. 1881–1885. 1886–1890. 1891–1895.
 25,090  28,792  35,402  43,052  63,318  78,777  87,272 110,356 158,942
1900. 1901. 1902. 1903. 1904. 1905. 1906. 1907. 1908.
180,093 174.851 164,560 170,128 182,262 189,830 165,640 184,894 203,186

Over two-thirds of the world’s supply is derived from Mexico and the United States. The Mexican mines first sent supplies to Europe in the 16th century, and during the period 1781–1800 yielded two-thirds of the world’s production. Although the production has decreased relatively, yet it has increased enormously absolutely; in 1900, it was 55,804,420 oz., being second to the United States; in 1905 it was 73,838,066 oz., establishing a record for any single country. The United States came into prominence in about 1860, and the discovery of the famous Comstock lode in Nevada led to an enormous increase in the production. The production of this lode declined in 1876, but the total production of this country was increased by discoveries in Colorado (Leadville) and Nevada (Eureka); and in more recent years silver-producing areas in other states (Montana, Utah, Idaho) have been exploited. In 1860 the production was 116,019 oz., which increased to 1,546,920 in 1861; in 1872 it was 22,254,002 oz.; in 1888, 45,792,682; in 1890, 54,516,300 oz.; in 1900, 57,647,000; and in 1905, 58,918,839 oz. S. America has furnished European supplies since the discovery of the Potosi mines of Peru in 1533; Bolivia and Chile are also notable producers. Of European producers, Germany, Spain and Austria are the most important; Greece, Italy, France, Turkey and Russia occupy secondary positions. The German mines were worked in the 10th century; at the beginning of the 16th century the production was over 400,000 oz. annually; this dropped in the following century to about one-half; it then recovered, and in recent times has enormously increased, attaining 12,535,238 oz. in 1905. The mines of Spain, neglected late in the 15th century on the advent of supplies from America, came into note in 1827; the output has since greatly increased, amounting to 3,774,989 oz. in 1905. Austria-Hungary was producing twice as much as Germany, and about one-half of the total European production, in the 16th century; the yield diminished in the ensuing century, to be subsequently increased. The output was about 1,800,000 oz. in 1905. The total European supply was about 17,000,000 oz. in 1900 and about 18,600,000 oz. in 1905. Of other countries we may notice Canada, which produced 4,468,225 oz. in 1900 and 5,974.875 oz. in 1905, and Japan, which produced about 670,000 oz. in 1880 and 3,215,000 oz. in 1905. Australia came into notice chiefly by reason of the discoveries at Broken Hill, New South Wales; these mines producing 36,608 oz. in 1885, 1,016,269 in 1886, and 7,727,877 oz. in 1890. The total Australasian production in 1900 was 14,063,244 oz. and 14,362,639 oz. in 1905.


From the metallurgical point of view, silver ores may be classified as real silver ores and argentiferous ores. The former consist of silver minerals and gangue (vein matter, country-rock). The leading silver minerals are native silver; argentite or silver glance, Ag2S, usually containing small amounts of lead, copper and tin; dyscrasite or antimonial silver, Ag2Sb to Ag13Sb, an isomorphous mixture of silver and antimony; proustite or light red silver ore, Ag3AsS3; pyrargyrite or dark red silver ore, Ag3SbS3; stephanite, Ag5SbS4; miargyrite, AgSbS2; stromeyerite, CuAgS; polybasite, 9(Cu2S,Ag2S)·(Sb2S3,As2S3); cerargyrite or horn silver, AgCl; bromite or bromargyrite, AgBr; embolite, Ag(Cl,Br); iodite or iodargyrite, Agl. Metalliferous products containing silver arise in many operations; the chief products which may yield silver economically are copper and lead mattes, burnt argentiferous pyrites and certain drosses and scums. Argentiferous ores consist of silver-bearing base-metal minerals and gangue. Lead and copper ores, carrying silver in some form or other, are the leading representatives. The silver is extracted from the gangue with the base metal, usually by smelting, and the two are then separated by special processes (see Lead).

Milling, i.e. amalgamation and lixiviation, is cheaper than smelting, but the yield in silver is lower. Often it is more profitable to smelt real silver ores with argentiferous ores than to mill them, the greater cost being more than balanced by the increased yield. Milling is practised mainly in isolated localities near the mine producing the ore. As any given region is opened up by railways, cheapening transportation, milling is apt to give way to smelting. Thus on the American continent, which produces the bulk of the world’s silver, milling is still prominent in S. America and Mexico, while in the United States it has to a considerable extent been replaced by smelting.

Amalgamation is based on the property of quicksilver to extract the silver from finely-pulverized ore and collect it in the form of an amalgam. When the rock has been separated from the amalgam by a washing operation, the quicksilver is recovered by distillation in an iron retort, and the remaining crude retort-silver melted into bars and shipped to a refinery, which removes the impurities, the leading one of which is copper. A silver ore is either free-milling or refractory, that is, the silver mineral is readily amalgamated or it is not. In free-milling ore the silver is present either in the native state, or as chloride or as simple sulphide. Complex silver minerals (sulph-arsenides and antimonides) which are difficult to amalgamate must be made amenable to quicksilver, and the simplest way of doing this is to convert the silver into chloride. This is imperfectly accomplished, in the wet way, by cupric and cuprous chloride solutions, but completely so, in the dry way, by roasting with salt (chloridizing roasting). According as a preliminary chloridizing roast has or has not been given, the process is classed as roast-amalgamation or raw-amalgamation. The leading raw-amalgamation processes are the Patio and Washoe; then follow the Cazo, Fondon and Krohnke; of the roast-amalgamation processes, the European Barrel or Freiberg, the Reese River and the Franck-Tina are the most important.

The Patio process, sometimes named the American-heap-amalgamation process, which is carried out principally in Mexico, aims at amalgamating the silver in the open in a circular enclosure termed a torla, the floor of which is generally built of flagstones. In order to facilitate the decomposition of the silver-mineral, salt and magistral, i.e. cupriferous pyrites roasted to convert the copper into soluble sulphate, which is the active agent, are worked into the wet pulp spread out on the floor. The amalgamation proceeds very slowly, as the sole extraneous heat is that of the sun. According to Laur (“Metallurgie de l’argent au Mexico,” Ann. des mines, series 6, vol. xx.), at Guanaxuato, Mexico, 92·79% of the total silver recovered was extracted after 12 days, 97·55% after 25 days, 99·1% after 28 days and 100% after 33 days. The loss of quicksilver in the process is large, owing to the formation of calomel which is not saved. The yield in silver is low unless the ores are exceptionally free-milling; the bullion produced is high-grade, as refractory silver minerals are hardly attacked. The process is suited to easy ores and a region where the climate is warm and dry, and horse- or mule-power, labour and quicksilver are cheaper than fuel and water.

The Washoe process of pan-amalgamation, named from the Washoe district in the United States, is the leading raw-amalgamation process of the United States, where it was introduced in 1861 by A. B. Paul. It consists in wet-stamping coarsely crushed ore, settling the sands and slimes produced, and grinding and amalgamating them in steam-heated iron pans with or without the use of chemicals (salt and copper sulphate). The ores may contain a larger proportion of sulphides and complex silver minerals than with the Patio process and still give a satisfactory extraction. They are crushed to egg-size in a rock-breaker, and pulverized to pass a 40-mesh sieve in a California stamp-mill, which treats in 24 hovirs about 3 tons per stamp. A 10-stamp mill is fed by one rock-breaker, and discharges the liquid pulp into 10-15 wooden settling tanks, 9 by 5 by 8 ft., the settled contents of which are shovelled out and charged into the pans. The

Ean in general use is the combination pan. It has a flat cast-iron ottom, 5 feet in diameter, and wooden sides about 30 inches high, the lower parts of which are lined with cast-iron. In the centre is a hollow cone, through which passes the driving shaft, geared from below. This turns the grinding apparatus (driver with “muller”), which can be raised and lowered. The speed is 60-90 revolutions per minute. To the bottom and muller are attached grinding plates (shoes and dies), which are replaced when worn; and to the sides three wings to deflect the moving pulp towards the centre, and thus establish the necessary pulp current. The lower side of the bottom has also a steam-chest. A 10-stamp mill has 4-6 pans, which receive 2-ton charges. In working, the muller is raised 1/2 in., the pan charged with water and then with ore; the muller is then lowered, salt and blue vitriol added, and the charge ground for 3-4 hours. The pulp is heated with live steam to about 90° C, and kept at that temperature by exhaust steam in the bottom-chest. After grinding, the muller is raised and quicksilver added, and the silver up to 81-04 % then amalgamated in 4 hours.

In amalgamating without the use of chemicals, finely divided iron, worn from the shoes and dies in the stamp-mill and the pan, decomposes cerargyrite and argentite, and the liberated silver is taken up by the quicksilver; the process is hastened by adding salt. When salt and copper sulphate are added to the charge, they form sodium sulphate and cupric chloride, both of which are readily soluble in water. Cupric chloride acts upon argentite (Ag2S+CuCl2=2AgCl+CuS), proustite (4Ag3AsS3+4CuCl2=8AgCl+2Ag2S+4CuS+2As2S3), pyrargyrite (2Ag3SbS3+3CuCl2=6AgCl+3CuS+Sb2S3), and is also reduced to cuprous chloride by metallic iron. This salt, insoluble in water but soluble in brine, also acts upon argentite (Ag2S+Cu2Cl2=2AgCl+CuS+Cu) and pyrargyrite (2Ag3SbS3+Cu2Cl2=2AgCl+Ag2S+2Ag+2CuS+Sb2S3), and would give with silver sulphide in the presence of quicksilver, the Patio-reaction; metallic silver, cupric sulphide, and mercurous chloride (2Ag2S+Cu2Cl2+2Hg=4Ag+2Cu2+Hg2Cl2), but the iron decomposes the quicksilver salt, setting free the quicksilver.

The amalgamation is rapid. Thus Austin found that at the Charleston mills, Arizona, 92·13% of the total silver recovered was extracted after 1 hour, 94·10% after 2 hours, 95·92% after 3 hours, and 100% after 4 hours. The loss in quicksilver is small, as there is no chemical loss inherent in the process; the yield is relatively high, but the bullion is liable to be low-grade, on account of copper being precipitated and amalgamated.

When the charge has been worked, the contents of the pan are discharged into a settler, in which the amalgam is separated from the sands. It has the same general construction as the pan. It is 8 ft. in diameter and 3 ft. deep. The bottom, slightly .conical, has a groove near the circumference to catch the amalgam, which is withdrawn through a discharge-spout into a bowl. In the sides at different levels are three discharge-holes for water and sand. The muller reaches to within 3 in. of the bottom and makes 12-15 revolutions per minute. In settling, the pulp is diluted by a small stream of water, and the thinned pulp drawn off, first through the top discharge- hole and then through the other two, the bottom one being about 8 in. above the amalgam. Settling takes about half the time required to work a charge in the pan, hence one settler serves two pans. The amalgam is dipped out from the bowl into a canvas bag (the strainer), to separate the excess of the quicksilver from the pasty amalgam, which is then retorted and melted. The cost of treating a ton of ore in the western part of the United States is from $3 to $7. At some works treating ores containing sulphides which do not yield their silver to quicksilver, concentration apparatus (see Ore-Dressing) is inserted between the stamps and the settling tanks to remove the sulphides, which are worked by themselves; at other works they are recovered from the sands after these have left the settlers. In order to do away with the handling of the wet pulp, and to obtain a higher extraction, M. P. Boss has modified the ordinary plant by making the pulp flowing from the stamps pass through a grinding pan, then through a series of amalgamating pans followed by a row of settlers.

A 20-stamp mill is served by 12 men in 24 hours. The Washoe process is independent of the climate, but it requires cheap power and an abundance of water.

In the Cazo, Caldron or Hot process the pulverized silver ore is boiled in a copper-bottomed wooden vat, first with brine until the silver has been reduced by the copper, and then with quicksilver. The Fondon is an improvement on the Cazo. Bars of copper drawn over the bottom by mules or water-power (like the stone drags in the arrastra) grind off fine particles of copper, which hasten the reduction of the silver and diminish the formation of calomel. In the Krohnke process introduced by B. Krohnke into Copiapo, Chile, in i860, the silver mineral of the pulverized ore is decomposed in a revolving barrel by a hot solution of cuprous chloride in brine in the presence of zinc or lead and quicksilver (see B. Kröhnke, Methode zur Entsilberung von Erzen, Stuttgart, 1900).

Chloridizing Roasting.—In a chloridizing roast chlorine produces its effect as nascent chlorine or gaseous hydrochloric acid. The leading reagents are salt (NaCl), sulphur trioxide (SO3, produced in the roasting), and steam (H 2 0). The decomposition of salt is expressed by 2NaCl+2SO3 = Na2SO4+SO2+Cl2. In the presence of water-vapour the following reaction takes place: 2NaCl+SO3+H2 = Na2SO4+2HCl. As some water-vapour is always present, hydrochloric acid will invariably be formed with the chlorine. The roasting is carried on in hand and mechanical reverberatory furnaces, and Occasionally in muffle-furnaces. A chloridation of over 90% silver is the rule.

The European Barrel or Freiberg process consists in roasting the ground ore with salt which converts the silver sulphide into chloride. The mass, along with certain proportions of water, scrap-iron and mercury, is then placed in barrels, which are made to rotate so that the several ingredients are thoroughly mixed. The salt solution dissolves a small proportion of chloride, which in this form is quickly reduced by the iron to the metallic state. This solution .and precipitation is continuous, and the metal formed unites with the mercury to form a semi-fluid amalgam. The amalgam is pressed in linen bags to eliminate a quantity of relatively silver-free liquid mercury (which is utilized as such in subsequent operations), and the remaining solid amalgam is subjected to distillation from iron retorts. This process was perfected at Freiberg, Saxony, but abandoned there in 1856. In the United States it was used quite extensively in Colorado and Nevada, but has now been given up. The main reasons for this are the length of time required to finish a charge, on account of the absence of any extraneous source of heat, and the great care with which operations have to be carried out in order to obtain satisfactory results.

The Reese River or pan-amalgamation process consists in dry-stamping crushed dried ore and dried salt (separately or together), charging them into a roasting furnace, and amalgamating the chloridized ore in an iron pan. The general arrangement and construction of a mill resemble those of the Washoe process. The apparatus for drying ore and salt varies greatly, drying-floors, dry-kilns and continuous mechanical reverberatory furnaces with stationary and revolving hearths being used. The general construction of the pan is the same as in the Washoe process; the management, however, differs. The steam-chest is not used to such an extent, as the bottom would be prematurely corroded; less water is used, as the pulp would become too thin on account of the soluble salts (sodium chloride, sulphate, &c.) going into solution; and the roasted ore is not ground, as the hot brine readily dissolves the silver chloride from the porous ore, and thus brings it into intimate contact with iron and quicksilver. Chemical reagents are sometimes added—lime or sulphuric acid, to neutralize an excess of acid or alkali; copper sulphate, to form cuprous chloride with sodium chloride; and iron and zinc, to make the galvanic action more energetic and reduce the consumption of iron. The rest of the apparatus (settler, retort, crucible, furnace) is the same as with the Washoe process. The Reese River process costs from half as much again to twice as much as the Washoe process.

The Francke-Tina process, named from Francke, German consul at Bolivia, and Una, the wooden vat in which the process is carried out. was developed in Bolivia for the treatment of refractory ores rich in zinc blende and tetrahedrite (fahl-ore). The ore is given only a partial chloridizing roast, on account of the great loss in silver that would be caused by the formation of zinc chloride. The large amount of soluble sulphates of iron and copper formed in the roast is made to act upon salt charged in a copper-bottomed amalgamating pan; the chlorides formed finish in the wet way the imperfect chloridation obtained in the furnace.

Lixiviation. — Ores suited for amalgamation can, as a rule, be successfully leached. In leaching, the silver ore is subjected to the action of solvents, which dissolve the silver; from the solution the silver is precipitated and converted into a marketable product.

The leading solvents are aqueous solutions of thiosulphates, un- systematically but generally termed hyposulphites. Sodium chloride, characteristic of the Augustin process in which the ores, after a chloridizing roast, were extracted with brine, and the silver pre- cipitated by copper, has almost wholly fallen into disuse; and potassium cyanide, which has become a very important solvent for finely divided gold, is rarely used in leaching silver ores. The use of sodium hyposulphite as solvent, and sodium sulphide as precipitant, was proposed in 1846 by Hauch and in 1850 by Percy, and put into practice in 1858 by Patera {Patera process); calcium hyposulphite with calcium polysulphide was first used by Kiss in i860 (Kiss process, now obsolete) ; sodium hyposulphite with calcium poly- sulphide was adopted about 1880 by O. Hofmann (Hofmann process) ; finally, sodium hyposulphite with cuprous hyposulphite was first applied by Russell in 1884, who included in his process the acidula- tion of the first wash-water (to neutralize any harmful alkaline reaction), and the separation of lead with sodium carbonate from the silver solution previous to precipitating with sodium sulphide (see C. A. Stetefeldt, The Lixiviation of Silver Ores with Hyposulphite Solutions, &c, New York, 1888).

In all processes the silver ore is finely crushed, usually by rolls, as, because making few fines, they leave the ore in the best condition for leaching. As a. rule the ore is subjected to a preliminary chloridizing roast, though occasionally it may be leached raw. The vats in common use are circular wooden tanks, 16-20 ft. in diameter and 8-9 ft. deep if the leached ore is to be removed by sluicing, 5 ft. if by shovelling. They ha^e a false bottom, with cloth or gravel filters.

The basis of the following outline is the Patera process. The ore, supposed to have been salt-roasted, is charged loosely into the leaching vat and treated with water (to which sulphuric acid or copper sulphate may have been added), to remove soluble salts, which might later on be precipitated with the silver (base-metal chlorides), or overcharge the solution (sodium chloride and sulphate), or interfere with the solvent power (sodium sulphate). The vat is filled with water from above or below, in- and out-flow are then so regulated as to keep the ore covered with water. Any silver dissolved by the first wash-water is recovered by a. separate treatment. After the wash- water has been drained off, the ore is ready for the silver solvent. This is a solution containing up to 2 % of sodium hyposulphite, of which one part dissolves 0-485 part silver chloride, equivalent to 0-365 part metallic silver, to form double hyposulphites. Silver arsenate and antimoniate are also readily soluble, metallic silver slightly so, silver sulphide not at all. (In the Russell-process double salts: 4Na 2 S503-3Cu2S 2 03, and 8NaoSo0 3 -3Cu3S 2 3 the metallic silver and silver sulphide are readily soluble ; thus it supplements that of Patera.)

After the silver has been dissolved by percolation, the last of the solvent still in contact with the ore is replaced by a second wash- water. The silver solution, collected in a circular precipitating vat (10 ft. in diameter and 10 ft. deep), is treated with sodium sulphide (or calcium polysulphide), unless sodium carbonate was first added to throw down any lead, present in the ore as sulphate, that had gone into solution. Silver sulphide falls out as a black mud, with about 50 % silver, and the solvent will be regenerated.

If the sodium cuprous hyposulphite was used as a solvent in addition to the simple sodium hyposulphite, cuprous sulphide will be precipitated with the silver sulphide, and the precipitate will be of lower grade. At some works the silver is precipitated with sodium sulphide, and the liquor, after having been separated from the silver sulphide, is treated with calcium polysulphide, that by the precipi- tation of calcium sulphate the accumulation of sodium sulphate may . be prevented. The precipitated silver (copper) sulphide is filtered, dried, and usually shipped to silver-lead works to be refined ; some- times it is converted into metallic silver at the works. The solution, freed from silver, is used again as solvent. Lixiviation has many advantages over amalgamation. It permits coarser crushing of the ore, the cost of plant is lower, the power required is nominal, the cost of chemicals is lower than that of quicksilver, less water is necessary, and the extraction is often higher, as silver arsenate and antimoniate are readily soluble, while they are not decomposed in amalgamation. On the other hand, silver and silver sulphide are readily amalga- mated ; and while they are not dissolved in the Patera process, they are in the Russell process.

Mention may be made of the Ziervogel process, introduced at Hettstadt in 1841 for the purpose of extracting silver from copper mattes. In principle it consists in oxidizing silver sulphide to the sulphate which is soluble in water, the silver being then precipitable by metallic copper. This process when carefully carried out, especi- ally as to the details of the roasting process whereby the silver sulphide is oxidized, yields 92 % of the silver originally present.

Electrolytic Methods. — Crude silver generally contains small amounts of copper, gold, bismuth, lead and other metals. To

eliminate these impurities, electrolytic methods have been devised; of these that of Moebius is the most important and will be described in detail.

Under his earlier patent of 1884, cast crude silver anode plates, about \ in. thick, and thin rolled silver cathodes, were suspended in a i%, slightly acid, solution of silver nitrate contained in tarred wooden tanks. The deposit from this solution even with low current- densities is pulverulent and non-coherent, and therefore during electrolysis wooden scrapers are automatically and intermittently passed over the surface of the cathode to detach the loose silver, which falls into cloth trays at the bottom of the tanks. These trays are removed at intervals, and the silver washed and cast into bars, which should contain over 99-9% of pure metal. The relatively electro-negative character of silver ensures that with moderate current densities no metal (other than precious metals) will be deposited with it ; hence, while the solution is pure a current-density of 30 amperes per sq. ft. of cathode may be used, but as copper accumulates in it, the current-density must be diminished to (say) 15 to 20 amperes per sq. ft., and a little extra nitric acid must be added, in order to prevent the co-deposition of copper. A pressure of 1 -5 volt usually suffices when the space between the electrodes is 2 in. The tanks were arranged in groups of seven on the multiple system.

Of the metals present in the anode, practically all, except gold, pass into solution, but, under the right conditions, only silver should deposit. The whole of the gold is recovered as anode slime in cloth bags surrounding the anodes. Practical results with a large plant indicate an expenditure of 1-23 electrical horse-power hours per 100 oz. (Troy) of refined silver. In later installations, under the 1895 patent, the anodes are placed horizontally on a porous tray resting within the solution above an endless silver band revolving, also horizontally, over rollers placed near the ends of a long shallow tank. The revolving band forms the cathode, and at one end makes a rubbing contact with a travelling belt placed at an angle so that the crystals of silver detached thereby from the cathode are con- veyed by it from the solution anddeposited outside.

Alloy scrap containing chiefly copper with, say 5 or 6% of gold, and other metals, and up to 40 or 50% of silver, is often treated electrolytically. Obviously, with modifications, the Moebius process could be applied. Other systems have been devised. Borchers uses the alloy, granulated, in an anode chamber separated from the cathode cell by a porous partition through which the current, but not electrolyte, can pass freely. The anode residue is collected in the angular bottom of the tank, the electrolyte passes from the anode chamber to a series of tanks in which the more electro-negative constituents (silver, &c.) are chemically separated, and thence to the cathode chamber, where the copper is deposited electrolytically, thence it passes again to the anode chamber and so completes the cycle. In one form of the apparatus a rotating cathode is used. Dietzel has described (Zeitschrift fur Elektrochem., 1899, vol. vi. p. 81) the working of his, somewhat similar, process at Pforzheim, where about 130 ft of the alloy was being treated by it daily in 1899. The alloy is cast into anode plates about -J in. thick, and placed in the anode chamber beneath the cathode cell, and separated from it by linen cloth. In the upper compartment are two large revolving horizontal cathode cylinders. Acidified copper nitrate solution is run into this cell, copper is deposited, and the more or less spent solution then passes through the linen partition, and, taking up metal from the anodes by electrolytic solution, is run out of the trough through a series of _ vessels filled with copper by which the silver _ is precipitated by simple exchange; after acidification the resulting silver-free copper solution is returned to the cathode cell for the deposition of the copper, the solution being employed again and again until too impure for use.

Chemically Pure Silver. — Even the best " fine " silver of commerce contains a few thousandth-parts of copper or other base metal. To produce perfectly pure metal the usual method is to first prepare pure chloride and then to reduce the chloride to metal. This may be effected by mixing the dry chloride with one-fifth of its weight of pure quicklime or one-third of its weight of dry sodium carbonate, and fusing the mixture in a fire-clay crucible at a bright red heat. In either case we obtain a regulus of silver lying under a fused slag of chloride. The fused metal is best granulated by pouring it into a mass of cold water. A convenient wet method for small quantities is to boil the recently precipitated chloride (which must have been produced and washed in the cold) with caustic soda and just enough sugar to reduce the silver oxide (Ag 2 0) transitorily produced. The silver in this case is obtained as a yellowish grey heavy powder, which is easily washed by decantation; but it tends to retain unreduced chloride, which can be removed only by fusion with carbonate of soda.

Stas in his stoichiometric determinations employed the following process as yielding a metal which comes nearer ideal purity. Slightly cupriferous silver is made into dry nitrate and the latter fused to reduce any platinum nitrate that may be present to metal. The fused mass is dissolved in dilute ammonia and diluted to about fifty times the weight of the silver it contains. The filtered (blue) solution is now mixed with an excess of solution of ammonium sulphite, and allowed to stand. After twenty-four hours about one-half of the silver has separated out in crystals ; from the mother-liquor the rest comes down promptly on application of a water-bath heat. The rationale of the process is that the sulphite hardly acts upon the dis- solved oxide of silver, but it reduces some of the cupric oxide to cuprous oxide, which reduces its equivalent of silver oxide to silver and reforming cupric oxide which passes through the same cycle.

Alloys of Silver. — Silver readily alloys with many metals, and the admixture generally differs in physical properties from the pure metal. Thus arsenic, antimony, bismuth, tin or zinc render the metal brittle, so that it fractures under a die or rolling mill; copper, on. the other hand, increases its hardness, makes it tougher and more readily fusible. Consequently copper-silver alloys receive extensive application for coinage and jewelry. The composition of the alloy is stated in terms of its " fineness," the proportion of silver in iooo parts of alloy. Generally copper-silver alloys separate into two layers of different composition on fusion; an exception is the alloy Ag 3 Cu 2 , investigated by A. I. F. Levol, corresponding to a fineness of 719, which remained perfectly homogeneous.

The extent to which the properties of silver are modified by addition of copper depends on the fineness of the alloy produced. The addition of even three parts of copper to one of silver does not quite obliterate the whiteness of the noble metal. According to Kamarsch, the relative abrasion suffered by silver coins of the degrees of fineness named is as follows : —

Fineness .... 312 750 900 993

Abrasion .... I 2-3 3-9 9-5

The same observer established the following relation between fine- ness p and specific gravity of alloys containing from 375 to 875 of silver per 1000: — sp. gr. =0-001647 />+8-833.

The fusing points of all copper-silver alloys lies below that of pure copper; that of British standard silver is lower than even that of pure silver

Compounds of Silver.

Silver forms one perfectly characterized oxide, Ag 2 0, from which is derived a series of stable salts, and probably several less perfectly known ones. Argentic or silver oxide, Ag20, is obtained as a dark brown precipitate by adding potash to a solution of a silver salt; on drying at 6o°-8o° it becomes almost black. It is also obtained by digesting freshly precipitated silver chloride with potash. It is sparingly soluble in water (one part in 3000) ; and the moist oxide frequently behaves as the hydroxide, AgOH, i.e. it converts alkyl haloids into alcohols. It begins to decompose into silver and oxygen at 250 . Silver peroxide, AgO, appears under certain conditions as minute octahedra when a solution of silver nitrate is electrolysed, or as an amorphous crust in the electrolysis of dilute sulphuric acid between silver electrodes. It readily decomposes into silver and oxygen. It dissolves in ammonia with the liberation of nitrogen and the formation of silver oxide, Ag 2 0; and in sulphuric acid forming a fairly stable dark green liquid which, on dilution, gives off oxygen and forms silver sulphate. It is doubtful whether the pure compound has been obtained. The compound obtained from silver nitrate always contains nitrogen; it appears to have the constant composition Ag 7 NOn, and has been named silver peroxynitrate. Similarly the sulphate yields 5Ag 2 2 , 2Ag 2 S0 7 , silver peroxysulphate, and the fluoride the peroxyfluorides' AguFsOis, Ag 7 FO s . The sesquioxide, Ag<03, is supposed to be formed when silver peroxide is treated with ammonia (Watson, Jour. Chem. Soc, 1906, 89, p 578).

Silver chloride, AgCl, constitutes the mineral cerargyrite or horn silver; mixed with clay it is the butter-milk ore of the German miners. Early names for it are Lac argenti and Luna cornea, the first referring to its form when freshly precipitated, the latter to its ap- pearance after fusion. It is readily obtained as a white curdy precipitate by adding a solution of a chloride to a soluble silver salt It is almost insoluble in water, soluble in 50,000 parts of nitric acid, and more soluble in strong hydrochloric acid and solutions of alkaline chlorides. It readily dissolves in ammonia, the solution, on evapora- tion, yielding rhombic crystals of 2AgCl-3NH s ; it also dissolves in sodium thiosulphate and potassium cyanide solutions. On exposure to light it rapidly darkens, a behaviour utilized in photography (q.v.). Abney and Baker have shown that the pure dry chloride does not blacken when exposed in a vacuous tube to light, and that the blackening is due to absorption of oxygen accompanied by a loss of chlorine. Hydrogen peroxide is also formed. It melts at about 460 to a clear yellow liquid, which, on cooling, solidifies to a trans- lucent resinous mass. It is reduced to metallic silver by certain metals — zinc, iron, &c. — in the presence of water, by fusion with alkaline carbonates or cyanides, by heating in a current of hydrogen, or by digestion with strong potash solution, or with potassium carbonate and grape sugar. Silver bromide, AgBr, constitutes the

mineral bromargyrite or bromyrite, found in Mexico and Chile. It is obtained as a yellowish white precipitate by mixing solutions of a bromide and a silver salt. It is very slightly soluble in nitric acid, and less soluble in ammonia than the chloride. It melts at 427°, and darkens on exposure to air. The minerals embolite, mega- bromite and microbromite, occurring in Chile, are variable mixtures of the chloride and bromide. Silver iodide. Agl, occurs in nature as the mineral iodargyrite or iodyrite, forming hexagonal crystals, or yellowish green plates. It is obtained as a light yellow powder by dissolving the metal in hydriodic acid, or by precipitating a silver salt with a soluble iodide. It is very slightly soluble in acids and ammonia, and almost insoluble in alkaline chlorides; potassium iodide, however, dissolves it to form Agl-KI. Silver iodide is dimorphous; at ordinary temperatures the stable form is hexa- gonal; on heating to about 138 the colour changes from deep yellow to yellowish-white with the formation of cubic crystals. Silver fluoride, AgF, is obtained as quadratic octahedra, with one molecule of water, by dissolving the oxide or carbonate in hydrofluoric acid. It is deliquescent, and dissolves in half its weight of water to form a strongly alkaline liquid. It is not decomposed by sunlight. It melts at 435 and, on cooling, forms a yellow transparent mass. In addition to the salts described above there exist sub-salts. Silver nitrate, AgNOs, one of the most important silver salts, is obtained by dissolving the metal in moderately dilute nitric acid; on evaporation it separates in the anhydrous form as colourless triclinic plates. It dissolves in water, alcohol and ether. It stains the skin and hair black: an ethereal solution having been employed as a dye for the hair. Mixed with gum arabic it forms a marking ink for linen. It fuses at 218°; and when cast in quill-like moulds, it constitutes the lunar caustic of medicine, principally used as a cauterizing agent.

Silver sulphide, Ag 2 S, constitutes the mineral argentite or silver glance, and may be obtained by heating silver with sulphur, or by precipitating a silver salt with sulphuretted hydrogen. Thus ob- tained it is a brownish solid, which readily fuses and resolidifies to a soft leaden-grey mass. It forms with silver nitrate the yellowish green solid, Ag 2 S-AgNC>3, and with silver sulphate the orange-red powder, Ag 2 S-Ag 2 S04. Silver sulphate, Ag 2 SO<, is obtained as white crystals, sparingly soluble in water, by dissolving the metal in strong sulphuric acid, sulphur dioxide being evolved, or by adding strong sulphuric acid to a solution of the nitrate. It combines with ammonia to form the readily soluble 2NH 3 -Ag 2 S04. Silver selenide, Ag 2 Se, resembles the sulphide. It occurs in the minerals naumannite, PbSe-Ag 2 Se, and eukairite, Ag 2 Se-Cu 2 Se. The telluride, Ag 2 Te, occurs in nature as the mineral hessite.

Fulminating silver is an extremely explosive black powder, first obtained in 1788 by Berthelot, who acted with ammonia on silver oxide (prepared by adding lime water to a silver solution). When dry it explodes even on touching with a feather. It appears to be silver nitride AgaN, but it usually contains free silver and sometimes hydrogen. It is to be distinguished from silver fulminate (see Fulminic Acid). The nitride AgN 3 , silver azoimide (q.v.), is also highly explosive.

See j. Percy, Metallurgy of Silver and Gold (London, 1880), part i. ; T. Egleston, The Metallurgy of Silver, Gold and Mercury (New York, 1887-1890), part i. ; M. Eissler, The Metallurgy of Silver (London, 1891); H. F. Collins, The Metallurgy of Lead and Silver (London, 1900), part ii. ; H. O. Hofman, Hydrometallurgy of Silver (1907); C. Schnabel, Metallurgy, translated by H. Louis, 2nd ed. vol. i. (I905)-

Medicinal Use.

Two salts of silver are used in the British pharmacopoeia. (1) Argenti nitras (United States and British pharmacopoeia), lunar caustic, incompatible with alkalis, chlorides, acids, except nitric and acetic, potassium iodide and arsenical solutions. From the nitrate are made (a) argenti nitras indurata, toughened caustic, containing 19 parts of silver nitrate and one of potassium nitrate fused together into cylindrical rods; (b) Argenti nitras mitigatus, mitigated caustic, in which 1 part of silver nitrate and 2 parts of potassium nitrate are fused together into rods or cones. (2) Argenti oxidum, incompatible with chlorides, organic substances, phenol, creosote, &c, with which it forms explosive compounds.

Therapeutics. — Externally the nitrate has a caustic action, destroying the superficial tissues and separating the part acted on as a slough. Its action is limited. It may be" employed to destroy warts or small growths, to reduce exuberant granulations or it may be applied to bites. In granular lids and various forms of ophthalmia solutions of silver nitrate (2 grs. to I fl. oz.) are employed. A I % solution is also used as a prophylactic for ophthalmia neonatorum. The effects of the nitrate being both astringent and stimulating as well as bactericidal, solutions of it are used to paint indolent ulcers, and in chronic pharyngitis or laryngitis. Salts of silver are most useful as an injection in subacute and chronic gonorrhoea, either the nitrate (1 to 5% solution) being employed, or protargol, which is a proteid compound containing 8 % of silver nitrate, is used in I % solution; they also benefit in leucorrhoea. In pruritus of the

vulva and anus a weak solution of silver nitrate will relieve the itching, and strong solutions painted round the base of a boil at the beginning will abort its formation. Internally the nitrate has been used in the treatment of gastric ulcer, in ulcerative conditions of the intestine and in chronic dysentery. For the intestinal conditions it must either be given in a keratin-coated pill or injected high up into the rectum The oxide has been given in epilepsy and chorea. Nitrate of silver is eliminated from the system very slowly and the objection to its employment continuously as a drug is that it is deposited in the tissues causing argyria, chronic silver poisoning, of which the most prominent symptom is dark slate-blue colour of the lips, cheeks, gums and later of the skin.

Taken in large doses nitrate of silver is a powerful poison, causing violent abdominal pain, vomiting and diarrhoea with the development of gastro-enteritis. In some cases nervous symptoms and delirium supervene. The treatment consists in the use of solutions of common salt, followed by copious draughts of milk or white of egg and water or soap in water, in order to dilute the poison and protect the mucous membranes of the oesophagus and stomach from its action.