"Westminster" Secies

THE PRECIOUS METALS

 " "
The Westminster Series
Crown 8-vo. Fully Illustrated. 6s. net each.

The Manufacture of Paper. By R. W. SINDALL,

F.C.S.
Timber. By J. R. BATERDEN, A.M.I. C.E.
Electric Lamps. By MAURICE SOLOMON, A. C.G.I.,
A.M.I.E.E.
Textiles and their Manufacture. By ALDRED
BARKER, M.Sc., Technical College, Bradford.
The Precious Metals. By T. KIRKE ROSE, D.Sc.,
of the Royal Mint.
Ornamental Glass Work. By A. L. DDTHIE.
The Railway Locomotive. By VAUGHAN PENDRED,
M.I.Mech.E., late Editor of" The Engineer."
Iron and Steel. By J. H. STANSBIE, B.Sc. (Lond.),
F.I.C.
Town Gas for Lighting and Heating. By W. H. Y.
WEBBER, C.E.
Liquid and Gaseous Fuels, and the Part they play
in Modern Power Production. By Professor VIVIAN B.
LEWES, F.I.C., F.C.S., Prof, of Chemistry, Royal Naval
College, Greenwich.
Electric Power and Traction. By F. H. DAVIES,
A.M.I.E.E.
Coal. By JAMES TONGE, M.I.M.E., F.G.S., etc.,
Lecturer on Mining at Victoria University, Manchester.
India-Rubber and its Manufacture, with Chapters
on Gutta-Percha and Balata. By H. L. TERRY, F.I.C.,
Assoc. Inst. M.M.
The Book: Its History and Development. By
CYRIL DAVENPORT, F.S.A.
Glass. By WALTER ROSENHAIN. Superintendent of
the Department of Metallurgy in the National Physical
Laboratory.
Patents, Trade Marks and Designs. By KENNETH
R. SWAN, B.A. (Oxon.), of the Inner Temple, Barrister-at-
Law.
Precious Stones. With a Chapter on Artificial
Stones. By W. GOODCHILD, M.B., B.Ch.
Electro-Metallurgy. By J. B. C. KERSHAW, F.I.C.
Natural Sources of Power. By ROBERT S. BALL,
B.Sc., A.M.I.C.E.
Radio-Telegraphy. By C. C. F. MONCKTON, M.I.E.E.
* further ^volumes 'will appear at short intervals.
 OF THE
UNIVERSITY
OF
p j
II
h
^ ~.
S -5
03
S
S s
o 5"
1-3
8 I
fl
rt
S
a
rF >
 THE
PRECIOUS METALS
COMPRISING

GOLD SILVER AND PLATINUM

BY

T. KIRKE ROSE (i
A.R.S.M. D.Sc.
CHEMIST AND ASSAYER OF THE ROYAL MINT VICE-PRESIDENT
J

OF 'THE INSTITUTION OF MINING AND METALLURGY: HONORARY

MEMBER OF THE CHEMICAL, METALLURGICAL AND MINING
SOCIETY OF SOUTH AFRICA J
FELLOW OF THE CHEMICAL
SOCIETY, ETC.

LONDON
ARCHIBALD CONSTABLE & CO. LTD.
10 ORANGE STREET LEICESTER SQUARE W.C.

1909
SPRECKEt.3

BRADBURY, AGNEW, & CO. LD. PRINTKK:

,

LONDON AND TONBRIDOE.

 PREFACE
" "
THE term precious metals has usually been applied
to gold and silver only, its use dating back to a time when
no regard was paid to platinum. As platinum is now in

common use and is more valuable, weight for weight,

than gold, it may very well rank as a precious metal.
Other metals of high price come under the heading of
rare metals.

In the preparation of this work, the aim has been to

provide an introduction to the study of the, precious metals,

and an elementary book of reference for those who do not
wish to pursue the subject further. With these objects in
view details have been generally omitted and a wide extent
of ground has been covered, including the technical pro-
cesses of minting and the manufacture of gold and silver

wares. It need hardly be mentioned, however, that all

questions of currency and finance have been left untouched.

It is also perfectly obvious that the book is not intended to
be a substitute for the larger treatises on metallurgy, some
of which devote the whole of a bulky volume to the full

consideration of a section of the subject dealt with here in

the course of a few pages.

The author desires to express his thanks to Mr. Edward

I
viii PREFACE

Rigg, I.S.O., Superintendent of the Operative Department

of the Mr. Arthur Westwood, Assay Master
Royal Mint, to

of the Birmingham Assay Office, and to others for kindly

reading the proofs of various chapters and for their valuable

suggestions.

NORTHWOOD,
February 1st, 1909.
 CONTENTS
CHAPTER I.

HISTORY OF GOLD.

Gold in prehistoric times Its ancient names Early methods of

Egypt Methods of refining gold practised by the
extraction in
Romans Alchemy Treatment of gold ores in the Middle
Ages Use of mercury Later processes Parting gold from
silver .pp. 1 10

CHAPTER II.

PROPERTIES OF GOLD.

Physical properties, colour, hardness, malleability, ductility, tenacity,

crystallisation, density, melting point, volatilisation, conduc-
tivity for electricity and heat Chemical properties Noble metals
Chemical relations of gold Solvents for gold Identification of
gold Reactions of

Preparation of pure gold

gold
.....
Colloidal gold Purple of Cassius
pp. 11 19

CHAPTER III.

COMPOUNDS OF GOLD.

Chlorides of gold Aurous chloride Auric chloride, its preparation,

propertiesand decomposition Tests for auric chloride Chloro-
auric acid Bromides of gold Iodides of gold Cyanides of gold
Aurocyanides, their formation and decomposition Auricyanides
Sulphocyanides of gold Oxides of gold Fulminating gold
Thiosulphates of gold Silicates of gold Sulphides of gold
Selenide, telluride, phosphide, arsenide, 'and antimonide of
gold pp. 2036
x CONTENTS

CHAPTEE IV.

ALLOYS OF GOLD.

Gold and silver Gold and copper Gold and mercury Gold and
zinc Gold and cadmium Gold and tin Gold and antimony
Gold and arsenic Gold and bismuth Gold and lead Gold and
iron Gold and nickel Gold and cobalt Gold and manganese
Gold and platinum Gold and palladium, rhodium, iridium, etc.
Gold and aluminium Gold and tellurium Gold and thallium
pp. 3763

CHAPTER V.

THE OCCURRENCE OF GOLD IN NATURE : GOLD ORES.

Gold dust and nuggets Placer deposits Crystals of gold -Ores of

gold Sulphide ores Tellurides of gold Dissemination of gold in

various rocks Gold in sea water Origin of gold ores Geological
distribution of gold ores Geographical distribution of gold
ores pp. 64 71

CHAPTEE VI.

EXTRACTION OF GOLD FROM ITS ORES: GOLD WASHING.

Classification of gold ores Classification of the methods of treatment

Appliances used in washing gold ores The miner's pan The

batea
Dredging
The cradle
for gold .......
Sluices Hydraulic mining Riffles
-

pp. 72 84

CHAPTER VII.

TREATMENT OF GOLD ORES BY CRUSHING AND AMALGAMATION.

Stamp-mills Parts of a stamp Rock-breakers The method of feeding-

ores to stamp-mills Mortars Screens Cams and cam shafts
Method of running stamp-mills Water in stamp-mills Mercury
in stamp-mills Method of catching gold on amalgamated plates
The preparation and care of amalgamated plates Cleaning- up-
Retorting of amalgam Melting the gold bullion Other machines
 CONTENTS xi

for crushing gold ores Ball and roller mills Other amalgamating
machines Loss of mercury in amalgamation Treatment of
tailings from the stamp battery Concentrating machines-
Classification of pulp from stamp-mills Treatment of concen-
trates pp. 85 109

CHAPTER VIII.

TREATMENT OF GOLD ORES BY WET METHODS.

The cyanide process Collection of tailings from stamp batteries

Dissolving the gold by cyanide solutions Washing out the dis-
solved gold Precipitation of the gold by zinc Cleaning-up and
recovery of the gold Treatment of slimed ore by decantation
Electrical precipitation of gold Treatment of slimed ore by filter-

pressing Re-grinding in tube-mills Treatment of telluride ores

by cyanide solutions
chlorination process . ......
Dry crushing Roasting gold ores
pp. 110
The
127

CHAPTER IX.

SILVER AND ITS ALLOYS.

The physical and chemical properties of

Historical account of silver
silver The preparation of pure silver The alloys of silver
Silver and copper Silver and zinc Silver and cadmium Silver
and lead Silver and mercury Silver and tin Silver and platinum
Silver and aluminium Silver and bismuth Silver and nickel
Silver and palladium Silver and thallium. pp. 128
. 143
.

CHAPTER X.

COMPOUNDS AND ORES OF SILVER.

Silver oxide Silver chloride

Silver sulphide Silver bromide, iodide
and Silver-potassium cyanide Silver sulphate The
cyanide
composition and mode of occurrence of the ores of silver
>. 144153
xii CONTENTS

CHAPTER XI.

THE EXTRACTION OF SILVER FROM ITS ORES.

Amalgamation processes Direct amalgamation in tubs and in the

arrastra The method of working the patio process of amalgama-
tion and the chemical reactions involved The Cazo process The
Francke-tina process The Kronke process Pan amalgamation of
roasted and unroasted ore Lixiviation processes Roasting to
sulphate and lixiviation with water Roasting to chloride and
lixiviation with brine Treatment of roasted ore by the Patera,
Kiss and Russell thiosulphate solutions The cyanide process
Smelting with lead ores in blast furnaces Desilverisation of lead
The Pattinson process and its modifications Desilverisation by
zinc Cupellation The German cupellation furnace The English
cupellation furnace Separation of silver from metallic copper by
electrolysis .
pp. 154 182

CHAPTER XII.

REFINING GOLD AND SILVER.

Preliminary refining at ore-treatment mills Refining at the centres of

industry The sulphuric acid process Gutzkow's modifications
Refining by means of chlorine gas Electrolytic processes The
Moebius process The Wohlwill process Advantages and dis-
advantages of parting by electrolysis . . .
pp. 183 196

CHAPTER XIII.

THE ASSAY OF GOLD AND SILVER ORES.

General description Sampling and crushing ores The assay ton

Fusion of the ore sample with fluxes in crucibles and concentra-
tion of the gold and silver in lead buttons The use of various
fluxes The method of conducting the fusion Scorification of
ores with lead Cupellation of the lead buttons Losses of gold
and silver in cupellation Inquarting gold beads and parting gold
from silver Proportions of silver
Weighing the parted gold ..... to gold used in parting
pp. 197 20o
 CONTENTS xiii

CHAPTER XIV.
THE ASSAY OF GOLD AND SILVER BULLION".

Classification of methods Parting assay of gold bullion The ratio of

silver to gold in the parting assay Cupellation in the muffle
Boiling the assay pieces in nitric acid Use of check assays
Parting assay of alloys of silver and gold Cupellation assay of
silver bullion Gray Lussac process of silver bullion assaying
Volhard method of silver bullion assaying India mint method
of silver bullion assay by weighing the chloride .
pp. 206 218

CHAPTER XV.
MINTING.

Historical introduction Manufacture of coin

Valuation of bullion
Melting of bullion and gas fuel Casting coinage bars
Coke, oil

Rolling Cutting-out and marking the blanks Annealing and

blanching the blanks Striking and weighing the coins Telling
machines Composition of gold and silver coins of the principal
countries of the world f>p. 219 241

CHAPTER XVI.
MANUFACTURE OF GOLD AND SILVER "WARES.

Composition of gold wares Hall-marking Legal standards of gold

plate in various countries Composition of silver wares Legal
standards of silver plate in various countries Melting and rolling
gold and silver for wares Wire rolling and drawing Manufac-
ture of wares from plate Soldering Colouring gold wares
Imitation gold and silver wares Electro-plating Gold and silver
plating baths pp. 242 253

CHAPTER XVII.

PLATINUM.

Historical sketch Properties of platinum Platinum black Com-

pounds of platinum Alloys of platinum The occurrence of
 CONTENTS

platinum in Nature Composition of native platinum The

extraction of platinum from its ores The Wollaston process
Deville and Debray's process The preparation of pure platinum
The uses of platinum Assay of platinum and its alloys

M>. 254276

CHAPTEE XVIII.

PRODUCTION AND CONSUMPTION OF THE PRECIOUS METALS.

Production of gold in the world in ancient and modern times Present

production in various countries Production of silver in the world
Consumptionof the precious metals in the industrial arts and in

coinage
consumption of platinum .....
World's stock of the 'precious metals Production and
pp. 277 280

Index i>f>. 2S7295

 LIST OF ILLUSTRATIONS
KNIGHT'S DEEP AND SIMMER EAST JOINT PLANT. VIEW
FROM SAND DUMP Frontispiece
1. ANCIENT EGYPTIAN FURNACE 3
2. GERMAN STAMP MILL OF THE SIXTEENTH CENTURY . . 8
3. CURVE SHOWING FREEZING POINTS OF THE ALLOYS OF
GOLD AND SILVER 38
4. CURVES SHOWING FREEZING POINTS AND ELECTRICAL CON-
DUCTIVITY OF THE ALLOYS OF GOLD AND COPPER . 43
5. GOLD, 25 PER CENT. ; 'COPPER, 75 PER CENT. X 90 . . -44

6. FREEZING POINTS OF THE ALLOYS OF GOLD AND ZINC . 47

7. GOLD, 31'5 PER CENT.; CADMIUM, 685 PER CENT. X 25 49
S. GOLD, 70 PER CENT. LEAD, 30 PER CENT.
;
X 70 . . 53
9. FREEZING POINTS OF ALLOYS OF GOLD AND LEAD . . 54
10. FREEZING POINTS OF ALLOYS OF GOLD AND IRON . . 56
11. FREEZING POINTS OF ALLOYS OF GOLD AND NICKEL . . 57
12. FREEZING POINTS OF ALLOYS OF GOLD AND PLATINUM . 59
13. FREEZING POINTS OF ALLOYS OF GOLD AND PALLADIUM . 61
14. FREEZING POINTS OF MIXTURES OF GOLD AND TELLURIUM 63
15. 13ATEA 75
16. CHADLE 76
17. DIAGRAM SHOWING THE VARIOUS PARTS OF A STAMP . So
18. SECTION OF MORTAR . 86
19. SIMMER EAST STAMP MILL 100
20. TROMMEL AND SORTING TABLE, KNIGHT'S DEEP AND SIM-
MER EAST JOINT PLANT .102
21. CYANIDE PLANT AND BATTERY HOUSE, KNIGHT'S DEEP AND
SIMMER EAST JOINT PLANT 116
22. INTERIOR OF EXTRACTOR HOUSE, KNIGHT'S DEEP AND
SIMMER EAST JOINT PLANT 118
xvi LIST OF ILLUSTRATIONS

FIG. PAGR
23. SLIMES DAM, KNIGHT' S DEEP AND SIMMER EAST JOINT
PLANT 119
24. TUBE MILL, KNIGHT'S DEEP AND SIMMER EAST JOINT
PLANT . 121
25. FREEZING POINTS OF ALLOYS OF SILVER AND COPPER 132 .

26. FREEZING POINTS OF ALLOYS OF SILVER AND ZINC .134 .

27. FREEZING POINTS OF ALLOYS OF SILVER AND LEAD . . 136

28. FREEZING POINTS OF ALLOYS OF SILVER AND TIN . . 139
29. SILVER, 80 PER CENT. TIN, 20 PER CENT.
; X 24 . . 140
30. FREEZING POINTS OF ALLOYS OF SILVER AND PLATINUM . 141
31. SILVER, 94 PER CENT. ; ALUMINIUM, 6 PER CENT. X 30 . 142
32. AMALGAMATING PAN 162
33. ASSAY FURNACE 202
34. CUPEL AND MOULD 203
35. STAS PIPETTE 213
36. HAND PIPETTE FOR DECINORMAL SOLUTION, GAY LUSSAC
PROCESS 213
SHAKING MACHINE USED IN GAY LUSSAC PROCESS .214
37.

38.

39.
SILVER MELTING HOUSE, ROYAL MINT
ROLLING ROOM, ROYAL MINT
.... .

226
228
40. CUTTING ROOM, ROYAL MINT 229
41. CUTTING MACHINES, ROYAL MINT 230
42. SCISSEL 231
43. ANNEALING FURNACES FOR SILVER BLANKS, ROYAL MINT . 232
44. PRESS ROOM, ROYAL MINT 234
45. WEIGHING ROOM, ROYAL MINT 236
46. TELLING MACHINE, ROYAL MINT 237
THE PRECIOUS METALS
CHAPTEK I

HISTORY OF GOLD

IT is probable that gold was tbe first metal to attract the

attention of primitive man, although no traces of it have
been found in the dwellings and other relics of the Stone

Age. It must be noted, however, that flint daggers with gilt

handles have been found in Egypt. Whether gold was the
first metal observed or not, however, it is certain that its

lustre, bright colour, and toughness, and its wide distribution

in a native state as pieces of all sizes in loose sands and
gravels, must have led to the collection of gold in very

early times.
ProfessorGowland considers 1 that the order in which the
metals were discovered was not the same for every region,
but that the metals which occur as such in Nature must
have been the first known to the men inhabiting the
which they occurred. According to his view,
localities in

although gold cannot have escaped the observation of the

men of the Stone Age, it must generally have been found in
the form of fine particles which could not have been applied
to any useful purpose until after the invention of the art of

melting. Even melted lumps or naturally occurring nuggets

1
Presidential Address, Anthropological Institute, 1906.

r.M. B
2 THE PRECIOUS METALS

of gold would have been too soft to be of value as weapons

or implements, and must have been used only as personal
ornaments. Hence gold could have played only a very
small part, if
any, in the development of culture in Neolithic
times.
It was well known in Egypt some 5,500 years ago, and
had probably at that time already passed into use as a
standard of value. Thus in the code of Menes, B.C. 3600,
one part of gold was declared equal in value to two and a
half parts of silver.
The words originally used to denote gold seem to have
been connected with its glittering appearance, and indicate
the ideas associated with it in the minds of the early
observers. Such words are the Egyptian nub, the Hebrew
zaliab, and the Sanscrit, 7 ra/ita and liiranya. In later times
the alchemists called gold sol, and denoted it by the same

symbol as the sun, not on account of its appearance,

,

but because it was the most perfect of the noble metals, the

King of Metals. Silver was lima, the moon, and the other
known metals were named after the planets.
The ancient methods of obtaining gold in Egypt are
illustrated in rock carvings which date back to 2500 B.C.

The auriferous gravels were washed with water in stone

basins, and the gold melted in little furnaces with the aid
of mouth blowpipes. The earliest furnaces being adapted
from camp fires were urged no doubt by the wind only, and
the introduction of artificial bellows marked a great advance
in metallurgical skill. An ancient Egyptian furnace closely

resembling a camp fire, but fitted with bellows, is shown in

Fig. 1. This is from a wall painting at Thebes and is

reproduced from the Journal of the Anthropological Institute,

 HISTORY OF GOLD ;i

Vol. XXXVI., 1906, by kind permission of Professor Gowland

and of the
Koyal Anthropological Institute.
Sloping tables of stone were also used, as well as basins,

for washing gold ores, and sheepskins spread on the tables

or on sloping rocks for entangling the particles of gold and

helping in its collection were employed at an early date.
The Golden Fleece by Jason in B.C.
seized 1200 doubtless
typified the gold obtained from the rivers by its use.

FIG. 1. Ancient Egyptian Furnace.

The introduction of ore-crushing to obtain gold contained

in solid rock took place at some unknown time, but it was

well established in Egypt in B.C. 59, when it was described
1
by Diodorus the Sicilian as follows :

"
The parts Egypt near Ethiopia and Arabia are rich
of

in gold mines. The ore is a black mineral, marked

. . .

with white veins and shining specks. The chiefs of the

undertaking employ a great number of workmen who are
all either condemned criminals or prisoners of war. Even
1
Diodorus Siculus, Bk.-IV., c. 11 and 12.

B 2
4 THE PRECIOUS METALS

the parents of the condemned are summoned when the

number of workmen is insufficient. They work day and
night without rest, and under the surveillance of barbarous
soldiers, who speak languages different from those of the

miners, so that they cannot be won over either by promises

or by prayers.
"
The man who recognises the veins of gold is placed at
the head of the workmen, and points out to them the place
to excavate. The rocks are broken by iron wedges. The
miners follow, in their labours, the direction of the metallic

threads, and are assisted by the light of lamps in the

subterranean darkness. The stones are carried outside,
and are there crushed and reduced to small fragments.
"
The workmen never cease from their toil; they are
forced incessantly to the work by bad treatment and by
blows of the whip. Even children are not spared some ;

are set to carry the blocks of stone, others to break them

into fragments. The fragments are taken by older work-
men, of over thirty years of age, and crushed in iron mortars.
The fragments thus crushed are then ground in mills with
arms, which are turned by women and aged men. Two or
three people work at each mill. It is impossible to describe

the sufferings of these unhappy ones exposed naked to cold

;

and rain, they are allowed no repose there is no feeling

;

of pity, either for a weakly woman or for an old man on

the verge of the tomb no regard to the sick, who may
;

be a prey to the shivering of fever they are all struck

;

indiscriminately with repeated blows until they die of their

sufferings on the very spot where they have worked."
When the ore had been reduced to powder, it was spread

on wide, slightly inclined tables, and a stream of water

 HISTORY OF GOLD 5

flowing over the tables carried off the earthy materials, and
left the gold separated by This operation was
its weight.
repeated by the workmen several times they rubbed the ;

powdered ore with their hands for some time, then they
wiped it with little sponges in order to remove the impurities
which water alone could not carry off. By this means the
gold dust became clean and shining.
Diodorus adds that these methods were very ancient and
had been discovered by the first Kings of Egypt.
Hollowed-out stone mortars and also stone grinding-mills
have been found in many parts of the world besides Egypt,
in proximity to ancient gold workings, and the methods
described above were probably carried by the Komans to
some of these countries. The use of sieves, which were
originally made of hair and fibres, was also probably
introduced by the Egyptians, but heavy pestles or stampers,

requiring the efforts of several men to raise them and let

them fall, had their origin in later times.
The gold produced by these simple mechanical methods
was naturally very impure, containing varying proportions
of silver and other metals. When the silver was in
sufficient proportion (20 to 35 per cent.) to affect the

colour perceptibly, the alloy was called electrum. This

natural mixture of gold and silver was used for some of
the earliest gold coinages, but the removal of silver by
cementation was begun in very ancient times. Gold,
electrum and silver wererecognised and mentioned by
all

Homer, and Dr. Schliemann found all three among the

ruins of Troy.
The purification of gold was effected,according to
Agatharcides, by enclosing it in an earthenware vessel with
6 THE PKECIOUS METALS

lead (proportionate in amount to the quantity of gold), salt,

a little tin and some barley husks. The vessel was luted
up and heated in a furnace for five days and five nights,
and was then allowed to cool and opened. The gold, though
it had lost a little of its
weight, was very pure, and retained
neither lead nor tin. Such gold was called aurum obryzum,
and the process itself was called obrussa. 1 It seems to

have been a combination of cupellation and cementation.

The lead and the base metals in gold would be
the

eventually oxidised and absorbed by the pot, and the silver

chloride, formed by the action of the common salt, would
also fuse and pass into the pot. The use of the tin is
obscure, but the bran would prevent the lead from being
oxidised and absorbed too rapidly. This process was well
known to the Romans and is mentioned by Cicero,
Suetonius, and others.
Strabo describes the use of cementation in Spain, and
Pliny gives a still more exact description. The process
differed from obrussa in the omission of leadand the
addition of misy or sulphate of iron. The removal of silver
from the gold was more perfect by this method.
The Middle Ages were characterised more by the work of
the alchemists than by any great progress in the metallurgy
"
of gold. Chemia " was defined by Suidas as the art of
making gold and silver. It had its origin in Egypt at the
beginning of our era, and great powers were attributed to
the Egyptian priests, who carefully guarded their reputed
secrets and made their experiments a part of the mystic
rites of their religion.

The history of alchemy affords very curious reading.

1
"Histoire de la Chiinie," F. Hoefer, Vol. L, p. 116.
 HISTORY OF GOLD 7

The alchemists were familiar with a number of chemical

changes which seemed to them no less remarkable than the
transmutation of mercury or copper into gold would be.
For example, they knew that the addition of antimony to

copper produced a beautiful violet-coloured substance, and

that the incineration of lead yielded a bright yellow body.
" "
Why, they asked, should not the projection of some sub-
stance into the base metals give them the yellow colour and
the other properties of gold ? They sacrificed everything to
experiments with this end in view and repeatedly announced
that they had discovered the secret. Nevertheless no one
seems to have prepared the transmuting agent for himself,

but to have received it from a mysterious stranger. The

credulity and desire for gold of the rich and powerful
kept the alchemists supplied with funds to continue their
researches, their patrons including many reigning mon-
archs. Alchemy continued to flourish throughout Europe

for over a thousand years, and the claims of the gold-makers

were not generally discredited until about the end of the

seventeenth century.
Few improvements were introduced in the extraction of
gold from its ores for several hundreds of years previous to
the beginning of the sixteenth century. In 1519 stamp
batteries for crushing ore were introduced at Joachimsthal,

and screens set at the outlet of the mortars were described

l
by Agricola in 1556 and were in use in the Hartz in 1767.
A stamp mill for dry crushing in use in the sixteenth
century is shown in Fig. 2, which is taken from Agricola,
"De Re Metallica," Lib. VIII. Here A is a screen frame,
and B a screen which consists of iron wire and is fastened
1 " "
Voyages Metallurgiques, Jars, Vol. II., p. 309.
 THE PKECIOUS METALS

to the frame by iron rods D. The screen E is used for

separating the fine material from the coarse ore which is

FIG. 2. German Stamp Mill of the sixteenth century.

fed to the stamps. The crushed material is sieved at F.

The stamps are raised by water power.
The use of separating gold from other
mercury for
materials was described by Pliny and was probably never
 HISTORY OF GOLD 9

forgotten. Thus Geber, who died in 777, was aware that

mercury would dissolve gold and silver but not earthy
materials, and Theophilus in the eleventh century described
the extraction of gold from the sands of the Rhine by means
of quicksilver. In the Tyrol the method of gold ore treat-
ment included crushing in stamp batteries and the stirring
of the crushed ore in circular bowls with large quantities
of mercury. A stream of water passing through the bowl
carried the tailings, leaving the gold to sink to the
off

bottom of the bowl and unite with the bath of mercury

kept there. This process is of great antiquity, and its
principle is stillused in the modern amalgamating
machines of Hungary.
The catching of amalgam by means of copper plates is
more modern, and was probably suggested by the Cazo
process, which was introduced in Peru in 1609. l No
mention is made of
charging mercury stamp into batteries
and catching the gold amalgam on copper plates before the
year 1850. The chlorination process of gold ore treatment
originated in 1848, and the cyanide process in 1887.
As for the parting of gold from silver, the cementa-
tion process continued in use by the side of some other
furnace processes until it was displaced by the use of nitric
acid, which dissolves silver and other metals and leaves

gold unattacked. The process was described by Albertus

Magnus in the thirteenth century, but was not used on a
200 years later in Venice. It was
large scale until about
introduced into Paris in 1514, and has become obsolete

only within the last hundred years. The sulphuric acid

method was established in 1802 in Paris, and has flourished
1 " Arte de los Metales," Lib. III., cap. xxi.
Barba,
10 THE PEECIOUS METALS

for about a century. It is now being superseded by the

electrolytic process, which has already replaced it

throughout
America. In Australia the chlorine process of refining
and parting, introduced in 1867, is the only one that has
ever been used, owing to the high price of acids on that
continent.
 CHAPTER II

THE PROPERTIES OF GOLD

PHYSICAL PROPERTIES.

GOLD has a characteristic yellow colour, the so-called

golden yellow. it contains silver the colour is paler

If

yellow, and if itcontains copper the colour is a more

reddish-yellow. The colour of pure gold is seldom seen,

although it is closely imitated by certain mixtures of gold,
silver, and copper, used for wares of 18- and 22-carat gold.
Some specimens of gold-leaf consist of nearly pure gold.
Finely divided gold is often red or purple, as may be seen
in purple of Cassius (q.v.) and in some ruby glass, although
most specimens of ruby glass are now coloured by oxide
ofcopper instead of by metallic gold. Gold-leaf is green

by transmitted light. Gold precipitated from solution

varies in colour from black to bright red (see Faraday's
gold). Volatilised gold condenses as a reddish-purple stain.
Molten gold is green.
Gold is softer than silver, but harder than pure tin. It

is the most malleable and most ductile of metals. Leaf

gold can be made only one three-hundred-thousandth of an
inch thick by hammering, and an ounce of gold can be
drawn out into fifty miles of wire. The tenacity of gold
isseven tons per square inch, or less than that of silver or
copper. A wire of O'l inch in diameter is broken by a
12 THE PEECIOUS METALS

weight of 123 Ibs. Gold has an elongation of 30'8 per

cent, before rupture.
Gold crystallises in the cubic system, but large, well-
formed crystals are only occasionally found in Nature, and
are difficult to produce artificially. There are some large
cubic crystals of pure gold preserved in the Mint, which
remained in a crucible when the molten charge was
poured out.
The density of cast gold is about 19*3, and this can be
raised by rolling to 19 '48, and by hammering to 19'65.
The density of precipitated gold is usually somewhat higher,
and one specimen was found to be as high as 20'72. The
melting point of gold is 1,064 C., and it begins to volatilise
at about the same temperature. The amount of volatilisa-
tion, however, remains small at the temperatures attained
in ordinary industrial furnaces. Gold boils readily in
electric-arcfurnaces, and its
boiling point has been
estimated by Krafft and Bergfeld to be about 2,530 C.
under atmospheric pressure.
Its electrical conductivity at ordinary temperatures is
76*7, that of silver being 100. Its conductivity for heat

is also less than that of silver and of copper. It is non-

magnetic.

CHEMICAL PROPERTIES.

Originally gold was called by the alchemists a noble

metal, because it was not affected by fire. When put into
the fire, emerged unchanged, whereas
and even melted, it.
" " "
the base metals were oxidised and "destroyed (i.e., they
lost their metallic character) under the same conditions.

In the cupellation process the furnace was observed to

 THE PROPERTIES OF GOLD 13

remove the base metals from admixture with gold, leaving

the noble metal unscathed. The other noble metal recog-
nised by the alchemists was silver, but it was deemed less

perfect in its nobility than gold, because it could be dissolved

in aqua fortis, by which gold is unaffected. Hence gold was
called the more noble and silver the less noble or less perfect
metal. This identification of physical with moral qualities
still has its effect on our nomenclature.
The symbol of gold is Au (from aurum), and its atomic
weight is 197'2, taking that of oxygen as 16. In its solution
in mercury it is apparently monatomic. It bears a close
relation to the heavy platinum metals, osmium, iridium,
and platinum. Silver is similarly related to ruthenium,

rhodium, and palladium, the light platinum metals. Thus

the atomic weights of gold and the heavy platinum metals
are not very different ;
their densities are all very high,
and they all form lower oxides which are feebly basic, and
higher oxides which are both acidic and basic in a slight
degree.
At the same time gold is also related to silver and copper,
compounds of the
type RX existing in each case. The
substances aurous chloride, AuCl, silver chloride, AgCl,
and cuprous chloride, CuCl, resemble each other in several
respects. They are insoluble in water (although AuCl is

decomposed by it), but are soluble in potassium cyanide,

hydrochloric acid, and ammonia. Gold differs from silver

and copper in the extreme facility with which its compounds
are decomposed and the metal isolated. At a gentle heat
chlorine and oxygen are expelled from the chlorides and
oxides of gold, and metallic gold alone remains. These
compounds, therefore, act as oxidising agents.
14 THE PRECIOUS METALS

Gold is not appreciably attacked at any temperature by

water or by air, and consequently it does not become
tarnished. Molten gold gradually absorbs oxygen from
the atmosphere, and retains at least part of the
oxygen According to Neumann, finely
after solidification.

divided gold absorbs from 32 to 48 volumes of oxygen

at 450, or 0'24 to 0'36 per cent, by weight. Gold also
occludes hydrogen and carbon monoxide under suitable
conditions.
Gold not perceptibly attacked by alkalies or by hydro-
is

chloric, nitric, or sulphuric acid. Finely divided gold is

slightly soluble in boiling hydrochloric acid and in boiling
nitric acid. the ordinary temperature
It dissolves readily at

in water containing chlorine, bromine, or a mixture of iodine

and potassium iodide. It also dissolves readily in a boiling
concentrated solution of ferric chloride, and in any mixture
producing nascent chlorine, bromine, or iodine. It is slowly
dissolved by a solutionof potassium cyanide at the ordinary

temperature. This action depends on the presence of oxygen

or an oxidising agent. The most rapid solvent for gold is
hot aqua regia (a mixture of three parts of hydrochloric
acid with one part of nitric acid). Another convenient
method of dissolving it is to pass a current of electricity
through a solution of an alkaline cyanide, using an anode
1
of gold.

The compounds of gold are not numerous and have not

been fully studied. One of the chief chemical characteristics
with which its compounds are formed
of gold is the difficulty

and the ease with which they are decomposed.

1
For further information on the solvents of gold see "The Metal-
lurgy of Gold," T. K. Eose, 5th edition, 1906, pp. 9 and 12.
 THE PKOPERTIES OF GOLD 15

IDENTIFICATION OF GOLD.

Gold in the metallic form can be readily identified by its

colour, its great density, and by the fact that it is not
attacked by hot nitric acid. For confirmation, the gold

may be dissolved in aqua regia, and the yellow liquid

evaporated to dryness with the occasional addition of
hydrochloric acid. The residue, which must not be heated
above 100, is redissolved in water, and sulphurous acid
gas passed through the liquid. A black or brown pre-
is

cipitate insoluble in nitric acid denotes the presence of gold.

Cupellation is also a good test of the presence of gold

(see chapter on Assaying). The lead remaining on the

cupel must be tested further, as it may consist of silver or
the platinum metals.
Gold in solution may be identified by acidifying with
hydrochloric acid and passing sulphuretted hydrogen or
sulphurous acid. If nitric acid or other oxidising agents
are present they must be destroyed before SH 2 or S0 2 is
passed through the solution. The precipitate with sul-
phuretted hydrogen is black if the solution is cold, brown if
it is boiling. The precipitate is soluble in ammonium
sulphide, and if it is cupelled or merely ignited it yields

metallic gold. Sulphurous acid gives a black precipitate

of metallic gold, insoluble in nitric acid, and yielding a
yellow bead when cupelled. In these points it differs from
tellurium, which also gives a black precipitate with sul-

phurous acid.

Other reactions gold in solution as chloride, in the

of

absence of nitric acid, are as follows :

A mixture of protochloride and perchloride of tin gives

a precipitate of purple of Cassius.
16 THE PKEGTOUS METALS

Ferrous salts yield a black or brown precipitate of metallic

gold. In very dilute solutions, the liquid becomes blue by
transmitted light and brown by reflected light.
Oxalic acid and alkaline oxalates precipitate metallic

gold, especially on the application of heat.

Ammonia gives a yellow precipitate of fulminating

gold.
Potash and soda give a precipitate of oxide of gold soluble

in excess. If a large quantity of hydrochloric acid is

present the precipitate is not formed.

In the absence of an excess of hydrochloric acid

potassium cyanide gives a yellow precipitate soluble in

excess.
Solutions containing aurocyanides of the alkalies and

thiosulphates of gold do not give the ordinary reactions of

gold, although sulphuretted hydrogen precipitates sulphide
of gold from thiosulphates. Solutions containing cyanides
or thiosulphates be tested for gold by heating them
may
with aqua regia, evaporating to dryness, taking up with
water and a little hydrochloric acid, and testing the solu-
tions in the ordinary way.

COLLOIDAL GOLD.

Colloidal gold suspended in water may be obtained in

various ways by precipitating dilute solutions of gold
chloride. Such gold was first prepared by Faraday, and
the ruby-coloured liquids containing it are known as
"
Faraday's gold." A good method of preparation is to

raise to boiling a solution containing from O'Ol to O'OOl

per cent, of gold chloride, made slightly alkaline by the

 THE PEOPEETIES OF GOLD 17

addition of a little magnesia, and then to add a few drops

of formaldehyde or turpentine, or a solution of acety-
oil of

lene, or ofphosphorus dissolved in carbon disulphide. The

red liquid may also be prepared cold, and in that case the
action is much more rapid if a drop of liquid containing
" "
colloidal gold is added to inoculate the solution.
The gold is in the form of minute particles, which are far

too small to settle in water under the influence of gravity,

so that the liquids remain unchanged even if left undis-
turbed for years. Nevertheless, the gold is not in solution,
as is shown by the fact that it can be removed from the

liquid by shaking with freshly precipitated barium sulphate

and other precipitates.
These particles of gold are rendered visible under the
microscope by means of a powerful ray of light thrown
from the side. The smallest particles that become visible
are rather less than one hundred-thousandth of a millimeter
in diameter, or about six times the diameter of a molecule,
so that they are aggregates of about 200 atoms of gold.

The gold is and the solutions decolorised by

collected

shaking with mercury. Larger aggregates of gold are

formed and the liquid gradually changed to a blue colour
by the action of acids, neutral salts, alcohol in excess, etc.

The blue colour in time changes to violet and then to black,

and the gold settles to the bottom.

The same red colour is seen in ruby glass containing
gold.
The goldpurples, such as the well-known purple of
Cassius, also owe their colour to finely divided metallic
" "
gold or a colloidal solution of gold. Purple of Cassius
is prepared by adding SnCl2 and SnCU to a solution of
P.M. c
18 THE PKECIOUS METALS

gold chloride. A purple precipitate of hydrated stannic

oxide is formed, mixed with finely divided metallic gold.
The same colour is seen when alloys of silver, tin and gold
are attacked by nitric acid. Eecently it has been found by
1
H. Moissan that when a gold-tin alloy is raised to its boil-

ing point in an electric furnace and the metallic vapours

allowed to escape into the air, the tin burns and a substance
is obtained with the properties of purple of Cassius, and

containing tin dioxide, calcium oxide and gold. Similar

purples are obtained by distilling mixtures of gold with
lime, alumina, magnesia, silica, etc. Miiller also prepared
in the wet way a number of purples,* containing magnesia,
lime, baryta and alumina instead of oxide of tin.

THE PEEPAKATION OF PURE GOLD.

" "
Parted gold, which contains little except a small propor-
tion of silver, is dissolved in nitro-hydrochloric acid, and
the excess of nitric acid is expelled by evaporation on a
waterbath with additional HC1. The product consists

chiefly of chloro-auric acid, HAuCl4. As soon as scales of

gold begin to appear on the surface of the black fused liquid

and a smell of chlorine is perceived, the. vessel is allowed to
cool, when the contents solidify. The gold chloride is then
dissolved in a little distilled water and poured into a larger
volume of water. A yellow solution is produced with a
white cloud of silver chloride suspended in it. This is

allowed to settle and the solution is siphoned, off, diluted

to about 150 c.c. for each gram of dissolved gold, and

precipitated by sulphur dioxide or oxalic acid. The former
1
Comptea Bendus, Vol. CXLI. (1905), p. 977.
 UNIVERSITY

THE PROPERTIES OF GOLD 19

is the more convenient reagent, as it acts at the ordi-

nary temperature. The brown precipitate of gold (largely

consisting of scales and plates, if oxalic acid is used) is
allowed to settle, washed at first by decantation, and then,
when the smell of 862 has been removed, by shaking with

pure cold distilled water in a large flask. After shaking

with many fresh additions of water, the flask is heated. As
soon as it can be boiled for twenty-four hours with distilled

water without showing a trace of hydrochloric acid in the

liquid (tested by a considerable excess of silver nitrate), the
gold is dried, melted in a plumbago or clay crucible and cast
in a blackleaded mould. No scum should appear on the
surface of the molten gold, a scum appears, it is
but if

advisable to clear it awa

by adding pinch of borax. The
a

ingot is scratch-brushed and cleaned in acid. It is rolled

down between The gold may some-
carefully cleaned rolls.
times be not quite pure, and in that case it can be
still

improved by passing it again through the whole process,

redissolving and reprecipitating it.

This method is used at the Eoyal Mint in the preparation

of fine gold trial plates. The most convenient amount of
gold to work on is from 10 to 20 oz. Larger amounts
than these must be subdivided, and smaller amounts may
give less satisfactory results owing to the introduction of
impurities in melting and casting.

c 2
 CHAPTER III

THE COMPOUNDS OF GOLD

THERE are two series ofsimple compounds of gold,

corresponding to the general formulae AuE and AuE 3 in
,

which E is a negative monovalent radicle. There are also

some complex compounds which do not exhibit the ordinary
reactions of gold. In these gold is regarded as forming
"
part of certain complex ions (see below under Cyanides of
Gold," p. 28).

GOLD CHLORIDES.
There are two well-known chlorides, AuCl and AuCla.
As their densities in the gaseous condition have not been
determined their molecular formulas are not certainly
known. The formulae given above are the simplest that
can be assigned to these compounds. They are formed
simultaneously by the action of chlorine on metallic gold at
moderate temperatures, and are split up into gold and
chlorine at slightly higher temperatures. Gold and chlorine
unite at the ordinary temperature in the presence of moisture,
and this property is taken advantage of in the extraction of
gold from its ores in the Plattner and other chlorination
processes. The rate of action increases as the temperature
rises to C., when the rapidity of absorption of
about 225
chlorine by gold in the presence of water vapour is at a
maximum. At this temperature a mixture of about 20 per
 THE COMPOUNDS OF GOLD 21

cent, of AuCl and 80 per cent, of AuCl 3 is formed in an

atmosphere of chlorine, approximately corresponding to
the empirical formulae Au 2 Cl 5 Meanwhile the opposing
.

tendency of the chlorides of gold to decompose into gold and

chlorine begins to be apparent at about 70 in an atmos-

phere of chlorine, and the rate of decomposition increasing

with the temperature becomes very rapid at about 300, so
that above this temperature, even in an atmosphere of

chlorine, gold chloride not formed to any great extent.

is

In the air, the decomposition of gold chloride takes place

at far lower temperatures. At 100, 25 per cent, of AuCla
is converted into AuCl in seven days. At 190, AuCla is
completely converted into AuCl in ten hours, and into
metallic gold in six days.
The importance in
volatilisation of chloride of gold is of
connection with the loss of gold in the roasting of ores with
salt. It was found by the author 1 that volatilisation
begins at about 180 and rapidly increases to a maximum
at about 290. It then falls off gradually until the

temperature reaches 800, when it again begins to increase

rapidly. The rate of volatilisation depends partly on the
vapour tension of gold chloride and partly on its pressure of
dissociation, which is at a maximum between 300 and 800.
The volatilisation between 500 and 800 (very dull red to
red heat) is small. The gold volatilises chiefly in the form
of AuCl 3 with which a little AuCl is mixed.
,

Aurous chloride, AuCl, can be best prepared by heating

AuCl 3 to 190 in air. It is a lemon-yellow hygroscopic

powder, insoluble in water and ether, but soluble in

potassium cyanide. It is decomposed by water at the
1
Chem. Soc. Jour., Vol. LXVII. (1895), p. 881,
22 THE PKECTOTJS METALS

ordinary temperature into gold, which remains as a fine

powder, and AuCl 3 which dissolves in the water.
, The
density of aurous chloride is 7'4. For its behaviour when
heated see above.
Auric chloride, AuCl 3 can be prepared by acting on
,

finely divided gold with chlorine, especially at about 200.

or by dissolving gold in aqua regia. In the latter case the
excess of acid is removed as far as possible by evaporation
or a water bath, and the reddish-black liquid is allowed
to crystallise. Some chloro-auric acid, HAuCLi, remains
mixed with the auric chloride. Pure auric chloride is

difficult to prepare. The product obtained by acting on

gold with chlorine at 200 to 300 always contains less
chlorine than corresponds to the formula AuCl 3 If the .

product is dissolved in water, the metallic gold separated by

decantation, and the solution evaporated to the crystallising
point, large orange crystals of the hydrate AuCl 3 .2H 2
separate. These can be partly dehydrated by stand-
ing on a porous tile over concentrated sulphuric acid for
some days.
Auric chloride is brilliant red to dark red in colour. It

crystallises in needles and leaflets. Its melting point is

288 and its density is 4'3. It is deliquescent and dissolves

in water readily with evolution of heat. It is also soluble

in ether and alcohol. Water dissolves about five times its

weight of AuCl 3 . The strongest solutions are dark red,
but weaker solutions are of a bright lemon-yellow colour.
These solutions are decomposed and gold precipitated by
many reducing substances, such as sulphurous acid, oxalic
acid, FeS0 4 FeCl 2 phosphorus, hypophosphites, nitrites
, ,

and nitrous acid, NO, N2 4? arsenious and antimonious

 THE COMPOUNDS OF GOLD 23

acids. The reaction in the case of precipitation by S(>2

may be expressed as follows :

2AuCl 3 + 3S0 + 6H 2 2 = 2Au + 6HC1 + 3H S0 2 4.

Pure hydrogen, unless nascent, does not appear to act

on chlorideof gold, but most of the metals decompose it

at once. Mercury forms a gold amalgam, lead gives

dendritic gold, and tin often yields a purple-coloured
precipitate resembling purple of Cassius as well as gold
in the form of brown powder. Many organic compounds
also precipitate gold, their action being assisted by light
and heat. Charcoal acts best on boiling. Solutions of

gold chloride leave a violet stain on the skin, which can be

removed by cyanide solutions. Cyanide of potassium, if
added to gold chloride in solution, produces at first a yellow
precipitate soluble in excess ;
if much free hydrochloric

acid is present, no precipitate is formed.

Sulphuretted hydrogen precipitates metallic gold at 100,
but at lower temperatures mixtures of gold, sulphur and
sulphides of gold are thrown down. Oxalic acid reduces

gold chloride very slowly at ordinary temperatures, but

rapidly at 100. The reaction may be expressed by the
following equation :

2AuCl 3 + 3C H 2 2 4 = 2Au + 6HC1 + 6C0 2.

The detection of minute quantities of gold chloride in

solution usually effected either by means of protochloride
is

of tin or of ferrous sulphate. Protochloride of tin, SnCl 2 ,

gives a brown precipitate of ill-defined composition with

concentrated solutions of auric chloride. A mixture of

SnCl 2 and SnCl 4 , however, prepared by digesting a piece
of metallic tin in a solution of SnCl 2 with the addition of
,
24 THE PBECIOUS METALS

a little hydrochloric acid, gives a purple precipitate of purple

of Cassius.

The test may be made a very delicate one. The best method
of carrying it out
heat the solution of gold chloride to
is to

boiling, and then to mix it suddenly with a few cubic centi-

metres of a solution of stannous chloride. This is most
easily effected by pouring the auric chloride solution into a
large beaker containing the stannous chloride. A yellowish-
white flocculent precipitate of tin hydrate is formed, and
this settles rapidly, and can be separated and transferred,

together with a few cubic centimetres of liquid, to a Nessler's

tube. In this way the author has detected the presence of
one part of gold in solution as chloride in 100,000,000 parts
of water (or one grain of gold in six tons of water). The
amount of solution used in this case was about three litres,
and the colour of the precipitate was a faint greyish-
violet, the difference between and stannous chloride pre-
it

cipitated by pure water being just visible. The presence

of one part of gold per million is signalised by a purplish-
red to blackish-purple colour.
The ferrous sulphate test is less delicate, but more easily
and quickly applied. In comparatively strong solutions a
brown precipitate of metallic gold is formed, and in weaker
solutions the colour of the solution becomes blue by trans-
mitted light. By comparing two Nessler tubes, one con-
taining distilled water and the other a solution of gold
chloride, the presence of one part of gold in about 720,000
parts of water can be detected by the addition of ferrous
sulphate.
Chloro-auric acid, HAuC^. When gold is dissolved in

aqua regia and evaporated with excess of hydrochloric acid

 THE COMPOUNDS OF GOLD 25

to the point of crystallisation, long yellow needles are

formed on cooling, having the composition HAuCl 4 .4H 2 0.
When heated these needles fuse and then decompose, yield-
ing hydrochloric acid, chlorine, and a mixture of aurous
and auric chlorides. The crystals are deliquescent and
readily soluble in water. The solutions are in general
reduced by the same agents as auric chloride.
A number of double salts of the same type are known
and are distinguished by the name chloro-aurates. They
are prepared by adding a slight excess of metallic chloride
to a solution of HAuCU,
evaporating to dryness, taking up
with pure water and crystallising out by evaporation in a
vacuum. They may be purified by redissolving and recrystal-
lising them. They are all soluble in water and alcohol, and
may be represented by the general formula, AuCl 3 .RCl.nH 2 0.
Potassium chloro-aurate crystallises in yellow needles,
having the composition 2KAuCl4.5H 2 0. Sodium chloro-
aurate, NaAuCl4.2H 2 0, which also forms yellow needles,
is used in photography in the toning bath. When pure it
contains 49*5 per cent, of gold.
Chloro-aurates of ammonium, barium, strontium, lithium,

magnesium (MgAu 2 Cl 8 .6H 2 0), calcium, manganese, nickel,

cobalt, zinc, cadmium, rubidium and caesium are also known
in crystalline form. They are all dehydrated at 100.
Double chlorides and phosphorus, gold and sulphur
of gold
and gold and selenium have also been prepared.

GOLD BROMIDES.
Two bromides, AuBr and AuBr3 have been
,
isolated cor-

responding to the chlorides. Their vapour density has

26 THE PBECIOUS METALS

not been determined, so that their molecular formulae are

unknown.
Aurous bromide, AuBr, is obtained by heating auric
bromide, or better still bromo-auric acid, to about 115. If

bromo-auric acid, HAuBr 4 .5H 2 0, is heated in a porcelain

basin, it melts and gives off HBr and water. After being

kept at 115 for some time, the colour becomes yellowish-

grey,and nearly pure AuBr remains. Thus prepared it is
a yellowish-grey amorphous powder, not deliquescent and
insoluble in water, but decomposed by it on boiling into

gold and auric bromide. Hydrobromic acid converts it

into metallic gold and bromo-auric acid. At temperatures
above 115 it is decomposed into gold and bromine.
Auric bromide, AuBr 3 is prepared by acting on finely
,

divided gold with bromine in the presence of water. The

action is slow and may be somewhat expedited by heat.
Gold is more readily soluble in a mixture of nitric and
hydrobromic acids, but in this case the nitric acid must be
subsequently expelled by evaporating at a gentle heat, if

auric bromide is to be prepared. In evaporating solutions

containing either bromide or chloride of gold, the bottom
only of the porcelain basin must be heated, while the sides
are kept cool. If this precaution is not taken, the salt

creeps up the sides and over the edge, and is found in part
on the outside of the vessel.
Auric bromide crystallises in black needles and in scarlet
plates. It also occurs as a brownish-black powder. It is

not deliquescent and is not very soluble in water, dissolving

in about 100 parts of water at the ordinary temperature.
Even highly dilute solutions are coloured red, and as the
proportion of auric bromide increases the solution darkens,
 THE COMPOUNDS OF GOLD 27

and concentrated solutions are nearly black. Auric bromide

is soluble in ether, and resembles auric cbloride in
many
of its properties. It is split up into gold and bromine at

about 140. Its solutions are reducible in the same way

as those of chloride of gold. Sulphur dioxide completely
decolorises a solution of auric bromide before any gold is

precipitated. The final result of the action may be repre-

sented by the following equation :

3S0 2 + 2AuBr + 6H = 3H
3 2 2 S0 4 + 2Au + 6HBr.
Auric bromide forms double bromides or bromo-aurates
with bromides of some other metals. The bromo-aurates
are of reddish-brown colour, and are similar in properties
to the chlor-aurates.

GOLD IODIDES.

Iodine combines with gold less readily than either

chlorine or bromine. Gold dissolves in a solution of

potassium iodide containing iodine, forming a solution of

potassium iodo-aurate, KAuI 4 This salt crystallises in
.

black, lustrous, quadrangular prisms which decompose on

heating. Auric iodide, AuI 3 ,
is obtained as a dark green

precipitate by adding AuCla little by little to a solution

of potassium iodide. Auric iodide is very unstable, and
breaks down into aurous iodide and iodine at the ordinary
temperature. It is decomposed by potash with precipita-
tion of metallic gold. Aurous iodide, Aul, is obtained as a
pale yellow precipitate by adding potassium iodide little by
little to a solution of auric chloride. It is decomposed into

gold and iodine at a temperature of about 120.

Some double iodides or iodo-aurates are also known, as,
for example, KAuI4 .
 28 THE PRECIOUS METALS

GOLD CYANIDES.
The only simple cyanide known is aurous cyanide, AuCy.
A number of double cyanides corresponding to a hypo-
thetical auric cyanide, AuCy3 ,
are also known, but no such

compound as AuCy3 has been isolated.

Aurous cyanide prepared by evaporating down a solu-
is

tion of potassium aurocyanide, KAuCy2, with hydrochloric

or nitric acid and washing it with water in the dark. The
residue when dried consists of yellow crystalline grains of
aurous cyanide. It is insoluble in water, alcohol and ether,
and is unaffected by exposure to air and light provided that
it has been well dried. It is not attacked by nitric, hydro-
chloric or sulphuric acid, but dissolves in aqua regia.
When heated, cyanogen and metallic gold
is given off

remains. It is ammonia, sodium thiosulphate,

soluble in

yellow ammonium sulphide, and in alkaline cyanides.

ammonium, potassium, sodium, barium,
Aurocyanides of
strontium, calcium, cadmium, zinc and cobalt have been
isolated. These compounds when in solution contain
"
the 'ion AuCy2, aurocyanidion," and not the simple ion
Au. They are colourless and do not show the ordinary
reactions of gold. When a solution of potassium auro-

cyanide, KAuCy2 ,
is electrolysed, it is believed that

potassium is set free at the cathode and the ion AuCyg

at the anode. The potassium
on water, liberating acts

hydrogen and forming potash, and also perhaps in part

replaces the gold in solution directly, precipitating the gold
at the cathode and forming potassium cyanide. The AuCy2
set free at the anode immediately breaks down into AuCy
and Cy. The AuCy is either precipitated in the anode
 THE COMPOUNDS OF GOLD 29

sludge or is redissolved in excess of KCy if that is present,

and the nascent cyanogen generally attacks the anode,
dissolving or oxidising it.

Aurocyanicle of potassium, KAuCy2, is formed by dissolving

AuCy, fulminating gold or Au 2 in a solution of cyanide of
potassium. It is also formed by passing a current of elec-

tricity through a solution of cyanide of potassium, using a

gold anode. Soon after this was discovered it was found
*
by Bagration that plates of gold are slowly dissolved by

cyanide solutions without the aid of an electric battery.

This is due to the presence of oxygen, as was pointed out
2
by Eisner. The amount of oxygen required was shown
3
by Maclaurin to be in accordance with the equation

4Au + 8KCy + 2 + 2H 2 = 4KAuCy + 4KOH. 2

4
According to Bodlander the action takes place in two
stages, the first stage being accompanied by the production
of hydrogen peroxide, which then acts as an oxidiser and
enables the cyanide to dissolve more gold, thus :

(1) 2Au + 4KCy + 2H 2 + 2 = 2KAuCy 2 + 2KOH + H2 2.

(2) H 2 2 + 4KCy + 2Au = 2KAuCy2 + 2KOH.

Oxygen can be replaced by many other oxidisers in this
reaction. Among these may be mentioned sodium dioxide,
barium dioxide, hydrogen peroxide, potassium permanga-
nate, potassium ferricyanide, chlorine, bromine, etc. The

1
Bull, de VAcard. des Sciences de St. Petersburg, Vol. II. (1843),
p. 136.
2
Erdm. Jour. Prak. Chem., Vol. XXXVII. (1846), p. 441.
3
Jour. Chem. Soc., Vol. LXIII. (1893), p. 724.
*
Zeit. fur Angewandte Chemie, 1896, p. 583.
30 THE PEECIOUS METALS

dissolution of gold is assisted in all cases by heat, if the

oxidiser is not destroyed by it too rapidly.

Aurocyanide of potassium is also formed without the

presence of an oxidiser by the action of the double cyanide
of mercury and potassium on metallic gold. In this case
the gold directly replaces the mercury in solution according
to the equation

HgCy2 .2KCy + 2Au = 2KAuCy + Hg. 2

The action is greatly expedited by boiling the solution.

Aurocyanide of potassium crystallises in colourless rhom-
boidal octahedra, which remain unchanged in air at the

ordinary temperature, but are decomposed by heat, cyanogen

being given off and a mixture of gold and potassium cyanide
remaining. The aurocyanide is soluble in seven parts of
cold water and in half a part of boiling water. When it is

heated with nitric, hydrochloric or sulphuric acid, hydro-

cyanic acid given off and aurous cyanide is precipitated.
is

On boiling with aqua regia all the cyanogen is removed

and gold chloride remains.
Ferrous salts are without action on solutions of KAuCy2,
but according to Lindbom oxalic acid and sulphur dioxide

precipitate aurous cyanide from hot solutions.

Gold is precipitated from solutions of potassium auro-
cyanide by the passage of a current of electricity or by the
action of certain metals, notably zinc, copper and aluminium.
The reaction is probably mainly one of simple displace-
ment, thus :

2KAuCy2 + Zn = K ZnCy + 2Au.

2 4

When cyanide of potassium is not present in excess, the

action stops almost at once, owing to the formation of a
 THE COMPOUNDS OF GOLD 31

layer of zinc hydrate zinc cyanide on the zinc, or

or
" "
perhaps to the polarising action of a layer of hydrogen.
When KCy is present in excess the zinc dissolves in it,
hydrogen is freely evolved, and potash is formed. In the
cyanide process, the zinc shavings used to precipitate the
gold tend to become coated with a " white precipitate,"
which consists of a mixture consisting mainly of zinc

hydrate, zinc cyanide, and ferrocyanide of zinc and

1
potassium.
Gold is also precipitated from solutions KAuCy2 by of

charcoal, and AuCy is precipitated by the addition of copper

sulphate after acidification with SCV Aurous cyanide is
also precipitated by neutralising a solution of KAuCy2 by
means of sulphuric acid and adding cuprous chloride.
Aurocyanide of potassium is used in gold baths for
electro-plating. The baths may be prepared by passing
an electric current through a solution of potassium cyanide,
using a large anode of gold and a small cathode.
Aurocyanides of ammonium, NH 4 AuCy 2 ,
and sodium,
NaAuCys, are similar in their properties to the potassium
salt. The ammonium salt is easily soluble in water, but
the sodium salt is only slightly soluble.
All these aurocyanides, when acted on by chlorine,
bromine, or iodine, yield compounds, of which KAuCy2.Cl2
+ 2H2 0, may be taken as a type. These compounds are
very soluble in water, and are decomposed by heat.
Auncyanides are salts of auricyanhydric acid, HAuCy4,
which can be prepared as follows A solution of AgNOa :

is added to a solution of auricyanide of potassium. A

1
Prister and Bay, Jour. Chem. Met. and Mmj. Soc. of S. A., Vol. V.
(1904), p. 75.
32 THE PRECIOUS METALS

precipitate of auricyanide of silver is obtained, and this

iswashed and then treated with a small quantity of HC1.

The liquid is then filtered, to remove the silver chloride,
and evaporated in a vacuum, when colourless tabular
crystals of HAuCy4 + 1JH 2 result.

Auricyanide of potassium, KAuCy 4 + 1JH 0,2 is formed

by adding a perfectly neutral solution of AuCl 3 to a warm
concentrated solution of KCy. On evaporating down, large
tabular colourless crystals of the auricyanide are formed.

They are easily soluble in hot water, and are less soluble
in cold water. Auricyanide of potassium begins to decom-
pose when heated to 200, but fuses at a red heat long

before its decomposition is complete.

Auricyanides of ammonium, silver and cobalt are also
known.
known only as double sulpho-
Sulpl10 cyanides of gold are
cyanides,such as aurous potassium sulphocyanide, KAuS2Cy2,
formed by adding AuCl 3 to KSCy at 80. This salt forms
insoluble precipitates with many metallic salts. Seleno-

cyanates of gold are also known, but cyanate of gold has

not been prepared.

OXIDES OF GOLD.

Three oxides have been isolated, Au 2 0, AuO and Au 2 3.

These are obtained by precipitating the haloid compounds.

They are not readily formed by the action of oxidising
agents on metallic gold, and are all reduced at a moderate
temperature by heat alone, oxygen being given off and
metallic gold left as a residue. Auric oxide, Au 2 03, is
obtained by the action of potash or magnesia on a boiling
solution of gold chloride. The orange-coloured precipitate
 THE COMPOUNDS OF GOLD 33

of hydrate of gold, Au0 3 H 3 dehydrated by drying at

,
is

about 100. At about 110 oxygen is given off, and at 160

the dark chestnut auro-auric oxide AuO is produced, soluble
in hydrochloric acid. Aurous oxide, Au 2 0, is produced by
the action of potash on AuBr. It is a violet-coloured

powder, soluble in cold water and in alkalies, and decom-

posed by hydrochloric acid. It breaks down into gold and
oxygen at about 250. Nevertheless, at 450 gold absorbs
and retains from 0*24 to 0'36 per cent, of its weight of
oxygen, an amount which corresponds to from 6 to 9 per

cent, of Au 2 0. In cupellation (q.v.) it is possible that some

aurous oxide is produced and absorbed as such by the cupel.
Auric oxide dissolves in hydrochloric acid, forming
HAuCLi, and it is also soluble in sulphuric and nitric acids.

On dilution, hydrated auric oxide is reprecipitated, but

by
cooling the solution in nitric acid by means of a freezing
mixture a nitrate of gold, Au(N03)3.HN0 3 .3H 2 may be
crystallised out. Schottlander also obtained sulphate of
gold from the nitrate. When metallic gold dissolves in
nitric acid the nitrate is probably directly produced.
Several crystallisable double nitrates have also been pre-

pared.
Auric oxide also dissolves in alkalies, forming aurates,
whence auric oxide is sometimes called auric acid.

Potassium aurate, KAu02.3H 2 0, has been obtained in

yellow crystals, and from its aqueous solutions various
insol ble aurates have been obtained by precipitation with
metallic salts. The aurates are
readily reduced by heat,
or by organic compounds with the production of metallic
gold.
When a solution of gold chloride is precipitated by
P.M.
34 THE PRECIOUS METALS

ia,, fulminating gold is produced, but this

perhaps is

better regarded as an amine than an aurate. Easchig

states that if prepared in this way it is a mixture of
auric diamine, (AuN.NH 3 2 .3H 2 0, with
) auric imidochloride,
NH.AuCl. It is also produced by the action of ammonia
on moist auric oxide. It is of a yellow colour, is soluble in

potassium cyanide and in hydrochloric acid, and detonates

when struck or heated to 143. Sometimes it explodes
spontaneously at ordinary temperatures. The products of

itsdecomposition are gold, ammonia, nitrogen and water.

Other fulminates are obtained by the action of ammonia
on aurous oxide, Au 2 0.
The aurates also give rise to double sulphites of gold and
the alkalies (or auro sulphites), when alkaline aurates are
acted on by means of alkaline sulphites. The ammonium,
sodium, potassium, and barium salts are known.
Thiosulphate of gold and sodium, Au2S203.3Na2S 2 03.4H 2 0,
crystallises in fine colourless needles, readily soluble in
water. Its solutions are sweet to the taste. It is pre-
pared by adding neutral chloride of gold to thiosulphate
of soda, and precipitating the salt with alcohol, in which

it very slightly soluble.

is It was formerly used, under
"
the name Sel de Fordos et Gelis," for intensifying the

images of
daguerreotype pictures, gold being deposited
on the silver amalgam. The salt still plays an important
" "
part in photography in toning silver prints, as it exists
in combined toning and fixing baths. The solution of the
salt exhibits neither the characteristic reactions of gold nor
those of thiosulphates. Neither ferrous sulphate, stannous
chloride nor oxalic acid precipitates gold from thiosulphate

solutions, and sulphuric or hydrochloric acid does not at

 THE COMPOUNDS OF GOLD 35

once precipitate sulphur or disengage sulphurous acid.

On heating for some time with these acids, or on passing
sulphuretted hydrogen through the solution it yields
sulphide of gold.
Silicates of gold areprepared by fusion of an aurate or of
gold chloride mixed with an oxidiser, such as antimony
oxide or nitre, with a large excess of a glass rich in silica.
The result is a clear yellow glass containing about 0'2 per
cent, of gold.

SULPHIDES OF GOLD.

Two sulphides are known, Au 2 S and AuS, corresponding

to the lower oxides, and a number of double sulphides or

thio-salts corresponding to the aurateshave also been pre-

pared. When H 2 S is passed into a cold solution of AuCla a
black precipitate of AuS mixed with sulphur is obtained, the
action being partly represented by the following equation :

8AuCl 3 + 9H S + 4H = 8AuS + 24HC1 + H S0

2 2 2 4.

If the solution is heated variable mixtures of sulphur,

metallic gold and gold sulphides are obtained. Sulphide
of gold begins to decompose at 140, and loses all its sul-

phur at about 270. It is unattacked by mineral acids

except aqua regia, but is soluble in alkaline sulphides, form-

ing thio-salts which are decomposed by mineral acids.

Gold sclcnide and gold telhiride are similarly obtained by
precipitation from a solution of AuCla by means of H 2 Se
and H 2 Te.
Phosphide of gold is a grey, readily fusible substance
formed by gently heating gold in phosphorus vapour or by
passing phosphoretted hydrogen into a solution of gold
chloride. It burns when heated in air to about 115.

D 2
36 THE PKECIOUS METALS

Arsenide of gold a similar grey, brittle body obtained by

is

heating together gold and arsenic or by acting on gold with

the vapour of arsenic. On being heated in air most of the
arsenic is expelled. On grinding arsenide of gold with
mercury, the gold is amalgamated and the arsenic expelled,
but the mercury and amalgam become subdivided into a
number of little globules coated with a greyish powder, and
it is difficult to collect them again.
Antiinonide of gold is similar. See also pp. 50 and 51.
 CHAPTER IV

THE ALLOYS OF GOLD

GOLD AND SILVER.

GOLD and silver unite in all proportions when melted
together, forming homogeneous alloys. On solidification

no separation of the constituents takes place, and the

microscopic structure resembles that of pure gold or pure
silver. If the mixture is kept melted for some time without

stirring, a partial separation of the two metals takes place

under the action of gravity, gold settling to the bottom.
The colour of gold becomes paler when small quantities
of silver are added to it, and is white with a scarcely

perceptible tinge of yellow when 50 per cent, of silver is

present. Alloys containing more than 60 per cent, of silver

are silver-white. Gold containing small quantities of silver

(10 to 20 per cent.) are stated by some observers to be of a

green colour (" green gold ").

The melting point of alloys containing less than 5 per

cent, of silver is the same as that of pure gold, and further

moderate additions lower the melting point only

of silver

slightly. The alloy containing 35 per cent, of silver melts at

1,061, or only three degrees below the melting point of gold.
As the percentage of silver increases the melting point falls
more rapidly, but if there is only a trace of gold in silver,
its melting point is still above that of pure silver, 1 see Fig. 3.
1
Koberts- Austen and Eose, Proc. Roy. Sac., Vol. LXXI. (1903), p. 161 .
38 THE PRECIOUS METALS

All the alloys are soft, malleable and ductile, being

intermediate in these properties between pure gold and pure
silver. The densities of the alloys are also intermediate
between those of the metals, and show no marked expansion
or contraction due to the mixing of the metals, although
Hoitsema 1 found evidence of slight contraction of about

1100

950

900
10 20 30 4-0 50 GO 70 80 100

Atoms oF Silver per cent

FIG. 3. Curve showing Freezing Points of the Alloys of

Gold and Silver.

0-4 per cent, insome of the alloys. At 15 the densities of

unworked cast alloys are as shown in the table on p. 39.
The electric conductivity of alloys of gold and silver was
found by Matthiessen 2 to be below that of either gold or

silver. The electric conductivity of silver being taken at

100, that of gold is 78, and that of the 50 per cent, alloy
only 15.
There no chemical compound of gold and silver known
is

to be formed, but they appear to form isomorphous mixtures,

Zeitschrift fur Anvrgan. Cliem., Vol. XLI. (1904), p. 66.

1

2
Chem. Soc. Jour., Vol. IV. (1867), p. 201.
 THE ALLOYS OF GOLD 39

the crystalline structure of all the alloys being precisely

similar.
The behaviour of these alloys when treated with nitric or

sulphuric acid is of interest, as upon it are based the usual

methods of refining gold. Alloys containing at least 60 per
"
cent, of silver are completely parted" by boiling in nitric

acid, the silver being dissolved and the gold left behind in
a brown porous state. The gold obstinately retains about
TABLE OF DENSITIES OF THE ALLOYS OF GOLD AND SILVER.
Gold. Density.
100
91-6
per cent
....
....
19*31
18-04
79-8
78-4 ....
....
17'54
16-35
75-0
66-7 ....
....
16-03
15-07
50-0
33-3 ....
....
13-60
12-88
25-0
16-7 ....
....
11-78
11-28
10-47

O'l per .cent, of its weight in silver, and if boiling is

continued in strong acid, gold dissolves whilst the amount
of silver is not appreciably reduced. If the alloy originally

contains less than 60 per cent, of silver, the parting is not

complete, but the alloy of gold 1 to silver 1*25, containing

55*5 per cent, of silver, yields gold containing only 0'3 per
cent, of silver. With equal gold and silver,
parts of

however, nitric acid leaves half the silver in the gold, and
with smaller proportions of silver the acid has less and less
40 THE PEECIOUS METALS

effect. Copper or other base metals can replace silver

without materially affecting the results.

Alloys of gold and silver containing over 50 per cent, of
gold are difficult to dissolve in acid, and should be melted
with more silver if it is required to part them. Nitric acid
has little effect on them, and nitro-hydrochloric acid, which
dissolves the gold, converts the silver into insoluble chloride,
which protects the alloy from further attack. The alternate
action of nitro-hydrochloric acid and ammonia eventually
results in the dissolution of all the gold, but the method is

a tedious one. A
mixture of aqua regia and common salt is
also sometimes used in the dissolution of these alloys, the
silver chloride being dissolved by the salt. A large volume
of liquid must be used, asonlyO'127 gram AgCl can be held
by 100 c.c. of a saturated solution of common salt.
in solution
and silver have been used from very early
Alloys of gold
times in jewellery and for coins. Phidon, King of Argos,
struck coins of electrum in ^gina about the year 720 B.C.

Electrum includes pale coloured alloys, containing from

15 to 35 per cent, of silver. A little later pure gold and

electrum were being coined side by side in Lydia, with

different money values, the electrum being taken as worth
25 per cent, less than pure gold. 1 The addition of copper
to harden both gold and electrum was made in Eoman

coins some time before the Christian era.

GOLD AND COPPEB. 2

These metals are miscible in proportions when molten,
all

and on solidification, separate only to a slight degree. The

1
Lenormant, "La Monnaie dans 1'Antiquite," Vol. I., p. 194.
2
Eoberts-Austen and Eose, Proc. Roy. Soc., Yol. LXVII. (1900),
p. 105.
 THE ALLOYS OF GOLD 41

firstadditions of copper to gold cause a rapid lowering of the

melting point. Standard gold, containing gold 91*6 per cent.,

copper 8*3 per cent., solidifies at 951, or 103 below the
melting point of pure gold, and gold of the French standard
coinage alloy containing 90 per cent, of gold, solidifies at
946. The alloy containing 82 per cent, of gold and 18 per
cent, of copper has the lowest melting point, viz., 905, and
this alloy may be a true eutectic. The melting point curve
rises continuously with the increase of copper from the
eutectic alloy to pure copper, which melts at 1,083. The
l
densities of the alloys are as follows :

Gold. Density.
100 per cent 19*31
91*7
90-0 ,. ....
....
. 17-35
17-17
83-3
75-0 ., ....
...
15-86
14-74
58-3
25-0 ....
...
12-69
10-03
8-7

These densities refer, however, to cast specimens. When

hardened by rolling or hammering the densities of the
alloys become greater. Thus the usual density of an
English sovereign (containing 91*66 per cent, of gold) is
about 17 '48, but these coins for the most part contain
a small and variable percentage of silver. From the
densities it is evident that gold and copper contract

slightly (from 0'5 to 1*0 per cent.) on being alloyed.

1
Eoberts- Austen, Ann. Chim. Plnjs., [5] XIII. (1878), p. 118;
Hoitsema, Zeit. An. Chem., Vol. XLI. (1904), p. Go.
42 THE PKECIOUS METALS

The colour of gold becomes redder when copper is added

to it, and if silver is added at the same time, the two metals
counteract the effect of each other, so that triple alloys can be
formed having a colour closely resembling that of pure gold.
The and copper is
electric conductivity of alloys of gold
1
less of either gold or copper, and is even less
than that
than that of the silver-gold alloys (see above). The chemical
compound and copper is known to
of gold exist, and the
alloys in general show no marked change in properties
when the proportions of the metals are slightly varied.

Copper hardens gold and decreases its malleability, these

effects continually increasing until the eutectic alloy is

reached. This alloy (containing 18 per cent, of copper)

is distinctly brittle,
breaking under the hammer with a
conchoidal fracture, and showing an elongation before
rupture of 3'3 per cent., that of pure gold being 3O8 per
cent., and that of pure gold alloyed with 0*24 per cent, of
lead being 4*9 per cent. The tensile strength of pure gold
is 7 tons per square inch, that of standard gold (containing
8*3 per cent, of copper) 16 tons, and that of the eutectic

alloy 7*87 tons per square inch.

2
According to Kurnakoff and Schemtchuschny the two
metals are isomorphous, and form a continuous series of
mixed crystals. In Fig. 4, which gives the results of their

work, the curve of electric conductivity of the alloys is shown

in the line PQE. The line ALMLiB is the liquidus
curve connecting the points at which solidification begins,
and ASMSiB is the solidus curve, denoting the completion
of solidification. The minimum M, corresponding to the
1
Matthiessen, Chem, Soc. Jour., Vol. IV. (1867), p. 201.
/Soc., Vol. XXXIX. (1907), p. 211.
2
Jour. Euss. Phys. Chem.
 THE ALLOYS OF GOLD 43

alloy containing about 40*5 atoms per cent, of copper, is

coincident with the eutectic observed by Eoberts-Austen
and Rose. At this point the two curves meet, and there
is no pasty stage. When copper is in excess, that is, when

<
700

10 20 30 40 50 60 70 80 90 100%

Percentage of Gold in Atoms.

FIG. 4. Curves showing Freezing Points and Electric

Conductivity of the Alloys of Gold and Copper.

more than 18 per cent, by weight or 40*5 atoms per cent,

of copper is present, the part first solidified is rich in copper,

forming crystallites, which are darkened by a mixture of

chloride of iron and hydrochloric acid. The peripheral
layers of the crystallites and the matrix are richer in gold
 THE PRECIOUS METALS

and remain unattacked : see Fig. 5, which is a photo-

micrograph, after
Kurnakoff, of
alloy containing the
25 per cent, gold magnified 90 diameters.
The gold-copper alloys are used extensively for coinage
and jewellery, pure gold and the gold-silver alloys being
too soft, so that they soon become defaced by wear. In
jewellery some silver is usually added to the copper and
gold (see Chapter XVI.).
When alloys of gold and

copper are heated in the

air they blacken from

the formation of oxide

of copper, but the black
scale can be removed by
dissolving in hot dilute
sulphuric acid.

GOLD AND MERCURY.

These alloys are
FIG. 5. Gold, 25 per cent. Copper,
;
known as amalgams, and
75 per cent. X 90. -, ,,, ,
are formed, although
with difficulty, by the direct union of the two metals
1
at the ordinary temperature. According to Kasentsof,
mercury dissolves O'll per cent, of gold at 0, 0*126 per
cent, at 20 and 0'65 per cent, at 100. The mixtures con-
taining less than these quantities of gold are the alloys, liquid
at ordinary temperatures, silver-white like mercury, and
passing through cloth or other filtering material unchanged.
On the other hand, gold will absorb about six times its

1
Bull. Soc. Chem., [2] Vol. XXV., p. 20.
 THE ALLOYS OF GOLD 45

weight of mercury, forming a silver- white solid alloy. If

this solid amalgam, which contains about 13*5 per cent, of

gold, is mixed with more mercury, the result is not a
homogeneous body, but a saturated solution of gold in
mercury, in which solid particles of gold amalgam are
suspended. These solid particles are heavier than mercury
and settle to the bottom. They can be separated from
mercury by filtration under pressure, but usually retain a
small excess of mercury.
The amalgams recovered in gold-mills, in which ores are
treated by the amalgamation process, are not true alloys.
Before straining they consist of mercury containing a
number of little "nuggets" of gold into which mercury
has penetrated to some extent. On straining, these nuggets
coated with mercury are separated, and the gold miner's

amalgam consequently of variable composition, generally

is

containing from 25 to 50 per cent, of gold, the percentage

of gold being highest when the average size of the gold
particles is greatest.
When mercury applied to the clean surface of a piece
is
" "
of solid gold, it and at once penetrates through
it wets
it, passing between the crystalline faces, making it brittle
and of silver-white colour. A gold coin or ring is imme-
diately whitened by mercury, and can be broken in the
fingers, the fractured surfaces (crystalline faces) being as
white as the outside. The amount of mercury so taken up
is, however, small, and most of it can be removed by careful
heating below a red heat, when the mercury volatilises and,
unless air is excluded, the coin blackens.
At 440, or about 150 below red heat, a large part of the
mercury is removed from amalgam by distillation, and a
46 THE PRECIOUS METALS

mixture containing about 75 per cent, of gold remains. At

a bright red heat almost all the remainder of the mercury
is expelled, but about 0*1 per cent, is retained and cannot
be driven off below the melting point of gold.

Amalgams are all attacked by hot nitric acid, the

mercury being dissolved and the gold left as a spongy

mass. If the action of the acid is kept very slow by dilution
and cooling, the gold is left behind, nearly pure, in the
form of crystalline needles.

No evidence has been adduced of the existence of any

true compounds of gold and mercury.

GOLD AND ZINC.

The alloys containing less than 14 per cent, of zinc are

pale yellow and of about the same hardness as gold. They
increase gradually in brittleness with the increase in per-

centage of zinc. The first additions of zinc rapidly reduce

the melting point and give rise to a long pasty stage

during solidification, so that gold containing one or two per

cent, of zinc can be readily welded. The alloy, gold 85*6,
zinc (approximately corresponding to the formula
14*4
Au 2 Zn, that is, containing gold 66*6 atoms, Zn 33*4 atoms,
see Fig. 6) , is a eutectic, solidifying at 648. As zinc increases
from 14 to 25 per cent. (33'4 to 50 atoms per cent.), the
colour of the alloy gradually changes from pale yellow to a
and the melting point rises from
beautiful reddish-lilac tint,
648 to a maximum of about 750. This is the melting

point of a distinct compound, AuZn, containing 25 per cent.

of zinc. It forms lustrous crystals and is brittle. With
further additions of zinc the lilac colour fades, the alloys

containing from 25 to 30 per cent, of zinc consisting of

 THE ALLOYS OF GOLD 47

lilac polygonal crystals of AuZn set in a white matrix.

The compound AuZn 2 , containing 39'8 per cent, of zinc

(66*6 atoms per cent.) is a hard, white, homogeneous alloy,
and will scratch steel, but is as brittle as glass and breaks
with a conchoidal fracture, which has a brilliant lustre. It
1
solidifies at 650. According to Vogel this compound is

1064

1000 1000

300 900

800 800
Auln
700
\
600 500
3
Au-In fo

500 500 fc

400 400^- I
10 20 30 4-0 50 60 7O 80 90 100

Percentage of Gold in Atoms

FIG. 6. Freezing Point of the Alloys of Gold and Zinc.

Au 3 Zn 5 and contains 35'5 per cent, of zinc. The alloys con-

taining more than 40 per cent, of zinc (67 atoms per cent.)
become less brittle and less lustrous as the proportion of

zinc increases. They are not homogeneous, and present a

dull bluish-white appearance resembling zinc. The com-

pound AuZn 8 containing 72*6 per cent, of zinc (88*9 atoms

,

per cent.), makes its appearance in these alloys. It is a

Anory. Chem., Vol. XLVIII. (1906),
1
Zeit. p. 319.
48 THE PRECIOUS METALS

hard white substance solidifying with surfusion at about

470 (along the lineXY, Fig. 6), and probably occurs in the
zinc crusts which are formed in Parke's process. The
alloys containing more than 72 per cent, of zinc melt at

temperatures between 419 and 470 (see Fig. 6).

All the alloys of gold and zinc are brittle, but those con-
taining from 30 to 80 per cent, of gold (13 to 58 atoms per
cent.) are more brittle than the others. The alloys poor in
gold are best prepared by adding solid gold to molten zinc
below a red heat. The gold is taken up by the zinc quietly.
In preparing the alloys containing over 50 per cent, of gold
which solidify above a red heat, the metals ignite with
momentary incandescence, and a small part of the zinc is

volatilisedand lost. On heating these alloys to about 1,000

the zinc can be boiled off for the most part.

GOLD AND CADMIUM.

These alloys present a general resemblance to the alloys
1
of gold and zinc. Heycock and Neville, by distilling off
the excess of cadmium from an alloy of the two metals,
obtained residues which corresponded approximately in
2
composition with the formula AuCd, but Vogel believes
that the compound is Au 4 Cd 3 containing 30 per cent, of
,

cadmium. The compound is greyish or silver-white, some-

what cadmium, and is very brittle. It solidifies at
like

623. Another compound, AuCd 3 containing 63 per cent,

,

of cadmium, melts at 493 and is extremely hard and

brittle. There is a eutectic alloy containing 87 per cent.

1
Chem. Soc. Jour., Vol. LXI. (1892), p. 914.
2
Anorg. Chem., Vol. XLVIII. (1906), p. 333.
Zeit.
 THE ALLOYS OF GOLD 49

of cadmium, and freezing at 303, the melting point of

cadmium being 320. The alloy containing 31'5 per cent,
gold and 68'5 per cent, cadmium, etched with concentrated
nitric acid, is shown under a magnification of 25 diameters

in Fig. 7. It consists of white crystallites of AuCd 3 set in a

black eutectic mixture. The photomicrograph is after Vogel.

The hardness of this series attains to a maximum in the

alloys containing 18 to
30 per cent, and 51 to
63 per cent, of cadmium

respectively. The re-

maining alloys have

about the same degree
of hardness as the
metals themselves. The
metals when melted
together unite with in-
candescence to form the

alloys,some of the
cadmium being
& volati-
.
FIG. 7. Gold, 31-o per cent.;
lised and oxidised. Cadmium, 68-5 per cent, x 25.

Like the alloys of

gold and copper and of gold and zinc, these alloys can be
"parted" by nitric acid and are completely dissolved by
aqua regia. The zinc-gold and cadmium-gold alloys are
also readily parted by dilute sulphuric acid.

GOLD AND TIN.

Three definite compounds of gold and tin have been

found to exist. 1 They are AuSn, containing 62*4 per cent.
1
E. Vogel, Zeit. Anorg. Chem., Vol. XLVI. (1905), p. 60.

P.M. E
50 THE PEECIOUS METALS

of gold,AuSn 2 containing 45*5 per cent, of gold, and AuSn 4

, ,

containing 29*4 per cent, of gold. AuSn is a metallic silver-

grey alloy, harder and more brittle than the other alloys,
and having a higher electrical conductivity than other
gold-tin alloys except those containing over 95 per cent, of
gold. AuSn 2 is also silver-grey, but consists of large crystals,
which are often 10 millimetres long. AuSn 4 is coloured
gold-brown by nitric acid, while AuSn 2 is unchanged.
AuSn is as resistant as pure gold to the action of acids.
The curve of fusion falls sharply from the melting point
of gold to a eutectic point at 280 and 20 per cent, of tin.
It less rapidly to a maximum at 418, the melting
then rises
point of AuSn, and again falls to a second eutectic point at
217 and 90 per cent, of tin, finally rising to the melting
point of tin, 226.
When alloys of gold and tin are treated with aqua regia
the colour of purple of Cassius is seen, and when a gold-tin

alloy is heated in an electric arc furnace and the metallic

vapours allowed to escape into the air, the tin burns and a
substance isobtained having the properties of the purple
of Cassius and containing tin oxide, calcium oxide and
1
gold. The residual ingot of gold and tin in the furnace is
richer in gold than the original alloy. The alloy containing
8 per cent, of tin is brittle and has a pale whitish-yellow
colour externally, with a yellowish-grey earthy fracture.
Gold containing only 1 or 2 per cent, of tin is, however,
perfectly ductile.
GOLD AND ANTIMONY.
About 1 per cent, of antimony makes gold brittle when

it is not annealed, and none of the alloys are malleable.

1
H. Moissan, Ccmptes Rmdus, Vol. CXLI. (1905), p. 977.
 THE ALLOYS OF GOLD 51

One definite compound, AuSb 2 , exists, containing 45 per

1
cent, of gold. This substance is of the same colour as
antimony and is extremely brittle and harder than its
components. It melts at 460. The lowest freezing point
of the seriesof alloys is at 360, corresponding to a

eutectic which contains 76 per cent, of gold. A small

quantity of antimony greatly reduces the melting point of
gold and gives rise to a long pasty stage. Hatchett showed
that molten gold absorbs the vapours of antimony and is
made brittle thereby. All the alloys are formed with con-
traction. They are not easily attacked except by aqua

regia. According to Cosmo Newbery, when they are finely

powdered and triturated with mercury they are slowly
decomposed, yielding gold-amalgam and black metallic

antimony.
GOLD AND ARSENIC.
These alloys have been little studied, but appear to
resemble the antimony alloys. They are easily fusible,
yellow or grey alloys.
brittle, pale Owing to the volatility
of arsenic they are not readily formed above a dull red
heat. See also p. 36.
GOLD AND BISMUTH.
Bismuth alloys with gold, forming pale yellow or greyish
alloys which are readily fusible and are all very brittle,

although none are hard. Of all metals bismuth has the

greatest effect in decreasing the ductility of gold. Even
0*25 part of bismuth in 1,000 parts of fine gold is enough
to make the metal fragile, 2 so that it breaks under the

hammer. This is due to the fact that bismuth is not

1
E. Vogel, Zeit. Anory. Chem., Vol. L. (1906), p. 151.
2
33rd Annual Report of the Mint, 1902, p. 72.
2
52 THE PEECIOUS METALS

dissolved by gold, and no definite compound of gold and

bismuth exists. The freezing-point curve of the series of

alloys consists of two straight branches which meet in a

1
eutectic point at 240 . The eutectic mixture contains 82

per cent, of bismuth. Molten gold absorbs the vapours

of bismuth and becomes brittle. All the alloys have a

strong tendency to segregate. Small quantities of bismuth

do not lower the melting point of gold rapidly as in the
cases of antimony and cadmium, but when the bismuth
exceeds about 5 per cent, there is a long pasty stage

terminating at 240.

GOLD AND LEAD.

These alloys are readily fusible, and very brittle, ranging
in colour from pale yellow by yellowish-grey to bluish-
white. About 0*15 per cent, of lead makes pure gold some-
what brittle. There are two definite compounds of gold
and lead, AuPb 2 and Au2Pb, containing 32 and 65*5 per
2
cent, of gold respectively. The former forms long white
needle-shaped crystals, which are readily distinguished
under the microscope from the large well-shaped crystals of
Au 2Pb.
Large white crystals of Au 2 Pb set in the eutectic are
shown in Fig. 8, which is a photomicrograph after Vogel of
the alloy containing 70 per cent, of gold, slowly cooled and

magnified 70 diameters.
There are three eutectic alloys, containing mixtures
of and AuPb 2 AuPb 2 and Au 2 Pb, and Au 2 Pb and
lead ,

gold respectively. The first of these (D, Fig. 9) has the

Anorg. Chem., Vol. L. (1906), p. 145.

1
E. Yogel, Zeit.

Anorg. Chem., Vol. XLV. (1905), p. 11.

2
E. Vogel, Zeit.
 THE ALLOYS OF GOLD 53

composition gold 14*8 per cent., lead 85*2 per cent., and
solidifies at 215 or 111 below the melting point of lead.
The second (C, Fig. 9) solidifies at 254 and the third
(B, Fig. 9) at 418. All the alloys containing no more than
70 per cent, of gold solidify below a red heat. The freez-

ing-point curve is shown in Fig. 9, which

Vogel. is after

Lead is separated from gold by solution in nitric acid,

but is still more readily
removable by exposure
of the molten alloys to

a current of air at a
red heat (cupellation,
q.v.). Under these
conditions the lead is

oxidised, and the

litharge, PbO, so formed
can be removed by
absorption by a porous
substance, such as
bone-ash or compacted
magnesia, or may J be
Fl0 8
' - Id 70
' r
f "*%
Lead -

30 per cent, x 70.

allowed to flow away
in shallow channels in the furnace bed, the molten metal

being held together by its higher surface tension. Almost

the last traces of lead are removable in this way, although
it is oxidised only when at the surface of the molten gold.

Moreover, the molten litharge dissolves and carries away

the oxides of other metals which are not readily fusible
alone, and consequently a convenient method of removing
base metals from gold is to melt it with lead and subject the
whole to cupellation. The method is used in refining and
54 THE PEECIOUS METALS

in assaying. platinum and other metals remain

Silver,
with the gold, and from this cause, as already stated, silver

10 20 30 40 50 60 70 80 90
noo
[

1100

Au
1
1000' I000

900< 900<

'

800' AuPb. 800

700' 700

1
BOO 600

500' 500'

400 (

400'

Pb
300 300'

200 200'

100' too'

10 20 30 40 50 60 70 80 90 100

Percentage weight of lead .

EIG. 9. Freezing Points of Alloys of Gold and Lead.

and gold were known as " noble metals," because they would
" "
resist the fire, whilst the base metals were destroyed as
 THE ALLOYS OF GOLD 55

such. The art of cupellation is of great antiquity, and was

known to the ancient Hebrews.

GOLD AND IRON.

Gold and iron unite without much difficulty and in all

proportions, forming alloys which are hard but malleable

and ductile, so long as the proportion of iron does not exceed
80 per cent. The only difficulty encountered in preparing
the alloys caused by the high temperature at which iron
is

melts, but it is gradually taken up by molten gold at a

temperature of 1,100 to 1,200. If cast iron is used, melt-
ing at about 1,130, this difficulty disappears. The freezing-
point curve of gold and iron has been determined by Isaac
and Tammann. 1 It is shown in Fig. 10. The first

additions of iron to gold cause a fall in the melting

point, the lowest point reached being 1,040, when 5 per

cent, of iron is present. With a further increase in the
proportion of iron the melting point rises again and reaches
1,168, when 26 per cent, of iron is present. This tempera-
ture marks the completion of solidification of all mixtures
containing between 26 and 74 per cent, of gold. The gold
does not affect the temperatures of the critical points

(changes in allotropic form) of iron.

The alloys containing 8 to 10 per cent, of iron are pale
yellow, very ductile and susceptible of receiving a high
polish. Alloys containing between 15 and 20 per cent,
of iron are used in jewellery in France under the name

or gris. They are greyish-yellow and very hard, but are

easily worked. The alloy with 25 per cent, of iron is also
employed by jewellers under the name or bleu. Those
1
Zei't. Anorg. Chem., Vol. LHI. (1907), p. 291.
56 THE PKECIOUS METALS

containing from 75 to 80 per cent, of iron are silver- white,

extremely hard> and are attracted by the magnet. The gold

1600

1500

1400

1300

1200

/I00

'000

900

800
 THE ALLOYS OF GOLD 57

dissolution in hot dilute sulphuric acid, or less conveniently

by hydrochloric acid or nitric acid. The alloys of gold and

iron are formed with expansion.

GOLD AND NICKEL.

Small quantities of nickel alloyed with gold produce
pale yellow, hard,
veryductile alloys,
J484-
which are magnetic
and take a good polish.
l
Hatchett found that
8 per cent, of nickel
made gold brittle with
a coarse-grained earthy
fracture. The two
metals form a eutectic, 2
which contains about
22 per cent, of nickel
and melts at 950. The
freezing-point curve of too
the alloys is shown in
Percentage weight of Nickel.

Fig. 11, which is after

FIG. 11. Freezing Points of Alloys of
Levin. Gold and Nickel.

GOLD AND COBALT.

Cobalt has a far greater than nickel in making gold
effect

brittle. According to Hatchett, standard gold becomes

brittle if only one-eightieth part of cobalt is melted with it.
1
Phil. Trans., 1803, p. 43.
2
Levin, Zeit. Anorg. Cham., Vol. XLV. (1905), p. 238.
58 THE PEECIOUS METALS

GOLD AND MANGANESE.

These alloys are very hard and melt at a higher tempera-
ture than gold. The alloy containing about 12 per cent, of
manganese is of a pale greyish-yellow colour with a lustre
like that of polished steel. It is somewhat malleable, and
shows a coarse-grained fracture. The manganese can be
removed by cupellation with lead. The alloy containing
67 per cent, of manganese is grey, hard, very slightly ductile,
and has a coarsely granulated fracture. The alloy of gold
10 per cent., manganese 90 per cent., is pale grey in colour,
and is very ductile with a fine-grained fracture.

GOLD AND PLATINUM.

These metals are not known to form any definite com-
pounds, but can be alloyed in all proportions. The first
additions of platinum at once raise the melting point of the

alloy above that of gold, and the alloys formed are malleable
and ductile. When the proportion of platinum exceeds 50

per cent, the alloys are brittle and greyish in colour.

A thermal study of the alloys containing up to 60 per
platinum has been made by Doerinckel with the
1
cent,
results shown in Fig. 12. The curve ABC marks the
beginning and the curve ADC the end of solidification.
The two metals appear to form a continuous series of mixed
crystals. When examined under high powers of the micro-
scope the alloys are seen to consist of crystals of which the
cores, richer in platinum than the edges, appear in relief.

By annealing at the temperatures indicated by the curve

ADC, the alloys become completely homogeneous.
Anorg. Chem., Yol. LIV. (1907),
1
Zeit. p. 345.
 THE ALLOYS OF GOLD 59

The yellow colour ofdisappears rapidly on the

gold
addition of platinum, and only a slight yellow colour
remains when 30 per cent, of platinum is present. If the
platinum is above 40 per cent, the colour is that of

platinum. The alloy of gold 10, platinum 90, has a

60 THE PKECIOUS METALS

the platinum moving towards the centre of the ingot on

solidification. These alloys cannot be separated into their
constituents by cupellation and parting in acid in the
ordinary way. They are not attacked by single acids, but
are dissolved by aqua regia.

GOLD AND PALLADIUM, EHODIUM, IRIDIUM, ETC.

These alloys have been very little studied. The gold-
palladium alloys form a continuous series of mixed crystals
which gives the curves of fusibility after Kuer).
(see Fig. 18,
Gold containing from 10 to 20 per cent, of palladium is
nearly white and is hard but ductile. The alloy of gold
50 per cent., palladium 50 per cent., is said to be iron-grey
in colour and deficient in ductility.

Gold forms difficultly fusible but ductile alloys with

osmium and rhodium. It isdoubtful whether any real

alloy with iridium is formed. When gold containing

iridium is melted, the iridium sinks to the bottom. On
treatment with aqua regia the gold is dissolved and the
iridium left in the form of a black powder which is mainly
if not entirely oxide of iridium.

GOLD AND ALUMINIUM.

These metals form at least two definite compounds. One
ofthem, Au 2 Al, containing 6'4 per cent, of aluminium, is a
hard white substance melting at 622 1 The eutectic alloy
.

between this compound and pure gold contains only 4 per

cent, of aluminium but melts at 525, over 500 below the

melting point of pure gold. The other compound, AuAl 2 ,

i
Phil. Trans., Vol. CXCIVA. (1900), p. 201,
 THE ALLOYS OF GOLD 61

containing 21*5 per cent, of aluminium, solidifies at a

higher temperature than the melting point of pure gold
1
itself. This compound is of a beautiful purple colour
which resembles but is deeper in tint than the compound
1600 1600

1500

1400 -

-
1200

-
1100

1000 woo
10 20 30 40 50 60 10 80 90 100

Percentage weight of Palladium,

FIG. 13. Freezing Points of Alloys of Gold and Palladium.

AuZn. Itformed whenever gold and aluminium are

is

fused together, provided that at least 10 per cent, of

aluminium is present. The purple compound is readily
seen in any of these alloys, occurring in patches in a white
1
Proc. Roy. Soc., Yol. L. (1892), p. 367.
62 THE PRECIOUS METALS

matrix consisting of the excess of aluminium or of the

white compound of aluminium. There is little evidence of
the existence of any other compounds of gold and aluminium

except these two.

GOLD AND TELLURIUM. 1
The curve shown in Fig. 14. A compound,
of fusibility is

AuTe 2 , containing 43'6 per cent, of gold exists, freezing at

452. This corresponds in composition to the mineral

species sylvanite and calaverite. There are two eutectics,
one containing 60 per cent, of gold and fusing at 432, and
the other containing 20 per cent, of gold and fusing at 397.
The presence of 0*05 per cent, of tellurium is enough to

cause the formation of a small quantity of the gold-rich

eutectic, and gold containing this amount of tellurium is

accordingly deficient in ductility. When more tellurium is

present, the alloys are very brittle.

The alloys rich in gold consist of crystals of gold set in
the greyish-white eutectic. Gold containing 2'4 per cent,
of tellurium is of a pale brass-yellow colour, but even when

40 per cent, of tellurium is present, a faint yellowish tinge

is still perceptible, which disappears before the pure com-
pound AuTe2 is reached. The behaviour of tellurides of
gold on roasting ispartly explained by the remarkably low
temperatures of fusion of these alloys, all mixtures con-

taining less than 60 per cent, of gold being completely fused

before a red heat is reached. The production of spherical
globules of gold must therefore accompany the roasting
of gold ores containing tellurides, however carefully the
operation may be carried out. These globules will not be
free from tellurium.
1
T. K. Eose, Trans. InsL o/Mng. and Met., Vol. XVII. (1908), p. 285.
 THE ALLOYS OF GOLD

GOLD AND THALLIUM. 1

These metals form neither compounds nor homogeneous
solid solutions. There is a eutectic alloy containing 27 per
cent, of gold and melting at 131. This eutectic occurs in

30 40 50 60 70 BO 90 100

Au

-
(000

- 900

10 20 30 40 50 60 70 80 90

Percentage weight of Gold

FIG. 14. Freezing Points of Mixtures of Gold and Tellurium.

all the alloys, so that the addition of a small amount of

thallium to gold gives rise to a long pasty stage, terminating

at 131. The alloys are all brittle.
1
Levin, Zeit. Anorg. Chem., Vol. XLV. (1905), p. 31.
 CHAPTEE V
OCCURRENCE OF GOLD IN NATURE I GOLD ORES

GOLDgenerally occurs in Nature in the metallic state

alloyed with a varying proportion of silver and smaller
quantities of copper and other metals. Native gold is never
found pure, but sometimes contains less than 1 per cent,
1
of impurities. According to Furman, the finest native
gold which has been met with is that from the Pike's Peak
mine, at Cripple Creek, Colorado, which contained 99 '9 per
cent, of pure gold.
" "
Gold occurs in the form of nuggets, grains and dust
in many river gravels. In this form it was known to the

ancients, and, as already pointed out, was probably the

earliest metal known to man, although it is unlikely that
any use was made of it until long after its first discovery.
In some gravels masses of gold of considerable size occur.
The largest piece recorded was the "Welcome" nugget
which was found in Victoria. It weighed 2,195 oz., and
yielded gold to the value of 8,376 10s. 6d.
Large nuggets
have also been found in other parts of Australia, in Kussia,
and in California. Auriferous deposits consisting of sands
and gravels formed by the erosion of older rocks are known
as placer deposits or placers. The gold in these deposits
is rounded by attrition, all its sharp angles having been
worn off. The nuggets of gold are often larger and purer
Manager and Metal Miner, October, 1896,
1
Colliery p. 89.
THE OCCUEEENCE OF GOLD IN NATURE GOLD OEES : 65

than the masses found in lodes in the same localities ; but

there now
is doubt that the gold in placer deposits
little

previously existed in ores in veins and has been derived

from them by the action of running water, or in rare cases
by the waves of the sea. A large proportion of the gold
production of the world was formerly obtained from placers
by washing, but placer gold is now of secondary importance,
most of the present enormous output of gold coming from
gold ores which require to be crushed before the values can
be extracted.
The occurrence of well-defined crystals of gold is rare,
but has been noted in lodes in Transylvania, California,
Australia and Brazil. The usual forms are cubes and
octahedra, with rod- and plate-like forms. The crystal faces
are frequently striated, and the crystals are generally

rounded, uneven, or even curved. Twinning gives rise to

dendritic and reticulated groups, but individual crystal

faces and edges are usually very small. Moss-gold, leaf-
gold and wire-gold are forms occurring sometimes. The
interior of nuggets when polished and etche'd shows a

crystalline structure similar to that of fused masses. No

traces have been observed of a laminated, concentric structure
such as might be expected if the nuggets had grown by
accretion by means of deposition from solution. It is of

course arguable that the deposited gold may have been

rearranged after deposition by the slow effects of solid

diffusion.

Gold ores are usually found in veins or lodes in rock

formations. The greater part of the material ("

gangue")
forming the lodes is usually quartz, but various silicates
also occur, and sulphides, arsenides or antimonides of base
P.M. F
66 THE PRECIOUS METALS

metals are almost invariably present. The upper portions

of these veins near the surface of the ground are often
found have undergone oxidation from the influence of the
to

atmosphere, so that the sulphides of iron, etc., have been

converted into oxides. These gossans, or oxidised portions,
may extend downwards to a considerable depth, and the
gold contained in them, associated as it is with a loose
mixture of silica and oxide of iron, is often readily separated.
When the oxides pass into sulphides with increase of depth,
the extraction of gold may become more difficult. The
great deposit of mainly composed of
ferruginous sinter,
silica, found at Mount Morgan in Queensland, from which
gold to the value of several millions of pounds has been
extracted, is the richest gossan known.

Generally, below the water level in the rocks of the district

gold is found associated with metallic sulphides.
According to Dr. Don gold is always present in lodes,
1

in the districts of Australia which he examined, when

pyrites occurs, and is absent when no pyrites is observable.
According to L. Wagoner, however, gold sometimes occurs in
Western America in rocks free from sulphides. The gold
in lodes is in the form of minute scales, threads or grains
disseminated through the ore. It is sometimes visible to

the eye without magnification, but is more often in too fine

a state of division to be seen. It also occurs in copper and
iron pyrites, gilding the faces of the crystals or otherwise
associated with them, probably in the metallic state, not as
a sulphide.
If sulphides of gold are rare or unknown, on the other
hand tellurides of gold, containing the compound AuTe2,
1
Trans. A. /, M. E., Vol. XXVII. (1897), p. 564.
THE OOOUBEBNOB OF GOLD IN NATURE GOLD ORES :
67

are very common. They occur in considerable quantities in

Western Australia, Colorado and Transylvania, and have
been reported from many other localities. Various rniner-
alogical names have been assigned to the tellurides from
different localities, the best known being calaverite, which

has the composition AuTe 2 sylvanite or graphic tellurium,

,

which appears to be a variable mixture of AuTe 2 and Ag3 Te 4 ,

petzite (Ag 2 Te, in which some silver is replaced by gold)

and nagyagite, or foliated tellurium, which contains a
considerable percentage of lead.
The most part dark grey or black in
tellurides are for the

colour, rarely silver-grey. They are often mixed with

metallic gold, which sometimes gives them a brassy-yellow
colour. When heated in air they oxidise, fuming and
giving off Te02, and fuse below a red heat. The removal
of most of the tellurium leaves the gold in the form of
round pellets which have solidified from fusion. Calaverite
has a density of about 9, and contains about 44 per
cent, of gold, part of which is usually replaced by silver.
Sylvanite has a density of about 8*1. Analyses of two
specimens from South Dakota and Cripple Creek respec-
1
tively gave the following results :

(1) Gold . 7*64 per cent. (2) Gold . 5'61 per cent.
Silver . 32'39 Silver . 34*23
Tellurium 59'96 Tellurium 60*16

99-99 100-00

Petzite, or Ag2 Te, has a density of 8'86 (Miers). A

specimen from California contained 25 '6 per cent, of gold
1
F. C. Smith, Trans. A. I. M. E., Vol. XXVI. (1896), p. 485.
F 2
68 THE PRECIOUS METALS

and 41*86 per cent, of silver (Genth).

Nagyagite is foliated
like graphite, but has a density of about 7. It contains
only a few per cent, of gold.
The proportion of gold present in ores or gravels which
can be treated profitably for its extraction is very small.
The mean return from the Rand ores is less than half an
ounce per ton, or about one part of gold in 70,000 parts
of worthless material. Auriferous gravels which do not

require to be crushed are sometimes treated at a profit

when they contain only two or three grains of gold per

ton, or say one part of gold in five millions.

Beyond the limits of profitable extraction gold isvery widely
disseminated. Minute quantities of gold appear to occur
1
in all ores of silver, copper, lead, antimony and bismuth.
has also been detected in igneous rocks in almost every
It

case in which a diligent search for it has been made. It

is perhaps even more generally distributed throughout
metamorphic rocks, and sedimentary rocks are seldom
quite free from it. L. Wagoner found 2 an average of
0'37 part of gold per million, or 5 grains per ton in granites,
0*03 part per million in sandstones, and 0'007 part per
million in limestones. It also occurs in many clays and
shales. For example, the bed of clay on which the city of
Philadelphia is built is an auriferous deposit. It has been

suggested that the gold in sedimentary deposits has been

derived from the sea, but it seems at least equally probable
that the gold previously existed in the older rocks from
which these deposits were formed. The comparatively small
quantity of gold in limestones which are formed in clear
"
Metallurgy of Gold," T. K, Rose, 5th edition,
1
p. 36.
a
Trans. A. I. M. E., November, 1901,
THE OCCURRENCE OF GOLD IN NATURE GOLD ORES : 69

water far from land points rather to the land as the place
of origin of the gold.
/ Gold was detected in sea water by Sonstadt, 1 who
first

states that the amount in British waters is far less than

one grain per ton. Liversedge found 2 between one-half
and one grain per ton in various samples of water taken
from the sea off the coast of Australia. L. Wagoner 3
found from 0'6 to 3'7 grains of gold per ton of water from
the depths of the Atlantic, and only 0*2 grains per ton in

Chesapeake Bay. The total quantity of gold in the sea is

evidently enormous, amounting, on the figures given by

Liversedge, to about 10,000,000 to each inhabitant of the
globe, but there appears to be no prospect of its successful
extraction on a large scale. It is not known how gold
came to be present in sea water. /
The origin of gold ores is not yet certainly determined.
Gold usually contained in quartz veins which pass
is

through rocks of many different kinds. The effects of

segregation, by which solid crystals of definite composition

grow in a fluid containing a number of different constituents,

seem be quite enough to account for the existence of

to

quartz veins in igneous and metamorphic rocks and of gold

in quartz veins. Given a molten mass of rock, there would
be nothing surprising in nearly pure silica beginning to
solidify at certain points and receiving additions of more
silicaby the operations of diffusion. When the whole
mass had become solid, the segregated silica would readily
become converted into quartz and also receive still further

1
Chem. News, Vol. XXVI. (1872), p. 159.
2
Roy. Soc. N.S.W., October 2, 1895.
3
Trans.'A. L M. #.,1907.
70 THE PBECIOUS METALS

additions by the agency of heated water, by which silica

would be dissolved, transferred and re-precipitated. The
concentration of gold in quartz veins and other ores
may have been effected in a similar manner. 1 It is of
course well known when
a crystalline nucleus of any
that

mineral, such as gold, has been formed, there is a tendency

for all particles of the same mineral which are brought

near the nucleus by diffusion or in other ways to be fixed to

it, so that the crystal grows. The almost universal dis-
semination of gold throughout rocks of all sorts and in the
waters passing through them would thus afford an unfailing
source for the creation of gold ores by the enrichment of
certain zones. The length of time required for the formation
of the gold ores which have been discovered is not known.
The veins containing gold ores mainly traverse meta-

morphic rocks, especially slates or schistose rocks, such as

hydromica and chlorite schists. Valuable gold ores are

sometimes met with in basalt, rhyolite, and other igneous
"
rocks, and also occur in many different forms, as replace-

ment-deposits in limestones, as disseminations in igneous

and sedimentary rocks, and as contact-deposits near intrusive
2
masses . . ."

Gold-bearing gravels are, generally speaking, valuable

when they occur in river-beds in districts containing gold

ores, and consequently they are for the most part confined
1
For a
full discussion of the origin of gold ores, see papers by

Posepny, Spurr, Bickard, Weed, Wagoner, Don, and others, in the

Transactions of the American Institute of Mining Engineers, and else-
where. For a short resume with references, see "Metallurgy of Gold,"
T. K. Eose, 5th edition, pp. 4244.
2 "
Igneous Bocks as related to Occurrence of Ores," J. E. Spurr,
Trans. A. I. M. E., 1902. ,
THE OCCURRENCE OF GOLD IN NATURE GOLD OEES : 71

to rivers flowing through metamorphic rocks, especially

slates and schists. The beds of ancient rivers which have
long ceased to flow are sometimes very rich. Such river
gravels have been covered by thick masses of eruptive basalt
and other deposits in California and Victoria. In the latter
"
country they are known as deep leads."
The geographical distribution of gold ores corresponds

roughly with what has been said as to the geological

distribution. Mountainous districts and the streams flow-

ing from them are frequently auriferous. Wide, flat, alluvial

plains generally contain no gold. In Europe the chief

gold mines are in Hungary and Transylvania. Some gold
is produced in North Italy, Norway, Wales, etc. The
Eussian gold production is mainly from the placers of
Siberia. In India gold occurs at Colar, in Mysore, and
in small quantities in Madras. The other goldfields of
Asia are in Borneo, Celebes, Sumatra, and some other
East Indian islands, in China, Korea, Japan, and in the
Malay Peninsula. In Africa the greater part of the gold
produced isfrom the Transvaal and Khodesia, but deposits
of less importance occur in West Africa, in Egypt, and in

Abyssinia. Gold is widely distributed in the western moun-

North America, alike in Canada, the United States,
tains of
and Mexico. Minor goldfields exist in the- Appalachians
and in Eastern Canada. The goldfields of Central and
South America were formerly more important than at
present.They occur both on the Pacific and the Atlantic
slopes. Gold occurs in many parts of Australia and in
several districts in New Zealand.
The production of gold for different countries is given
in Chapter XVIII.
 CHAPTER VI

EXTKACTION OF GOLD FEOM ITS ORES : GOLD WASHING

WHEN the treatment of gold ores is under consideration

they fall naturally into two divisions (a) loose aggregations,
from which the gold can be removed by concentration in
virtue of its high specific gravity, and (b} relatively hard

ore, which must be crushed before the gold can be extracted.

It is convenient to include in the first division gravels which
have been cemented by infiltration of silica, oxide of iron,

lime, etc., so that they have become too coherent

to be
"
treated without crushing. Many examples of such cement
"
gravels occur in California. The second division includes
some quartzose ores, in which the gold is mainly in the
form of grains of moderate size, so that after crushing
most can be collected by simple concentration.
of the gold

Ores of this character have been met with chiefly in

Australia. Gold in a finer state of division escapes under
such treatment, but is caught by mixing the crushed ore
"
with mercury. The gold then becomes amalgamated,"
and the small particles of gold-amalgam are recoverable by
taking advantage of their tendency to coalesce with one
another, or with fresh mercury, so as to form large globules,
which readily sink in pulp diluted with water. Gold-
amalgam may also be separated from the crushed ore or
pulp by bringing it into contact with a metal plate, the
surface of which has been previously amalgamated by
 GOLD WASHING 73

rubbing with mercury. In the latter case the gold-amal-

gam adheres to the amalgamated plate, and is removed
at intervals by scraping. Mercury is also largely used in

the treatment of auriferous gravels.

Very finely divided gold,

however, may escape amalga-
mation, and the resistance to amalgamation is still more
marked in the case of compounds of gold, such as tellurides,
which show little or no tendency to yield up their gold
when brought into contact with mercury. Many ores, too,
contain materials which exercise a deleterious effect on
mercury, preventing globules from coalescing, so that
its

they are lost in the tailings, together with the gold which
they have picked up. All such ores which do not yield a
satisfactory percentage of their values on treatment with
mercury are known as "refractory" ores, as distinguished
"
from " free-milling ores.
In a sense all ores are partly refractory, inasmuch as a

part of the gold contained in them is not in a condition to

be recovered by means of mercury, however carefully the
treatment is carried out. Accordingly it is now customary
to treat the tailings from the amalgamation process by
mixing them with a solution of cyanide of potassium. Very
finely divided gold is dissolved by cyanide solutions, which
are then separated from the ore. The gold is obtained from
them by precipitation. Sometimes tailings containing rich
pyrites or other sulphides are treated by concentration,
and the concentrates are smelted with lead, or roasted and
subjected to the action of a solution of chlorine in water by
which the gold is dissolved, with a view to subsequent

precipitation.
Refractory ores are treated by smelting, cyaniding, or
74 THE PRECIOUS METALS

chlorinating, according to their nature and to various other

considerations. When the gold has been obtained, whatever
methods have been employed, it is melted and cast into
bars and afterwards refined to fit it for use in the arts.

The methods of treatment of gold ores may accordingly

be tabulated as follows :

"
Simple "washing or concentration, with or without
1.

the aid of mercury. This is suitable for auriferous sands

and gravels (placer deposits).
2. Crushing and concentration, usually with the aid of

mercury : suitable for free-milling quartzose ores.

Subjecting crushed ores, usually without roasting, to

3.

the action of cyanide of potassium suitable for the tailings

:

from free-milling ores and for certain refractory ores.

4. Boasting and chlorination suitable for concentrates
:

from free-milling ores which have already been amal-

gamated, and for certain refractory ores.
5. Smelting in blast furnaces to obtain the gold as an
alloy of lead : suitable for certain refractory ores and con-
centrates. This is described under the treatment of silver
ores, Chapter XL
The methods given above will now be briefly described,
and some of their chief modifications touched upon.

1. GOLD WASHING.
Small quantities of rich sands are washed in the miner's
pan, which was originally the cooking-pan of the prospector.
It is a flat-bottomed pan of sheet-iron, about 16 inches in

diameter, with sloping sides 4 or 5 inches wide. It is often

made with one or more " riffles " or shallow trenches in the
 GOLD WASHING

side or on the bottom near the angle to catch mercury and

When used, it is nearly filled with sand and
amalgam.
water, and is first to enable the gold to
well shaken

settle to the bottom, and then dipped in and out of water in

an inclined position to enable the sand to escape over the
edge by degrees. The final operation of separating
the last

remaining sand from the

gold requires a con-

siderable degree of skill.

This implement was

much used in early days
in California and Aus-
tralia, when a single
panful of dirt sometimes
yielded several ounces
of gold. A somewhat
similar wooden (usually
mahogany) bowl, the
batea, is especially
favoured by the negro
race. It is a conical
vessel about 18 inches
in diameter, and 2 inches FIG. 15. Batea.

deep in the centre (see

Fig. 15). The gold clings to the wood when it would slide

over iron. In the East a wooden trough is used.

For concerted work by a small party of diggers, the
cradle, or rocker, is more advantageous. It consists of a

wooden box, resting on two rockers DD, like a baby's

cradle (see Fig. 16, which is a sectional elevation). One
man shovels gravel into the box A, the bottom of which
76 THE PEECIOUS METALS

consists of a sieve, while another pours on water with one

hand and rocks the cradle with the other. Mercury is
sprinkled on at intervals. The fine sands and gold pass
through the sieve and over some inclined plates B, to which
cross-bars of wood, C, are nailed. The heavy materials
gold, mercury, amalgam, pyrites, etc. accumulate behind
"
the cross-bars, or riffles," and every now and then are

scraped out and panned. The amalgam is heated on a

shovel and the mercury driven off, leaving the gold behind.

PIG. 16. Cradle.

When larger quantities of auriferous gravel are available,

a sluice is generally used. A sluice-box is a slightly
inclined trough made ofwood, through which the gravel is
carried by a stream of water. Mercury is sprinkled on at
short intervals of time. A large number of boxes are
sometimes fitted together, making a sluice which may be
some hundreds of yards long.
The methods of conveying the auriferous material to the
sluices vary with the scale of the operations and the other
conditions. When small sluices or rocking cradles are
used the gravel is piled near at hand and shovelled into
them. In Siberia, where the valleys are shallow and the
 GOLD WASHING 77

inclination of the ground small, the gravel is carried in

cartsup an inclined plane to an elevated wooden platform
whence the sluice starts.
" "
In California, where the gulches are deep, the fall
of the ground rapid, water plentiful, and the auriferous
deposits of great thickness, the banks of gravel are some-
times attacked by jets of water under high pressure, and
the earth washed down and carried through the sluices
without being touched by hand. This is called hydraulic
"
mining, or hydraulicking." When the gravel beds are
below the general level of the country they may be raised
"
by the hydraulic elevator," a jet of water under a head of
as much as 400 or 500 feet, carrying water, sand and
boulders alike up a pipe inclined at some 60 degrees to the
horizon, so as to deliver them all at the head of the sluice,
the vertical lift being sometimes over 50 feet. When
gravels are buried beneath great thicknesses of lava or
" "
other materials, as in the case of the deep leads of

Bendigo, Ballarat and other places in Victoria, the gravel

is mined and raised through shafts. If the gold-bearing

material is hard and compacted, it is coarsely crushed in a

stamp mill resembling those described in Chapter VII., but
sometimes called a " cement mill."
The sluice is paved in various ways to take up the wear
and to catch the gold and amalgam. An irregular bottom
is required with plenty of depressions, chinks and crannies
into which the gold can settle. The obstructions are
known as riffles. The simplest contrivances are fir poles
about four inches in diameter, left rough with the bark on.
These poles are cut to the required length and nailed or
wedged in the bottom of the sluice either transverse to the
 78 THE PRECIOUS METAL

direction of the current or longitudinally. The latter is

generally preferred. Other riffles in use consist of cut slats

of wood nailed to the bottom of the trough or sluice. In
" "
Siberia square pigeon-hole depressions have been con-
tinuously used for more than fifty years. In California
the sluice is sometimes paved with 6-inch square blocks
of wood placed about two inches apart, or with large
rounded stones, or ordinary iron rails placed longitudinally.
All these afford plenty of crevices where the amalgam can
lodge. In later years the use of perforated iron plates and
" "
of expanded metal has largely increased. To catch
light spangles of gold, blankets are spread, the loose fibres
of which become charged with pyrites and gold, and in New
Zealand plush is a favourite gold-catcher. Amalgamated
plates, both stationary and shaking, are also used.
On dredges, where the sluice is necessarily short, a
number of different kinds of riffles are often used in the
same sluice. The view has been expressed 1 that this is
"
advantageous because the ripple in the water as it flows
down the box is altered, and consequently the material
being treated is tossed about, its course altered, and the

gold which may be adhering to the stones or grit has more

chance of being liberated." There are, of course, two some-
what opposed operations to be carried out in a sluice. The
process of disintegration of the gravel, if not previously
"
completed, must be finished in the sluice by tossing
"
about the auriferous material, but the main function of

the sluice must be to save gold. This is done by affording

opportunities to the denser particles of matter to settle to

Marks and of Mng. and Met.,

1
E. S. Gr. N, Marks, Trans. Itwt.
Vol. XV. (1906), p. 479.
 UNIVERSITY
CF

GOLD WASHING 79

" "
the bottom, and any eddies in the water, boiling-up or
disturbances of any kind interfere with the process. The
flakes of gold, instead of continuing to settle towards the
bottom of the sluice, are in such cases carried up again
towards the surface of the stream. Consequently, where
further disintegration is not required the water is allowed

smoothly as possible.
to flow as
Sometimes disintegration in the sluice is necessary, and
this is effected partly by the boulders occurring in the

gravel and partly by arranging drops of a few feet in the

sluice. When boulders are allowed to work down the sluice,

the wear of the pavement is increased.

At the La Grange hydraulic mine in North California, 1
the sluice-boxes are 6 feet wide and 4 feet deep, and are

paved with steel rails weighing 40 per yard, but the

Ib.

number of boulders passing through the sluice is large, and

the rails are worn out in about six months.

In some sluices the larger stones and lumps of clay
are removed by passing the stream of gravel and water

through a set of parallel iron bars (a "grizzly "), or in

small operations by means of a fork or by hand.
The dimensions of the sluice and
angle of slope vary
its

with the quantity and nature of the material to be treated

and the amount of water available. The art of sluicing
consists in arranging that the heavy materials gold,
mercury, amalgam, and "black sand" shall sink to the
bottom of the stream and lodge in the irregularities in the
sluice-bed, but the lighter worthless sand shall be
that
carried away by the water. If sand accumulates in the

sluice, the object is defeated and the amount of water must

and Scientific Press, October 10, 1908, p. 492.

 80 THE PKECIOUS METALS

be increased or the angle of inclination raised. The diffi-

culty lies chiefly in catching finely divided gold, which is liable

to be swept away with the sand. With this object the

stream is kept as shallow as possible, and various
special
devices have been introduced, such as the undercurrent. In
this, a
part (say one-fourth to one-third) of the fine material
and water are drawn off through a sieve placed in the bed
of the sluice and spread in a thin, even stream over the
surface of a wide, inclined table, sometimes supplied with
transverse riffles. After flowing over the table the material
is returned to the main sluice. The fine gold settles on the
table and " "
is caught by the burlap or other material with
which it is covered.
" "
At intervals the work is stopped, and a clean up takes

place to recover the gold caught in the sluice. The water

is allowed to run until it passes through clear, and it is then

deflected from the and the material lodged in the

sluice,
riffles, etc., is scooped out and panned or ground in a revolving
barrel withmore mercury and a few cannon balls. The
amalgam and quicksilver collected in this way are squeezed
through chamois leather or canvas, as in the days of Pliny.
Nearly pure mercury passes through, and pasty amalgam
containing 30 or 40 per cent, of gold remains in the
bag. Pliny describes the process in the following words :

"
ut et ipsum [i.e., argentum vivum] ab auro discedat, in

pellis subactas effunditur, per quas sudoris vice defluens

purum relinquit aurum." The pasty amalgam is retorted,

the mercury driven off and condensed by cold water, and
the residue melted and cast into bars.
The tailings, or refuse material from which the gold has
been extracted, are carried off by the water and discharged
 GOLD WASHING 81

into the nearest river or collected and impounded in

brushwood dams.
One of the main difficulties in hydraulic mining is in the

disposal of the tailings, which may amount to millions of

cubic yards in a year from a single mine. Vast quantities
of debris were discharged by the hydraulic miners into the

between 1862 and 1884, with the result

rivers of California
that the river-beds were gradually filled up and floods
became more frequent and extensive. The farmers' lands
were being covered more and more by deposits of barren
gravel and sand and their industry was threatened with
destruction.They accordingly obtained an injunction from
the Courts which put an end to hydraulic mining in a
1
great part of California in 1884.
Within the last few years the use of dredges has been
rapidly extending. A dredge is a flat-bottomed boat, with
machinery for raising gravel from the bottom of a stream
or pond, hoisting it on board, washing it over inclined
tables to save the gold, and then throwing or " dumping "
the tailings overboard at the stern. A dredge may thus
work its way up-stream, or may proceed across a flat plain
floating in a pond, cutting out the bank in front of it and
piling up the tailings behind. The latter method is known
as "paddock dredging." A
continuous supply of clear
water is necessary for washing in the latter case, as very
muddy water is found to interfere with the gold-saving.
Dredges were first successfully used on the rivers of New
Zealand, but are now at work in many countries. Some-
times suction pumps or grab-bucket dredges are used to lift

1
See Production of Precious Metals in the United States, 1884,
pp. 524530.
P.M. G
82 THE PEEOIOUS METALS

the gravel, but ladder-bucket dredges similar to those used

in ordinary dredging operations are much more common.
The following is a description of a typical New Zealand
1
ladder-bucket dredge : The pontoon is 119 feet long and
50 feet wide at the stern, and its greatest draught is 9 feet
6 inches. The engine is of 25 h.p. nominal. The capacity
of the buckets is 7 cubic feet, and they are worked at a

speed of ten buckets per minute. The ladder carrying

the buckets can be raised and lowered by means of a
winch. The pontoon is moored to the bank and moved
from place to place by a head-line and four side-lines, each
of which is supplied with a separate winch. The buckets
empty the gravel into a trommel, or revolving screen, which
is 31 feet long and 4 feet 6 inches in diameter. The coarse
material is delivered into the buckets of the main tailings
elevator, and the through the screen
fine material falling

passes to the washing where the gold is saved.

tables, The
tailings are deposited in a settling tank, whence they are
lifted out and delivered to the main tailings elevator, which

is 145 feet long and capable of stacking tailings at a height

of 80 feet above the water level. The capacity of the dredge
is reckoned at about 120 cubic yards per hour.

are not supplied with a revolving screen,

Some dredges
and the material is tipped from the buckets direct into a
sluice. In either case the length of the gold-saving tables
strictly limited, and the gravel must be thoroughly
is dis-

integrated and an opportunity afforded for the gold to

settle in a short distance. The tables are supplied with
expanded metal or perforated plates placed on cocoanut
" New Zealand, for 1899
Eeport of the Department of Mines,
1

1900," p. 43.
 GOLD WASHING 83

matting or other material. Angle-irons and Hungarian

riffles are also used, but the losses of finely divided gold

are frequently heavy.

Tailings in river dredging are often discharged into the

water over the stern of the pontoon. In paddock-dredging
it is of course necessary to stack them on the bank. One
of the drawbacks that stacked tailings do not usually form
is

ground capable of cultivation, and arable land is often laid

waste by dredging operations.
Cleaning-up usually takes place weekly. The expanded
metal or other plates are lifted and the mats washed in
tubs until free from all gold, grit, etc. The concentrates
" "
may again washed on a
be streaming-down table,
which is covered with plush or green baize, and the final
concentrates are amalgamated and panned.

Dredges are very economical in work, some large ones

treating from 3,000 to 6,000 cubic yards or more per
day, at a total cost of 2d. or 3d. per cubic yard, as in the
subjoined table.
84 THE PEECIOIIS METALS

are given and G. N. Marks. 1

by E. S. The first is from
a dredge working on the Buckland river, Victoria, and

dealing with gravel containing a quantity of sunken logs,

which delayed the operations. The second worked in the
Australian Alps on the Encumbene river at an altitude of
5,000 feet, and time was lost in the winter by the dredge
being frozen up.

1
"The Bucket Dredging Industry," Trans. Inst. of Mny. and Met.
Vol. XY. (1906), p. 491.
 CHAPTER VII

TREATMENT OF GOLD ORES BY CRUSHING

AND AMALGAMATION

MOST gold ores are crushed in

stamp-mills, which consist of heavy

stampers or pestles raised by machi-
nery and let fall in cast-iron mortars.
A stamp consists of the stem (A,
Fig. 17), a rod of wrought iron or mild
steel,12 or 15 feet long and 3 or 4
inches in diameter, the head B, and
the shoe C. The stamp is raised

by the double cam D, acting on the

tappet E, which is keyed to the stem.
The stem moves in guides E, of which
there are two. The total weight of
the stamp is generally from 1,000 to
1,500 lb., but even heavier stamps, up
to a ton in weight, are now preferred.
The stamps are raised and let fall 6

or 8 inches about 90 or 100 times a

minute. The ore is crushed between
the shoe C and the die G. Five FIG. 17.

stamps are let fall side by side in Diagram

showing the
each mortar. various
parts of a
A section of a mortar is shown in Stamp.
86 THE PRECIOUS METALS

Fig. 18, where a the aperture through which the ore and
is

water are fed-in constantly, I is the screen opening set in

the side of the mortar, and d is the die. An amalgamated
copper plate often bolted to the inside of the mortar at c
is

just below the screen. This is called the " chuck-plate."

The fall of thestamps splashes the pulp against the
screens. The water and fine particles of ore pass through
and are then allowed to run
over the amalgamated tables
where the gold is saved.
The ore to be fed to a

stamp-mill is now usually

broken to moderate size by
being passed through a rock-
breaker. These machines
have approaching and reced-
ing jaws which crush the
rock. They are divisible
into two classes, the recipro-
FIG. 18. Section of Mortar.
cating jaw breakers, such as
the Blake crusher, and the gyrating breakers, which are
continuous in action, such as the Gates crusher. In the
Blake machine a steel plate or jaw, pivoted above and having
a swinging motion, alternately advances towards and recedes
from a fixed vertical jaw. When the swinging jaw advances
some rock is crushed. When it recedes the opening at the
bottom between the jaws is widened and the crushed ore
drops through. This machine can be set to break rock down
to about two-inch cubes, and a large size machine can easily
break 500 tons of ore per day to this size. The ore is

usually shovelled into the mouth of the breaker by hand.

 TREATMENT OF GOLD OEES BY AMALGAMATION 87

The gyrating machines contain a spindle carrying a

crushing head, which rotates inside conical crushing faces
when the machine is empty and gyrates when ore is fed in.
The largest of these machines are of greater capacity
than the reciprocating machines, and work at a lower
cost.

The ore from the rock-breakers is usually broken down

in them to pass a 2-inch ring. Such ore is in the right
condition to be fed to the stamps. It is stored in a bin

with sloping bottom, and sliding down on to the table of

an automatic feeding machine is passed forward into the
mortar.
A typical automatic feeder is Hendy's Challenge Feeder,
which consists of a circular tray inclinedtowards the
mortar at an angle of about 12| degrees and slowly revolved
by bevelled gearing placed below it. The ore is delivered
into a sheet-iron hopper and slides thence on to the tray,
the amount being regulated by an adjustable door. When
the level of the ore on the dies in the mortar is low enough,
the tappet of the stamp-stem strikes a bumper-rod which
moves the bevelled gear-wheel and partly rotates the tray.
As the tray rotates, some of the ore is scraped off by
stationary side-plates and falls into the mortar. One
feeder is usually supplied to each battery of five stamps.
The foundations of astamp battery are of the first
importance, for if
they are badly made, the battery shakes
itself to pieces. A trench is dug out to receive the vertical
timbers or mortar blocks, which usually rest on concrete,
and are surrounded by concrete or masonry. Sometimes
the mortar blocks consist of concrete or iron instead of
wood. The mortars rest on and are bolted to these blocks,
88 THE PEECIOUS METALS

but are separated from them by sheet-rubber or blanketing,

toreduce the vibration.
The mortars themselvesare of cast iron, usually cast in
one piece, and vary considerably in width and in depth of
discharge that is, in the vertical distance from the bottom
of the screen to the top of the die. If the depth of

discharge is low, the ore is thrown more directly and with

greater "violence against the screens than if the depth of
discharge is high. The result is that the screens wear out
sooner, but, on the other hand, the capacity of the mill is

increased and the discharged pulp is coarser. The shorter

time during which the ore remains in the mortar when the
depth of discharge is low, checks amalgamation inside the
battery ;
but the practice is now growing, and finds general
acceptance on the Band, of adding no mercury in the
mortar. In this case the battery proper is used as a
crushing machine only, and not as an amalgamator. This
method is a reversion to the older practice which was
universal before 1850.
The mortars are lined with plates of chilled iron or steel
to takeup wear and preserve the mortar. The lining
plates are about 1 inch thick and last for some months.
The open top of the mortar is covered with planking which
surrounds the stamp stems, and is provided with holes for
feeding-in the water.
The screen-openings are usually at the front of the mortar
only, as in Fig. 18, and are more rarely provided at both
front and back, when maximum crushing capacity is
desired and inside battery amalgamation is not practised.
The screens are carried in wooden frames and consist either
of iron plates with holes or slots cut in them or of wire
 TEEATMENT OF GOLD OEES BY AMALGAMATION 89

cloth. In the latter case the proportion of the opening to

the area of the screen is greater than in punched plates, so
that the crushed ore is more readily discharged from the
battery. On the other hand, wire screens have a shorter
lifethan plates. Sheet-iron or steel plates or wire cloth
are in wide use, but suffer very rapid corrosion if the water
in the battery is acid from the oxidation of pyrites in the

ore. This is especially the case when the water used in

the battery is pumped from a mine. The acidity can be
neutralised by the addition of quicklime, and it is not so

rapid in its action on brass wire screens or on aluminium

or phosphor bronze, which are sometimes used. Eichards
1
cites a case where aluminium bronze cloth lasted seventeen
weeks while Eussia iron lasted only three weeks.
The slope of the screens is generally 10 degrees from the
vertical. Gravity thus assists the splash in passing the
grains of ore through the screen. Outside the screen is
fixed a canvas shield or a splash-board of wood to stop the

splashes caused by the fall of the stamps. The crushed ore

and water, striking the canvas, drips down and is enabled to
flow evenly on to the
amalgamated plates.
The cam-shafts, stamp-guides, etc., are supported by a
framework usually consisting of heavy timbers, which are
better able to withstand the jarring than iron or steel.
The guides are required to keep the stamps vertical in spite
of the side-thrust of the cams. Sectional guides are best,
so that those belonging to any single stamp can be renewed
when worn out. The cam-shaft is made long enough to

" Ore
Dressing," Vol. I., p. 178. This case is quoted from a paper
1

by T. A. Eickard. Trans. Am. Inst. Mmj. EMJ., Vol. XXIII. (1893),

pp. 137, 545.
90 THE PEECIOUS METALS

carry the cams for ten stamps. The lifting faces of the
cams are made in the form of the involute of a circle
modified at the end so as to allow the upward motion of
the stamp to stop gradually. The radius of the circle is
equal to the distance between the centres of the stamp stem
and the cam-shaft. The cams rotate the stamp as well as
lift it, and this gives a more even wear to the head than
would otherwise be the case. Cams are fastened to the
cam-shaft by means of tapered keys which are driven into
slots placed opposite each other in the shaft and cams.

Cams are removed by driving out the keys, but there are
self-tightening cams, such as the Blanton cam, by which the
operation of replacing a broken cam is made much less
tedious. The Blanton cam is fastened to the shaft by a
tapered wedge which wraps round the shaft and is kept in
position by two pins. When the cam lifts the shaft it slips
up the inclined plane of the wedge and makes itself tighter.

loosened by being knocked backwards.

It is

The cam-shaft is driven by a belt and pulley, and as

the stamps are dropped alternately it is not difficult to

arrange that the strain on the cam-shaft shall be nearly

constant. The order of drop of the stamps in a battery is

often 1, 3, 5, 2, 4, but other orders are in use. The dis-

tribution of ore on the dies depends partly on the order of

drop, the blow of a stamp throwing ore upon the adjacent

dies in readiness for the next blows. The wave of water
must be prevented from heaping ore on certain dies and
leaving others bare. It is laid down that two adjacent

stamps should not fall in succession, and that when a stamp

falls its neighbours should be rising.
The jack, or finger, is a bar of wood which is pivoted below
 TREATMENT OF GOLD ORES BY AMALGAMATION 91

on a subsidiary beam, and is used for hanging up the

stamp. When this has to be done, a piece of wood is laid

upon the cam as it rises, so that the stamp is lifted above its
usual level, and when the stamp reaches its highest point
the finger is pushed under the tappet and supports the

stamp out of reach of the cam. The stamp is brought

into action again by a reversal of this proceeding.
The wear of crushing falls upon the stamp heads and dies.
The tappets are moved on the stamp stems when required
by the wearing down of the dies, and false dies are also put
underneath worn dies to raise them and prolong their life.
In running a stamp-mill, the best thickness of ore to be
arranged for on the die when a blow is struck is such that
the ore be crushed without injury to the die.
may If the

layer of ore is too thick, part of the force of the blow is spent
in packing the particles together without crushing them.
If the layer is too thin, the stamp head clashes on the die.

The amount of ore crushed by a stamp in a day depends

on the nature of the ore, the size of the mesh of the
screens, the height of drop of the stamp and the number of

drops per minute, as well as on the weight of the stamp.

Clayey ores are troublesome to crush, and soft friable ores
are crushed most easily. Fast-running stamps cannot be
given so high a drop as slow-running ones. Time must be
given to raise the stamp, let it fall and allow a short period of
rest before beginning to lift it again. From D. B. Morison's

appears that the maximum safe lifting

1
researches it

velocity of a stamp is 31*9 inches per second, and that about

0'15 second must be allowed, after the stamp has completed
1
N. E. Coast Inst. Eng. and Shipbuilders, Vol. XIII, April 30, 1897.
Richards' " Ore Dressing," Vol. I., p. 216.
92 THE PRECIOUS METALS

its descent, for rebound and repose, in order that the rising
cam may not strike a falling tappet. These figures afford

data for calculating the maximum number of blows possible

forany given height of drop. With a fall of 5 inches, 115*2

drops per minute may be given, a fall of 6 inches corresponds

to 105'4 drops per minute, a fall of 7 inches to 97'6 drops

per minute, 8 inches to 91'3 drops per minute, and 9 inches

to 85*5 drops per minute. These are the usual heights of
drop, and the numbers of drops per minute arranged for in

practice are not far short of the theoretical maxima given

above.
The amount of water supplied in a stamp-mill is propor-
tioned to the ore as the occasion requires. A larger
proportion of water corresponds to more rapid stamping
owing to the discharge being facilitated. When inside
amalgamation is practised, rapidity of discharge is not the
only point to be considered, and the ore treatment outside
the battery also limits the amount of water to be used. In
a number of cases cited by Eichards
l
the water used varied
from 2*92 to 15*97 tons of water per ton of ore stamped.
The average was 6'68 tons of water per ton of ore.
Mercury is still often sprinkled into the battery. The
amount depends on the quantity of gold and silver contained
in the ore. The average quantity added is about 1J oz. to

each ounce of gold contained in the ore. If too much

mercury is added the outside amalgamated plates become

soft with excess of quicksilver. The use of amalgamated
and the sprinkling of mercury in
plates inside the battery
the mortar, which originated about 1850, have of late years
been tending to fallinto disuse, especially since regrinding
1 " Ore
Dressing," Yol. I., p. 222.
 TEEATMENT OF GOLD OEES BY AMALGAMATION 93

in tube-mills has been introduced. On the Band goldfield,

battery amalgamation has been entirely discontinued.

The "pulp," or crushed ore and water issuing from the
battery screens, is led in a very thin stream over inclined
tables where the gold is caught. These tables were
formerly covered with blanket material, in which the heavy
particles of gold, pyrites, etc., became entangled. The
blankets were taken up and washed at intervals. About
the year 1860, however, the blankets began to be replaced by

amalgamated copper plates, the use of which became almost

universal soon after 1870. Similar copper plates had been

placed inside the mortars about 1850, and the practice of

feeding mercury into the mortars with the ore and water
began at about the same time. The gold amalgam accumu-
lates on the copper plates, and the stamps are occasionally

stopped and the amalgam is scraped off and squeezed in a

filter-bag to separate the excess ofmercury. The pasty
amalgam, containing from 25 to 50 per cent, of gold, is then
placed in a retort and heated gradually to redness. The
mercury distils and is condensed by means of water, and
the residue of gold is ready to be melted.
The amalgamated plates are made of the purest and
softest copper, usually Lake or electrolytic copper. It is
annealed in order to make the surface porous and so enable
ifc to absorb mercury readily. In annealing, the plate is
supported at a little distance from the ground and a fire
kindled underneath. The plate is made hot enough to
char paper placed on it, and is then allowed to cool. Any
buckling is removed by the blows of a mallet, and the
plate is cleaned by scrubbing with sand and ammonia or
alkali. Cyanide of potassium is even more often used for
94 THE PEECIOUS METALS

removing the oxide of copper, grease, etc., which would

prevent the mercury from adhering to the surface of the

copper. The cyanide may be painted on the plate and
washed off after a day, or less, or it
may be used in con-

junction with a brush and some fine sand. Plates are also
cleaned with emery cloth, but however prepared they are
liable to become tarnished, and so to make amalgamation
difficult. The clean plate is sprinkled with mercury and
rubbed with a cloth, sal ammoniac or cyanide of potassium
solution being added to dissolve any oxide of copper that

may make its appearance. The mercury then coats the

plate, giving it a uniformly bright silvery appearance.
Gold amalgam or silveramalgam is rubbed on to give
the plate a better "catching" surface, and the plate is
1
ready for use. According 16 oz. mercury and
to Eoskelley,

6 oz. gold amalgam are enough for the preparation of a

plate 16 feet by 4 feet 9 inches.

Copper plates are often electro-plated with silver before

being amalgamated. This prevents the appearance of green

spots, the so-called "verdigris," due to the oxidation of

copper, which greatly reduces the value of the plate as a
gold-saver. The plates are prepared for electro-plating
by coating the back with asphalt varnish and cleaning the
face with acid, followed by scrubbing with pumice powder
and washing with cyanide of potassium. The plate is
final

then immersed in a bath made by dissolving chloride of

silver in cyanide of potassium, and a current of electricity

passed from silver anodes to the copper plate as cathode.

The bath contains about 3 oz. of silver per gallon,
and the plating is complete when about 1 oz. of silver
1
Jour. Chem. Met. and Mmj. Sac. of S. A., February, 1904, p. 286.
 TREATMENT OF GOLD OEES BY AMALGAMATION 95

isdeposited per square foot of cathode. The plate is then

amalgamated as before. Electro-silvered plates are more
easily kept in good condition than plain copper ones.
Amalgamated plates are supported on wooden tables, to
which they are fastened by copper screws or by buttons or
wedges at the sides to make them easily removable. A
number of plates are laid on each table, either overlapping

or laid side by side. The main catching plates are called

apron plates, or simply amalgamated tables. They are
laid on a slope to enable the pulp to run over them by

gravity, the inclination varying from about 1 to 2J

inches per foot. If the slope is too great the pulp runs

over it too fast, and the gold has less opportunity of

settling down and coming into contact with the plate. If

the slope is insufficient, sulphides accumulate on the plate

and interfere with its proper working. The variation of the

slope is due to differences in the ore, pyritic ores requiring

more slope than those nearly free from sulphides.Coarsely

crushing batteries with a high output require a high angle
of inclination of the plates, but the greater the proportion

of water used, the less the slope.

When pulp flows over the plates, the gold settles through
the stream, and coming in contact with the amalgamated
surface, adheres to it. It would appear from this that the

stream should be as thin as possible, and should flow

smoothly and uniformly without any tumbling action. In
practice, however, these conditions are probably unattain-
able. The discharge through the screens is intermittent,
and the pulp flows down in a series of little waves which
tumble over and over and bring the grains in the pulp
all

in contact with the plate irrespective of the action of

96 THE PEECIOUS METALS

gravity. With the idea of increasing the tumbling action,

many amalgamated tables are broken
up by steps or drops
placed transversely, so that the pulp falls from the lower
edge of one plate upon the upper end of the next plate
from a height of from 1 to 3 inches. If the fall is of the
correct height, the amalgam builds up under it faster than

anywhere gold being caught there. If the fall is

else, float
too great, the amalgam is scoured away instead of being
saved. These drops act as distributors of the pulp, and
are sometimes placed every 2 feet, but are often entirely
omitted. In the latter case the area of the plates is

made greater.
The apron plates are usually as wide as the mortar is
long, and their length varies from about 6 to 21 feet.
When sluice plates, or narrower plates below the apron

plates, are used, the apron plates may be even shorter, but
in sluice plates the stream of pulp is necessarily deeper,
and they are falling into disuse.
1
The depth pulp on the plates is given by Eichards
of the

at from 0'19 inch downwards. At the trough of the wave

it may be as little as 0*014 inch at the lower end of the

plates. He
gives the velocity of the waves at from 19 to
42 inches per second.
Mercury traps are sometimes placed between the plates
instead of drops. In one form they consist of simple
transverse troughs or riffles, 2 or 3 inches wide and

deep. Sometimes they are made wider and deeper, and

partitions or gates are placed in them, under which the
pulp passes. The intention is to catch mercury and
amalgam and to let the sand and pyrites pass through.
1 " Ore
Dressing," Vol. II.., p. 746.
 TREATMENT OF GOLD ORES BY AMALGAMATION 97

Sometimes they have a pool of mercury placed in them,

when they become mercury a gate is
wells or riffles. If

placed in the well almost or quite touching the surface of

the mercury, the pulp in escaping must pass over the
surface of the mercury, or even through it. A mercury
well acts as a trap for mercury, but has little other

advantage to recommend it, and such wells are not now

used much. They are useful in catching very coarse free

gold, but large particles are not likely to escape whatever

method of gold-saving is adopted.

Amalgamated plates are in the best condition to catch

and retain gold and silver when they are clean and bright
and the amalgam is soft marked by the
enough to be

pressure of the finger, but not soft enough for the mercury
to gather into drops and run off. If the amalgam becomes

hard it tends to flake off under the scouring action of the

pulp, and fresh mercury is sprinkled on, or the amount of

mercury fed into the battery is increased. On the Rand 1

mercury is sprayed over the upper 3 feet of the plates

(the part that is scraped daily) every four hours, when the
plates are dressed, and this mercury oozing downwards
supplies the lower part.
Dressing is necessitated by the discoloration of the
plates, anddesignedis to make them more active in
catching gold. In dressing, the stamps are hung up, the
water turned off and the sand brushed off the plates. If

they are discoloured they are cleaned with potassium

cyanide. On the Rand mercury is sprayed over first, and
then a solution containing 0'08 per cent, of cyanide is
sprinkled on, and the mercury is rubbed in with a hard
1
Eoskelley, C/tem. Met. and Mng. Soc. of 8. A., Vol. IV. (1904), p. 286.

P.M. H
98 THE PRECIOUS METALS

brush. Afterwards the amalgam is brushed smooth with

a soft brush, and the plates are finally washed do\\n.
Cyanide is used to a less extent than formerly, as it makes
the mercury drain away from the amalgam and so tends to
cause the plates to become hard and dry.
Cleaning-iip, or the removal of amalgam from the plates,
takes place at intervals, which may be as short as two hours
if extraordinarily rich ore is being treated, or as long as a
month. A daily clean-up is very common. In cleaning-up
the amalgam is softened by the addition of mercury well
rubbed in, and the amalgam is then removed with a scraper
of rubber or iron, care being taken not to scratch the

plates. A thin layer of amalgam is left on the plate to

form a good catching surface, and the plates are then

dressed. Hard amalgam accumulates on the plates and
cannot be entirely removed even by means of a chisel. It
isremoved at long intervals of two months or more by long
" "
scraping, or by steaming or sweating the plates before

scraping, or by buckling the plates with a hammer, a pro-

cedure which causes the amalgam to strip off. Old worn-
out plates are scaled by being heated over a fire and

plunged into water, or they may be dissolved in acid or

melted up and sold to a refinery.
The dirty amalgam obtained from the plates and battery,
containing sand, sulphides, metallic iron and base metal
amalgam, mixed with more mercury and treated in a
is

rotating iron clean-up barrel with water and cannon balls.

The water is then poured off and the mercury squeezed
through chamois leather or canvas. The squeezing is often
by hand, but hydraulic amalgam-squeezers are also in use.
The amalgam retained by the filter contains from 10 to
 TREATMENT OF GOLD ORES BY AMALGAMATION 99

50 per cent, of gold and silver, coarse gold yielding richer

amalgam than finely divided gold. Hard amalgam contains
more gold and silver than the soft amalgam obtained in the
usual way by cleaning the plates.
Retort in fi is carried out in cast-iron retorts, of which
there are two kinds, the pot retorts and the cylindrical
retorts. The pot retorts are smaller than the others and
are placed on any fire. The cylinders are placed hori-
zontally in special furnaces. Both kinds have a delivery
pipe leading to a Liebig's condenser. The amalgam is
squeezed into ball- or cake-shaped masses and placed in the
retorts on trays, or simply on paper or lime-wash linings.
The fire is lighted after the retort is in position, and at

first a gentle heat is used which is increased gradually.

The mercury is volatilised, and after condensation in the

cooled pipe is received in a vessel of water. When no more

mercury comes over, the retort is heated to redness for a

few minutes, and is then allowed to cool. The residue in

the retort contains gold, silver, base metals, sulphides and
some mercury. It is cut up, if necessary, and melted in
graphite crucibles with borax, soda carbonate, nitre and
sand. The base metals are partly oxidised by the nitre
and partly left in the molten gold and silver. The
metallic oxides form a fusible slag with the borax and sand,
and the carbonate of soda takes up sulphur from the sul-

phides, and makes the slag more liquid. When the charge
is melted it is well stirred and poured into ingot moulds,
which are overturned into water as soon as the charge is
solid in order to detach the metal readily from the slag.

The fineness of the bullion and the methods of sampling

are dealt with in Chapter XIV.
H 2
100 THE PRECIOUS METALS

Fig. 19 shows a general view of a stamp battery in

operation. It is a reproduction of a photograph (kindly

supplied by Mr. W. A. Caldecott) of the interior of the

 TREATMENT OF GOLD ORES BY AMALGAMATION 101

Simmer East Mill, Witwatersrand, Transvaal. It shows

the guide-posts, the guide-timhers, the stamp-stems, cams
and tappets, the cam-shafts and the cam-pulleys. The
stamp heads are hidden inside the mortar boxes. The
screens are shown and the amalgamated tables in front of
them. The weight of the stamps is 1,550 Ib. in this

battery, and the amount of ore crushed per stamp is about

6*3 tons per day of twenty-four hours.

On the Eand the ore is sorted by hand before delivery to

the stamp-mill and the waste or barren rock picked out and
discarded. In Fig. 20 l the sorting table and trommel at
the Knight's Deep and Simmer East joint plant are shown.
The trommel delivers large pieces of ore on to the rotating

tables, where it is sorted.

The ordinary stamps described above are known as

gravitation stamps or
gravit}^ stamps. There are other
forms of stamp in which power in addition to the weight of
the stamp is used to deliver the blow. Steam stamps are
lifted and forced down by a steam piston. In pneumatic
and spring stamps the power is applied by a crank with an
air piston or a spring to take up part of the shock to the

machine. These forms of stamp are not much used in

crushing gold or silver ores.

A number of other pulverisers are also in use. Arrastras
and amalgamating pans are described in Chapter XL, which
deals with the treatment of silver ores. Ball-mills and
roller-mills are more often applied to the crushing of gold
ores. Like stamps they can be used for either wet or dry
crushing. The Krupp ball-mill is widely used, and is a very
1
This figure is also from a photograph kindly supplied by Mr.
W. A. Caldecott.
102 THE PBECIOUS METALS

efficient machine. It consists of a cylindrical drum con-

taining a number of chilled-iron or cast-steel balls of

 TEEATMENT OF GOLD OKES BY AMALGAMATION 103

different sizes, and is rotated on a horizontal axis. The ore

is crushed by the grinding of the balls against each other

and against the inside of the periphery of the drum, which

is composed of five or more overlapping steel plates so

arranged that the balls drop a few inches in passing from

one plate to the next. The crushed ore passes out between
the plates on to a screen, and the oversized particles are
returned to the interior of the mill. The Krupp mill is

applicable to soft ores, especially in preparing them for

treatment by the cyanide or chlorination process.

Among roller-mills, the Huntinyton mill has perhaps been

more used than any other. It is applicable to the crushing
of soft gold ores as a substitute for the stamp battery. It

consists of a pan, inside which an iron ring revolves in a

horizontal plane, and hanging from the ring are three
stems, each having a roller attached to its lower extremity.
When the ring rotates, the stems swing outwards by centri-

fugal force, and the rollers press against and roll round
inside a ring-die near the bottom of the pan. The ore,
previously broken to less than |-inch pieces, is fed into
the mill with water and is crushed between the rollers and
the ring-die. Mercury is used in the mill, and the pul-
verised ore passes out through peripheral screens placed
above the ring-die, and is allowed to run over amalgamated

plates. The Huntington mill is much used in America,

and with careful feeding gives excellent and economical
results.

Amalgamating machines, designed do the work gene-

to

rally assigned to amalgamated copper plates, are numerous,

but not so much used as formerly. Amalgamating pans
are referred to elsewhere. In Transylvania the old system
104 THE PEECIOUS METALS

of crushing in stamp batteries arid amalgamating in

separate machines is still followed to a considerable extent.
The so-called Austrian theory of amalgamation is that gold-

saving is best effected by giving the particles of metal an

opportunity of settling by gravity through the pulp and of

coming into contact with the surface of mercury lying
underneath. In the old Tyrolean bowls described by
Agricola in 1556, the crushed ore was kept in agitation by
revolving stirrers above a bath of mercury, and according
1
to him the machines "separated even very minute flakes
of gold from the crushed ore. These flakes, settling to the
bottom, are drawn to itself and cleansed by the quicksilver
(lying in the tubs), but the water carries off the dross."
The modern forms of Hungarian amalgamators, such as the
Schemnitz mill and the Lazzlo amalgamator, resemble the old
bowls. They have revolving mullers which force the ore to
come in contact with the pool of mercury, but do not, as in

ordinary amalgamation pans, stir the mercury itself so as

to mix and grind it with the ore.

Shaking amalgamated plates have lately been used to a

great extent for the treatment of reground ore coming
from tube-mills in the Transvaal. They are sometimes
placed on old True vanner frames, and are given a slight
rapid side-shake of about 200 throws per minute of half
an inch or so each. The pulp
kept by this device from
is

settling on the plate, so that the inclination of the plate

can be reduced to about half an inch per foot, or even less.
Silver-plated copper is the material usually employed. A

longitudinal shake is sometimes given instead of a side-

shake. The width of the plates is made as great as possible

1 " De Ee Metallica," Lib. VIII., p. 233.

 TEEATMENT OF GOLD OEES BY AMALGAMATION 105

to reduce the depth of the film of pulp flowing over them,

but their length need not be very great, as they are efficient
gold-catchers. Since the introduction of regrinding and
treatment with shaking plates, the proportion of gold
extracted by amalgamation from ordinary Rand banket
ore has greatly increased, and is now nearly 70 per cent,
on a number of the mines. The extraction by means of

cyanide has suffered a corresponding diminution.

The loss of mercury in amalgamation varies with the kind
of ore as well as with the method of treatment. The chief
cause of loss isthe flouring or sickening of mercury that
is, its subdivision into minute globules which are too small
to settle in mercury traps, etc. The globules of mercury
are prevented from coalescing by a film of some foreign

material, such as grease, oxides of base metal, sulphides,

arsenides, hydrous silicates, etc. The loss by flouring is

increased by using impure mercury containing amalgams

of base metals, which are readily oxidised. It is also
increased by stamping grinding ore and mercury
or

together, especially when the ore contains arsenides, anti-

monides, etc., or soapy materials, such as clay, serpentine,
talc, or other hydrous silicates. Base oxides are removed
from quicksilver by sodium amalgam, but as the base
metals re-enter the mercury the improvement is only
temporary. Apart from flouring, a considerable loss of
mercury takes place in the pan-amalgamation process, and
also in the patio process, from the formation of sulphates
and chlorides of mercury. The loss owing to this cause
is trivial in the stamp battery. In the treatment of gold ores
1
in stamp mills the loss of mercury is given by Richards
1 " Ore 206.
Dressing," Vol. I., p.
106 THE PRECIOUS METALS

as 0*339 oz. per ton of ore crushed. This was the average
at sixteen American mills, but it is much less in the
Transvaal, where mercury not charged into the mortar.
is

Mercury is purified by distillation with lime and iron

filings. An addition of charcoal prevents the formation
of volatile oxides.Floating impurities may be removed by
running quicksilver through a funnel and rejecting the
last portion. The most effectual method of purifying
mercury is to cover it with dilute nitric acid and force air
through the mercury. The base metals are then oxidised,
and rising to the surface are dissolved by the acid.

TREATMENT OF TAILINGS FROM THE STAMP BATTERY.

After treatment in the stamp battery, the crushed ore
and water flowing from the amalgamated plates still con-
tains some of the gold, either embedded in the pyrites or

locked-up inside coarse grains of sand, or even in the form

of finely divided free flakes of native metal. The treatment
of these "tailings" for the extraction of the remainder of

the gold varies according to the nature of the ore.

The older method of treatment was generally some form
of concentration, the heavier particles being separated from
those of lower density and saved, whilst the lighter particles,

being comparatively poor in gold, were allowed to be carried

away in the stream of water and lost. Blankets were
stretched on sloping tables, and the pyrites, grains of
gold, etc., which adhered to the fibrous surface were
washed off every now and then and sent to a smelter or
amalgamated in grinding pans or chlorinated. Canvas
tableshave been substituted for blankets, and sloping
wooden tables were also used, but some form of " vanner "
 TREATMENT OF GOLD OEES BY AMALGAMATION 107

or shaking table has now generally superseded these

devices.
The Frue vanner is the best known of the vanners. It

consists of an endless rubber belt, slightly inclined to the

horizontal, and subjected to a rapid side-shake. The belt

has a continuous slow motion up the slope. The pulp
coming from the amalgamated plates is fed on to the upper
end of the inclined table formed by the belt, together with
some additional water. The pulp runs down the belt in
a thin stream, and the particles of high density settle to
ths bottom of the stream of pulp, and, adhering to the
surface of the belt, are carried up against the stream and
over the top. They are washed off into a tank by means
of jets of water. The side-shake keeps the ore loose, and
so assists the heavy minerals to settle to the bottom.

Finely divided sulphides are saved on this machine even

better than very large particles, which tend to roll or slide
down the belt with the worthless material. Two Frue
vanners are generally sufficient to treat the product of a
battery of five stamps.
The Wilfley table has no travelling belt, but is provided
with a number of wooden strips nailed on the surface of
the table. The lighter grains are washed down across
these wooden riffles, and the sulphides, etc., of any size

whatever, travel along the riffles under the influence of a

side-jerk and over the side of the table. The Wilfley

table is stated not to be of the highest efficiency in saving
fine slimes, and is sometimes
supplemented by other
machines. It has largely replaced other forms of shaking
tables and vanners of late years.
For certain purposes it is desirable to classify the pulp
108 THE PEECIOUS METALS

according to the size of the particles before other treat-

ment. This may be done by inclined shaking or revolving
screens or by the use of hydraulic classifiers, which generally
consist of some form of pyramidal box or inverted roof-
shaped trough. The apex of the box or trough is, 'of
course, downwards. A stream of pulp enters at one side
of the box, which is full of water, and overflows at the
other side. The current slackens when it enters the box,
and the larger and heavier particles of ore settle in the
box and accumulate in the apex, where they are discharged
either continuously or intermittently. Successive boxes
are used, each larger than the previous one, so that two
or three or more grades of material are collected, besides
the slimes which do not settle in such boxes. In order to
clean the separated classes of material an ascending current
of clear water is sometimes introduced at the apex. The
sand falls through the upward current, but cannot carry
down any slimes with it. The slimes can be collected in
large pits or tanks.
Classified material is better fitted than unclassified pulp
for some methods of concentration, such as jigging, and
also for treatment in leaching vats, as, for example, by
the cyanide process. The concentration of gold ores is,

however, less generally practised than formerly now that

the cyanide process has come into almost universal use.
Concentrates usually amount to 1, 2, or 3 per cent,
of the original ores. They consist of sulphides, oxides of
iron, flakes of gold, coarse grains of sand, etc., and may be
very rich. Their values can often be extracted in great part
by grinding with mercury in iron pans. The ordinary
amalgamating pan described in Chapter XI. answers the
 TREATMENT OF GOLD OEES BY AMALGAMATION 109

purpose sufficiently well, but its capacity is often too great

for the small quantities of concentrates collected, and a
smaller type of pan is used. This method, however, is

not now very widely used. When amalgamation of con-

centrates is not adopted, they may be roasted and chlorinated
or simply cyanided. Extremely rich concentrates can be
treated by smelting with lead in reverberatory furnaces,
followed by cupellation, but finely divided material is not
suitable for reduction in blast furnaces, so that the smelting
of concentrates is exceptional.
 CHAPTER VIII

TREATMENT OF GOLD ORES BY WET METHODS

THE CYANIDE PROCESS.

AT the present day the usual method of treating the

tailings from stamp battery amalgamation is by means of

the cyanide process without any previous concentration.

Eough classification of the pulp is, however, necessary.

The tailings are run into large vats, from which the surplus
water overflows, carrying with it the " slimes," or fine mud.
The vat may be originally filled with clean water, in which
'case the slimes overflow from the beginning, leaving the
sands to bottom, or some kind of rising
settle quietly to the

gate may be used, beginning with an empty vat and raising

the height of the overflow from time to time as the sand
accumulates. When the vat is filled with sand, it
may be
drained and discharged into trucks and the sand conveyed to
another vat for leaching with cyanide. This is the method
of intermediate filling with the use of a collecting vat, which
may be of any depth. The cyanide vat may of course be
placed vertically below the intermediate filling vat to
facilitate transference. As an alternative, direct filling

may be employed, the sands being collected in the cyanide

vat itself and leached without being disturbed.
In leaching, dilute solutions are used, usually containing
from 0'05 to 0'35 per cent, of cyanide of potassium. The
 TREATMENT OF GOLD ORES BY WET METHODS 111

sand rests on a false bottom placed inside the

vat, and con-
wood covered by cocoa-nut fibre
sisting of a filter-frame of
matting and hessian, or some other kind of filter-cloth.
The vats are round and contain as much as 400 tons of ore.
The cyanide solutions are either run on from above, or are
introduced below the false bottom and rise up through
the charge. When the ore contains partially oxidised
"
pyrites, which act as cyanicides," decomposing the

cyanide, the acidity may be removed by treatment with

caustic lime. This may be mixed with the ore before it

is crushed, or an alkali wash of lime or soda may be

applied to the ore in the vat, before the cyanide solution

is run on.
Two strengths of solution are often applied in succession,
"
a
"
strong solution and a
"
weak " solution. The strong
solution is usually left on for some time before being
allowed to run through, but the solution is not pumped
back again to the surface of the ore. Three or four
additions of solution are usually sufficient, the time of

treatment by the strong solution being often one or two

days. The solution is sometimes applied continuously,

that is to say, as soon as the liquid has sunk below the
surface of the ore, fresh solution is run on. Sometimes
the solution allowed to drain away completely before the
is

next charge is applied. This causes air to be drawn into

the interstices of the ore, and the fresh oxygen enables

subsequent dissolution of the gold to be more rapidly

effected (see p. 29).
Some ores absorb oxygen and require to be re-oxygenated.
This difficulty has been met by the method of double treat-
ment, the ore being treated with cyanide solution in one
112 THE PRECIOUS METALS

vat, then allowed to drain, and removed to another vat and

again leached. The turning over and exposure to air of
the charge during transference affords an opportunity for
its thorough oxygenation. The treatment of concentrates
is made owing to the absorption of all
especially difficult
free oxygen by the sulphides. In the Transvaal, concen-
trates formerly often underwent cyanidation during three
or four weeks, the charge being drained and spaded over

frequently to facilitate aeration between the applications of

cyanide solution.
When the strong solution has been drawn off, the weak
solution is run on, and finally a number of water washes
are applied, each wash being small in amount. Washing
is discontinued when cyanide can no longer be detected in
the water issuing from the vat.
The vats are discharged through doors in the bottom
into trucks which are often hauled up inclines and emptied
on the top of a heap of tailings. The heap thus grows
higher and higher as its area is extended. The sand in
the vat is sampled for assay before being discharged.
The auriferous solution drawn from the charge is passed
to the zinc-boxes for the precipitation of the gold. These
are usually troughs containing shavings of zinc coated with

precipitated lead, formed by dipping the shavings into a

strong solution of acetate of lead. The boxes are divided
into compartments by partitions, and the solutions generally
flow upwards through each compartment in succession, the

overflow from one compartment being conveyed through a

small passage to the bottom of the next. The zinc shavings
are very thin and light, and are carried on iron gratings

through which the solution passes. The gold is precipitated

 TEEATMENT OF GOLD ORES BY WET METHODS 113

on the zinc which displaces it in solution according to the

equation
2KAuCy 2 + Zn = K ZnCy + 2Au.
2 4

The lead forms a galvanic couple with the zinc and

strengthens its action as a precipitant, but the lead itself
remains unchanged. The zinc is, however, also dissolved
by the cyanide solution direct, potash being formed and
hydrogen set free, A white precipitate tends to form on
the zinc, checking its action (see above, p. 31). It con-

sists in part of cyanide of zinc, which would be dissolved

by a large excess cyanide of potassium. If the per-
of

centage of free cyanide of potassium in the solution is

too small, the gold is incomplete or is
precipitation of
entirely stopped. The gold forms a black mud on the surface
of the shavings, which disintegrate as the action proceeds.
In cleaning-up, the shavings are washed and filtered in
order to separate the residue of zinc and soluble salts, and
the sludge charged into a large tub with dilute sulphuric
is

acid or bisulphate of sodium. Finely divided zinc is dis-

solved and the cyanides decomposed with violent frothing,

owing to the disengagement of gas. After all action has

ceased the charge is allowed to settle and the clear liquid
decanted off. The slimes are then washed with hot water
and separated by a vacuum filter or a filter-press. They
may then be dried, calcined and fused with fluxes in

graphite crucibles. The fluxes used are soda carbonate,

borax and sand. Nitre or manganese dioxide are also
added sometimes to oxidise some of the base metals and
produce purer bullion.
Another method, the Tavener process, consists in mixing
the slimes, after filter-pressing, with fluxes and litharge, and
P.M. i
114 THE PKECIOUS METALS

smelting for auriferous lead in a reverberatory furnace.

Sawdust is used as a reducing agent. The slags are very
poor, and the lead may be regulated so as to contain as
much as 8 per cent, of gold. This is cupelled in a second
operation in an English cupellation furnace. The cupelled
gold is allowed to solidify on the test, and is then detached,
broken up while still hot, and melted in graphite crucibles
for casting into bars.

The treatment of the slimed ore separated from the sands

before the latter are cyanided differs considerably from the
method described above. In the Transvaal the decantation
process is in general use. Milk of lime is added to the
water from the separating plant containing the slimes, in
order to coagulate them and assist them to settle, which

they do in a flocculent form when lime is added. Large

pointed boxes are used for settling the slimes, which are
drawn off at the bottom and pumped to the treatment
vats, about 90 per cent, of the water contained in them
having been separated. The cyanide solution used to

dissolve the gold contains about 0*01 per cent. KCy.

Dissolution is rapid and is aided by agitating the charge,
which done by means
is of pumps. The slimes are then
allowed to settle, and the clear liquid is withdrawn and sent
to the zinc precipitation boxes. Slimes cannot be washed by
the ordinary method of percolation, as they resist the passage
of liquids through them, even when a vacuum is created
underneath the They are accordingly washed
filter-bed.

by decantation, the vats being repeatedly filled up with

fresh water with which they are stirred. The charge is
then allowed to settle and the supernatant liquid removed
by decantation until it is too poor to pay for treatment.
 TREATMENT OF GOLD ORES BY WET METHODS 115

The solutions are poor, both in gold and in cyanide, and

are treated either by passage through zinc-lead shavings
or by electrical precipitation, in which the smallness of
the percentage of cyanide present is not disadvantageous.
The cathodes consist of sheet-lead and the anodes of iron
plates or of sheet-lead coated with peroxide of lead. The
current used is not great, in order to avoid the destruction
of cyanide as far as possible. The gold gilds the lead, or is

deposited in a black granular form, and to recover the gold

merely necessary to melt and cupel the lead, which con-

it is

tains from 2 to 12 per cent, of gold. The electrical method

of precipitation was introduced as the Siemens-Halske process.
In Western Australia slimed ore is treated by filter-

pressing instead of by the decantation process. The pulp

is usually partly freed from water by settlement or by a

niter-press, and is then treated with cyanide solution in.

vats by agitation applied by means of paddles. The
mixture is then pumped into filter-presses and the liquid
forced through the filter-cloths under high pressure applied

by compressed air. Dissolution of the gold may also be

effected either
wholly or partially in
filter-presses by

forcing cyanide solution through them. In any case the

ore is washed clean in the presses. Filter-presses consist
of a series of parallel ribbed or corrugated iron flanged
plates tightly fitting together at their planed margins.
Each pair ofplates separated by filter-cloths, which
is

constitute the packing between the planed margins. The

liquid is forced through the cloth into the corrugations of

the plate and is drawn off in a series of channels. The
cakes of slimes are removed from the presses by opening
them individually. The whole operation of filter-pressing
i 2
110 THE PEECIOUS METALS

occupies only about two hours, so that the treatment is far

more expeditions than the method of decantation. The

precipitation of the gold from solution is effected as usual.
 TREATMENT OF GOLD ORES BY WET METHODS 117

Cyanide plants are now of great and very

capacity
complicated on the Band. A general view of a portion
of the reduction works (stamp battery and cyanide plant
combined) of the Knight'sDeep and Simmer East joint
plant is shown in Fig. 21. The battery house is on the
right of the picture.
Another view of the same plant is shown in the frontis-

piece, which is taken from the top of the enormous sand

heap or dump where the tailings, or ore after treatment, are
thrown.
In Fig. 22 a view is given of the interior of the extractor
house at the same works. On the right the zinc-boxes
divided into compartments are shown, and on the left a
and the pumping machinery.
filter-press
The slimes tailings cannot be piled up like the sands,
and are settled and caught in huge dams, one of which, the
Knight's Deep and Simmer East darn, is shown in Eig. 23.
Figs. 21, 22and 23 are reproduced from photographs kindly
supplied by Mr. W. A. Caldecott.
There are numerous modifications of the cyanide process
in use at different mines. The strength of the cyanide
solutions, the time of treatment, and the method of leaching

vary with almost every ore.

Latterly it has become apparent that the tailings from
stamp-mills are often too coarse to yield their gold without
previous regrinding, and they are accordingly reduced to
fine slime by being passed through tube-mills or other
1
machines. Tube-mills (see Fig. 24) are rotating iron
cylinders lined with flint or chert and half filled with
round flint pebbles, or more rarely with iron rods. The
1
Fig. 24 is from a photograph kindly supplied by Mr. W. A. Caldecott.
118 THE PRECIOUS METALS

tailings are fed in at one end of the cylinder, ground

between the pebbles and discharged at the other end. Any
coarse material which escapes from the mill is returned to
TREATMENT OF GOLD OEES BY WET METHODS 119
120 THE PRECIOUS METALS

it. A
photograph of one of the three tube-mills in use at the
Knight's Deep and Simmer East joint plant is reproduced
in Fig. 24.
The silica lining of tube-mills may be replaced by steel,
and in the Transvaal the flint pebbles have been discarded
in favour of banket ore, which consists of quartz pebbles

set in a siliceous cement. This material is found to answer

very well and contains a little gold which is set free as the

pieces of ore are ground up.

The speed of revolution of tube-mills is such that the

pebbles, carried up the side of the mill by the friction of

the lining, in part roll over one another and in part fall

freely through the air and strike the bed of pebbles at the
bottom. For a mill 5 feet in diameter the speed is about
twenty-eight revolutions per minute. The pulp is thickened
as much as possible by the removal of water before it is fed
to the mill. No screen is used at the discharge aperture,
but the product is classified in spitzkasten (pointed boxes),
and the coarser material is returned to the mill.
Since the introduction of tube-mills in the Transvaal
(where there were 103 at work in September, 1908) the
stamp-mill screens have been made coarser, so that the
output of the stamps has materially increased and the
tailings from the stamp-mills consist of particles of greater
average size than was formerly the case. Thus at the
Luipaard's Ylei Estate mill in August, 1908, the duty per
stamp was 9'67 tons of ore per day, and the average duty
on the Witwatersrand in September, 1908, was 6*49 tons
per day, while the average duty in the early part of 1904
was only 4'78 tons per day. Part of this change is,
however, due to the increase in weight of the stamps. The
 TREATMENT OF GOLD ORES BY WET METHODS 121

product of the tube-mills is of course finer than the former

product of the stamps.

122 THE PEECIOUS METALS

The line material from the mills is passed over shaking

amalgamated copper plates, which have already been
described. The extraction of gold by mercury on these
and on the ordinary stationary plates is as great as was
the case before the introduction of regrinding, so that
the material for cyaniding is poorer than formerly, and
the proportion of the gold obtained in the cyanide plant is
about the same in spite of the fact that, owing to the finer
grinding the tailings after being treated by cyanide are
poorer than before. "All-sliming" of ore appears to be
almost impossible. Even after very fine grinding in tube-
mills, the pulp can
be separated into easily leached
still

sands and slimes which cannot be leached by percolation.

The telluride ores of Cripple Creek, Colorado, and of
Kalgoorlie, West
Australia, cannot be treated successfully

by the cyanide process or by amalgamation without special

precautions. Telluride of gold is acted on by cyanide of

potassium only very slowly, and is not amalgamable at all.

Two methods have been used largely. In the Diehl process
the ore is crushed with water in stamp-mills, amalgamated,
concentrated, and the tailings separated by classifiers into
sands and slimes. The concentrates are dried, roasted and
sent back to the crushing mill. The sands are reground in

tube-mills, and the whole of the fine material is then

agitated with cyanide solution, to which some bromo-

cyanide is added, and filter-pressed. In the Marriner
process the ore is crushed dry in roller-mills or ball-mills
(q.v.) and roasted dead. The product is ground with mer-
cury and cyanide solutions in pans, and the tailings are
agitated with dilute cyanide of potassium and filter-pressed.
At Cripple Creek the telluride ores are crushed dry by
 TEEATMENT OF GOLD ORES BY WET METHODS 123

means of rolls, roasted and treated with cyanide by

percolation in large vats.
crushing has not been previously dealt with, but is
Dry
generally required as a preliminary to roasting or smelting.
Stamp batteries can be used, the discharge screens being
placed on both sides of the mortar instead of on one side
only, liolls are more generally employed. They are
revolving cylinders of steel or cast iron, placed with their
axes parallel and their peripheries a short distance apart.
Pieces of ores are drawn between two and crushed by
rolls

compression. Ore is generally passed through two or more

pairs of rolls in succession, the last pair having their
revolving faces placed close together. Kolls are not
adapted for very fine crushing, and are seldom used in
conjunction with screens finer than about 20-mesh. The
ore which will not pass the screen is returned for another

passage through the rolls.

Hoaating furnaces for use in conjunction with the cyanide

or chlorination process are of many different designs. The
old-fashioned hand-worked reverberatory furnace has now
almost passed out of use for such purposes. It consists of
a large hearth, on which the ore is spread, surmounted by
a low roof. The flame from a fire passes over the hearth,
heating the ore, which is stirred by iron tools or rabbles
introduced through doors at the sides of the furnaces. Air
is supplied through the working doors. The sulphur in the
ore is oxidised and partly removed as sulphur dioxide.

Automatic furnaces are now almost universal. In one

type, which has many forms, the rabbles are carried by
endless moving chains and the ore is stirred and pushed
forward towards the hot end of the hearth by the rabbles,
124 THE PRECIOUS METALS

and finally drops through a hole in the hearth near the

fire. In another type the ore is fed in at the upper end of

an inclined rotating cylinder, and, tumbling over and over

as it is carried up the sides of the cylinder, gradually finds
its way to the lower end, discharged. The flame
where it is

from a fire enters, together with air, at the lower end, and
passing through the cylinders escapes into the stack at the
upper end. In both types of furnace the inclination of the
hearth, the arrangements for moving the ore, and the heat
from the fire are adjusted so that the ore is roasted as

nearly completely as required before it is discharged.

is

The cyanide process, which was introduced in South

Africa in 1890, and has since spread to all parts of the

world, is now of the highest importance in the treatment of

gold ores. In the Transvaal 2,264,994 oz, of gold were

produced by use in 1907, as against 4,142,939 oz. by
its

means of amalgamation. The importance of the cyanide

process is even higher than might be deduced from these
figures, as few of the mines could be worked profitably if
the gold extracted by amalgamation represented their total

output. The same thing can be said of most of the gold-

fields in other parts of the world.

THE CHLORINATION PROCESS.

Chlorination has been to some extent displaced by the

cheaper cyanide process, but is still used for the treatment

of certain ores and concentrates, and is applicable in certain
cases in which cyanide is useless.

Ores must always be roasted before chlorination, to

prevent the waste of chlorine in acting on oxidisable
materials. The roasted ore is sometimes damped and
 TREATMENT OF GOLD OEES BY WET METHODS 125

subjected to the action of chlorine gas in closed vessels, the

soluble chloride of gold being afterwards washed out by
percolating water (Plattner process). It is now more usual
to treat the ore with a solution of chlorine in water, either
in vats open to the air (vat solution process) or in closed

rotating barrels (barrel process).

The Plattner process was that originally employed, and is

chiefly applicable to small quantities of rich concentrates.

These contain much pyrites and are roasted dead. If copper
pyrites is present, salt is added towards the end of the

roasting operation to chloridise the copper, which would

otherwise absorb chlorine in the vat. Chloridising roasting
of gold ores is, however, liable to be attended by a heavy

loss of gold by volatilisation. The roasted ore is damped

by sprinkling with water and thorough mixing, the amount
of water being kept low enough to allow the mass to be

porous, so that chlorine gas can pass through it readily.

The vat is lead-lined and has a filter-bed and false bottom,
and the gas,made by the action of common salt, manganese
dioxide and sulphuric acid in a separate generator, is
passed upwards into the ore through the filter-bed. When
the whole charge smells strongly of chlorine, the vat is
closed and allowed to remain for twenty-four hours or more,
fresh gas being added if required. The cover is then
removed and water poured on the ore and drawn off at the
bottom until gold can no longer be detected in the wash-
water. The gold chloride is precipitated by ferrous

sulphate, the equation being

2AuCl 8 + 6FeS0 4 = 2Au + Fe 2 Cl G + 2Fe 2 (S0 4 ) 8 .

The brown precipitate of metallic gold is allowed to settle,

126 THE PRECIOUS METALS

and is collected by filtering. It is then washed, dried and

melted.
The barrel process of chlorination was originally intro-
duced in the United States, and has never been used for
any great length of time in any other country. The
chlorine is applied under pressure with a view to shorten
the time of action. The gas was originally pumped into the
barrel through one of trunnions, but the later practice
its

is to generate the chlorine inside the barrel by means of

sulphuric acid and bleaching powder. Care is taken to

prevent these chemicals from coming into contact with each
other until the manhole of the barrel has been closed and
rotation begun. More water is used than in the Plattner
process, the charge being required to be an easily flowing
pulp. Iron barrels lined with lead and holding a charge of
from five to ten tons of ore are used, and are rotated for
about five hours, when the dissolution of the gold is usually

complete. The rotation is of value in mixing the charge

and bringing all parts ofequally under the action of the

it

chlorine, and also in scouring and brightening the particles

of gold and laying them open to attack.

When emptied into

dissolution is complete, the charge is

a false-bottomed vat to be washed, or the washing is some-

times carried out in the barrel itself, which in this case is
provided with a filter-bed forming the chord of an arc of the
circle of the barrel. The precipitation of the gold is effected
by sulphuretted hydrogen, and the sulphide of gold is
filter-pressed, roasted on iron trays and melted with fluxes.
If muchbase metals are present in the solution, the gold is

precipitated by ferrous sulphate.

The vat solution process is in use at the Mount Morgan
 TREATMENT OF GOLD ORES BY WET METHODS 127

Mine, Queensland, the largest chlorination mill in the

world. The vats are oblong, and of a capacity of 100 tons
of ore each. They are constructed of concrete lined with

cement and faced with pitch and tar, which are unattacked
by chlorine. The solution is run on to the ore through
earthenware pipes, and is drawn off through the filter-bed

as soon as the vat is full. Fresh chlorine solution is added

continuously until the gold is completely dissolved, which
appears to take about thirty-six hours. Water for washing
is then applied and passed through the ore until ferrous
sulphate gives no precipitate of gold in the issuing liquid.
The tailings are discharged from the tanks into trucks.
The gold solution is heated to expel all free chlorine, and
the liquid while still hot is allowed to run through tanks
with coarsely powdered charcoal resting on a filter-bed
filled

and covered with perforated sheet-lead to keep the charcoal

from floating. The gold is precipitated on the charcoal,
and when the latter is sufficiently coated it is burnt in small
furnaces, and the ashes melted with fluxes in a reverberatory
furnace.
 CHAPTEE IX

SILVER AND ITS ALLOYS

1
Historical. Silver was known to the ancients, probably
almost as soon as gold itself. Although it did not attract
so much attention as gold, the name which it carried in
ancient languages was founded on its colour and appear-
ance. Thus in Hebrew " khesef " (silver) was derived
from the verb " khasaf," to be pale, and the Latin argentum
was derived from its Greek equivalent, which meant white.'2
Silver was in use as money probably as early as gold.
It isrecorded that Abraham paid Ephron in silver for the
land bought as a burial place. 3 Pliny states that mines of
silver occurred in the mountains in almost all the Eoman
4
provinces, and that Spain contained the richest deposits.
Strabo 5 describes the treatment of silver ore at New
Carthage in Spain as consisting of roasting, followed by
washing through sieves. This operation was repeated five
times. The residue was fused with lead and cupelled, with
the result that pure silver was obtained.

According to Hoefer the only compound of silver known

to the ancients was silver chloride, which was formed in

refining gold and thrown away with the refuse.

1
Some notes on the history of silver are given in Chapter I.
2 " Histoire de la
Hoefer, Chimie," p. 43.
8
Genesis, chap, xxiii., v. 15.
4 " Hist.
Nat.," Lib. XXXIII., cap. 31.
s "
Geogr.," Lib. III.
 SILVEE AND ITS ALLOYS 129

THE PROPERTIES OF SILVER.

Silver is pure white in colour and is capable of receiving

a brilliant polish. When in the form of fine powder, as for
instance when reduced fromthe state of chloride by iron
and acid, silver is
grey and earthy in appearance. It is
extremely malleable and ductile, though less so than gold.
It is hardened by hammering or rolling, but its softness is

restored by annealing at a red heat. It is harder than gold

and softer than copper. the most perfect conductor of

It is

heat and electricity known, and its conductivity is increased

by annealing. Its specific gravity is 10*50 after having

been cast, and 10*57 after being struck in a coining press.

Pure however, is difficult to cast quite solid, and

silver,

generally contains minute cavities which reduce its density.

The density of English silver coin (925 fine) is 10'37. The
specific gravity of molten silver is 9*51. The atomic
weight of silver is 107*93 (taking = 16), and its symbol
is Ag.
Silver melts at 962 (Berthelot), and volatilises when
heated by the oxy-hydrogen blow-pipe, yielding a pale blue
vapour. It can be distilled in quantity by a hydrogen
flame in a suitable furnace, or by being heated in the
electric arc furnace, when it boils vigorously. Molten silver
absorbs about twenty-two volumes of oxygen from the air,
and the gas is given off again during solidification, usually
with much vigour. The solidified crust of silver is gene-

rally broken and blistered, and particles of metal are

projected by the gas disengaged from the interior (the
" "
so-called spitting or "sprouting"). This effect can be
prevented by stirring the metal under a cover of charcoal,
P.M. K
130 THE PRECIOUS METALS

or by adding a small quantity of copper, zinc or other

oxidisable metal, by which the oxygen is retained. It is

probable that the oxygen exists in the molten silver in

the form of the oxide, Ag 2 0, as this substance has been
shown to be stable at very high temperatures, although it

is instantly decomposed at a red heat.

Silvernot oxidised at ordinary temperatures either in
is

dry or moist air, and is only slowly oxidised at high

temperatures. It is rapidly dissolved by boiling nitric

acid, and if it is slightly impure it is attacked at the

ordinary temperature. It is soluble in boiling, concen-

trated sulphuric acid, but is not attacked by the dilute acid

unless alloyed with such a metal as iron. Hydrochloric
acid acts very slowly, if at all, on pure silver, but at a red
heat silver chloride is readily formed by the action of HC1
gas. The formation of silver chloride by the action of aqua
regia, hydrochloric acid on silver is soon
chlorine, or

stopped by the formation of a protective layer of AgCl,

which is insoluble in acid.

Sulphide of silver is readily formed by throwing

sulphur on to molten silver, or by exposing silver to the

action of sulphurous fumes at the ordinary temperature.

It is for this reason that silver blackens in the air in towns,
a film of sulphide of silver being formed on the surface,
H
presumably by the action of 2 S. The blackening of silver
egg-spoons is also due to the formation of sulphide of
silver.

PREPARATION OF PURE SILVER.

The simplest method of preparing pure silver is to dis-

solve commercial silver in nitric acid, allow the liquid to

 SILVER AXD ITS ALLOYS 131

stand, syphon off from precipitated gold, etc., precipitate

silver chloride in a hot solution with hydrochloric acid,
stir well, wash by decantation. The
allow to settle and
silver chloride becomes granular by continuous stirring and

settles readily. When the decanted liquid no longer shows

a trace of impurity, the silver chloride is transferred to a

porous pot which has been previously steeped in hydro-

chloric acid for some days and thoroughly washed by long,
continued standing in clean distilled water frequently
renewed. The porous pot is placed in water in an outer
vessel, in which carbon electrodes are dipped. A cathode
ofplatinum or pure silver is plunged into the silver chloride
and connected with the carbon anodes through an electric
battery. The white chloride at once begins to blacken next
the cathode, and is entirely converted into silver in the
course of a few days. The outer liquid becomes strongly
impregnated with chlorine gas and should be renewed once
or twice a day.
The carbon anodes may be replaced by a cylinder of

wrought iron, and a little hydrochloric acid added to start

the action. No chlorine is then evolved, and the action is

far more rapid. In this case the chloride of iron diffuses

into the inner pot and contaminates the silver, but may be
removed by washing.
The grey silver powder, after being carefully washed by
shaking with pure water renewed many times, is dried
it

and melted under a little carbon in a graphite or clay

crucible and poured into a blackleaded iron mould.

This method is used in the preparation of pure silver for

trial plates at the Royal Mint.

K 2
132 THE PEECIOUS METALS

ALLOYS OF SILVER.

The Silver-copper alloys are by far the most important of

these bodies commercially. The addition of copper to silver

"
lowers melting point, prevents it from
its sprouting" on
solidification, and makes it harder without sensibly impairing
its malleability or altering its colour. The melting points

I.IOO r

Coppet

1000

.Si/vet

\
800

10 20 30 40 SO 60 70 80 90 100

Percentage of Copper in Atoms.

FIG. 25. Freezing Points of the Alloys of Silver and Copper,

of the series of alloys as determined by Koberts- Austen 1

and by Heycock and Neville 2 are shown in Fig. 26. Level's

alloy, containing 71 '89 per cent, of silver or 60 atoms of

silver per cent., is the eutectic alloy, melting at 778, the

lowest temperature of any of the series. This alloy is per-

fectly homogeneous, and has been used for coinage purposes
for circulation in the Dutch Indies. It is not, however,
1
Jour. Soc. Arts, March, 1897.
2
Phil. Trans. Boij. Soc., Vol. CLXXXIX. (1897), p. 25.
 SILVER AXD ITS ALLOYS 133

especially well fitted for manufacture into coin, as it is

harder and less ductile than the other silver-copper alloys.

Alloys containing more silver than is contained by the
eutectic alloy consist of non-homogeneous mixtures, and on
solidification become richer in silver in the interior than on
the outside. The difference, however, never exceeds a few
parts per 1,000. These alloys are largely used for coinage
and for wares of In England the standard for
all kinds.
both is 92*5 per cent, silver, with 7*5 per cent, copper.
"
This is known as standard silver." In the United States,
France, Germany and most
other countries the principal

coinage alloy contains 90 per cent, silver, and melts at a

lower temperature than the English standard alloy and is

slightly harder. The colour of these alloys is almost the

same. English silver coin melts at 890 and American
coin at 873. For full details as to the standards of alloys
used for coinage in various countries, see Chapter XV., on
Minting. The silver-copper alloys blacken on being heated
in the air owing to the formation of oxide of copper, which
can be removed by hot dilute sulphuric acid. They dissolve
completely in nitric acid and in boiling concentrated sul-

phuric acid, but are scarcely affected by hydrochloric acid.

The Silver-zinc alloys rich in silver present some resem-
blances to the corresponding silver-copper alloys. They are
malleable and ductile, and take a high polish. The freezing-
1
point curve of the series according to Heycock and Neville
2
and Petrenko is shown in Fig. 26. The addition of zinc lowers
the freezing point of silver, and this continues throughout
the series, pure zinc having a lower melting point than any

1
Jour. Chem. Soc., Vol. LXXI. (1897), p. 407.
2
Chem., Vol.
Zeit. Ar.orcj. XL
VIII. (1906), p. 347.
134 THE PKECIOUS METALS

of the alloys. The alloys with less than 20 per cent, of

zinc are malleable, bat those with 20 to 30 per cent, are
brittle. These alloys contain the first eutectic, which melts

- 950'
350

- 900'

- 850'

- 800'

20 30 40 50 60 70 80 90 100

Percentage Weight oF Zinc .

FIG. 26. Freezing Points of the Alloys of Silver and Zinc.

at 705 and contains 28 per cent, of zinc. The alloy corre-

sponding to the formula AgZn, and containing 37*7 per
cent, of zinc, shows a lilac-red tint when heated to 300 and
suddenly cooled, resembling in its colour the alloys AuZn
and AgCd. This colour disappears if the alloy is heated
 SILVER AND ITS ALLOYS 135

with dilute nitric acid, and hence probably due to a is

superficial layer of oxide. alloys containing about 48

The
to 50 per cent, of zinc have a conchoidal fracture with a
brilliant glassy surface, and are very brittle. These alloys
contain the compound Ag 2 Zn 3 . All the alloys, except those
whose composition is nearly equivalent to AgZn, are of a
silver-white colour. The best method of preparing them,
as in the case of the silver-cadmium alloys, is to pour
molten silver into molten zinc. The latter is kept at a low
red heat, considerably below its boiling point. When this
method of preparation of the alloy is adopted the loss of zinc
by volatilisation or oxidation is very small. Besides AgZn
and Ag 2 Zn 3 the compounds Ag 2 Zn 3 and Ag2 Zn 5 and one of
,

about the composition AgZni 2 have been shown to exist.

Silver and Cadmium. These alloys are softer, whiter and
more malleable and ductile than the silver-copper alloys.
They isomorphous mixtures of various compounds
consist of
and cadmium, and the alloys rich in silver are
of silver

homogeneous and suitable for the manufacture of trial

plates for assay purposes. The compounds AgiCd, Ag2 Cd,
1
Ag3Cd 2 AgCd, Ag
, 2 Cd 3 and AgCd 3 have been shown to exist.
The alloy containing 40 per cent, of silver (Ag 2 Cd 3 ) has a
brilliant lustre and a conchoidal
fracture, and is intensely
hard and The compounds Ag2 Cd 3 and Ag 2 Zn 3 thus
brittle.

bear strong resemblances to each other, as do the compounds

AgZn and AgCd. A violet colour distinguishes the alloys

with composition nearly equivalent to AgCd, and these
alloys, like those near in composition to AgZn, undergo
allotropic changes at temperatures considerably below their
melting points.
1
T. K Hose, Proc. Roy. S c., Vol. LXXIV. (1904), p. 218.
136 THE PRECIOUS METALS

Silver and Lead. Silver and lead mix together in all pro-

portions while they are molten, but on solidification they

separate, so that the alloys are not homogeneous. The
freezing-point curve is shown in Fig. 27, which is after
1
Petrenko, and is remarkable for the persistence of the

1000

900

800

700

600

500

400
C Pb
300

200

100

" 10 20 30 40 SO 60 70 80 30 100

Percentage Weight of Lead

FIG. 27. Freezing Points of the Alloys of Silver and Lead.

eutectic point, which is observable in an alloy containing

over 99 per cent, of silver. The two metals are in fact
quite insoluble in each other in the solid state. The eutectic
alloy, B, contains 2*5 per cent, of silver and solidifies at

303. There is no evidence that any compounds of silver

and lead exist. These properties explain the effects of
the Pattinson process (q.v.) for concentrating the silver

Anonj. Chem., Vol. LI1I. (1907),

1
Zeit. p. 200.
 SILVEE AND ITS ALLOYS 137

contained in lead. The curve AB is of a remarkable

shape.
Molten lead takes up silver readily, removing it from its
compounds such as the chloride, sulphide, arsenide, etc.
Lead also silver from its alloys with copper at
removes
temperatures between the melting point of lead and those
of the silver-copper alloys. This property was used in the
old liquation process, now abandoned. On the other hand,
zinc removes silver from molten lead, forming a zinc- silver

compound, probably AgZni2, which is lighter than lead

and floats to the top with the dross and surplus zinc, so
that it can be skimmed off the lead. This is the basis of
Parke's process (q.r.).

When silver-lead alloys are melted in the air, the lead is

oxidised and the silver remains unaltered. At a red heat

the action is rapid and the litharge produced becomes
molten and is far less viscous than the molten metals.
The cupellation process is based on these properties.
In 1853 was shown by Levol 1 that alloys of silver and
it

lead undergo segregation and are not uniform in composi-

tion. A result of this is that the valuation of argentiferous
lead produced in smelting operations is a matter of difficulty.
2
According to A. C. Claudet, satisfactory samples cannot be
cut from pig lead, and it is necessary to remelt the alloy
and to take dip samples.

Silver and Mercury. The alloys are called amalgams like

the other alloys of mercury. Surface action between silver
and mercury is easy and rapid at the ordinary temperature,

1
Ann.de Chim. et Phys., Vol. XXXIX. (1853), p. 173; Percy's
" Silver and
Gold," p. 173.
2
Tram. Just, of Mug. and Met., Vol. VI. (1898), p. 29.
138 THE PEECIOUS METALS
"
so that silver is at once " wetted by mercury. The inter-

penetration of the metals is, however, exceedingly slow,

though more rapid at the boiling point of mercury than at
the ordinary temperature. The crystalline amalgam AgHg
(containing 35 per cent, of silver and 65 per cent, of mer-
cury) produced in the branching form known as Arbor
is

Diance, by prolonged contact of an amalgam of silver and mer-

cury with a saturated solution of silver and mercury nitrates
"
in nitric acid (see Percy'sSilverand Gold," p. 176). If too
much mercury is present a liquid amalgam only is formed.
The amount of silver dissolved by liquid mercury, and
remaining in it after the solid amalgam has been filtered
off, is stated to vary with the pressure to which the amalgam
is subjected. This seems to be improbable. The amount
of dissolved silver is variously stated at from 0*05 to T5
per cent. The composition of solid amalgam undoubtedly
varies with the pressure during filtration and with the size

of the particles of silver, which are rarely fully amalgamated

to their centre. Before filtering, the solid amalgam floats

to the top of the mercury.

Mercury decomposes a number of compounds of silver,

such as the chloride and bromide, forming mercury chloride,
bromide, and amalgams of silver.
etc.,

Silver-tin Alloys. These alloys are silver- white in colour

and are in general somewhat brittle and hard. The alloys
containing between 40 and 60 per cent, of silver are used
1
by dentists. They are mixed with about an equal weight
of mercury, and amalgams are formed which are pasty at
first and subsequently become hard. The
degree of full

hardness is not attained until 24 or 48 hours after the

1
E. A. Smith, " Silver Alloys of Industrial Importance," 1901.
 SILVER AND ITS ALLOYS 139

mixing with mercury, and some molecular change evidently

takes place gradually.
The thermic study of these alloys by Petrenko 1 shows
that only one compound, Ag 3 Sn, exists, containing 26'7

per cent, of tin, and that there is only one homogeneous

1000

300 - - 900'

800' - 800'

700'

- 600"

500

400 - 400

300'

200'

C
100

0"

Percentage Weight oF Tin .

FIG. 28. Freezing Points of the Alloys of Silver and Tin.

alloy, the eutectic mixture, which contains 3*5 per cent, of

tin and 220,
solidifies at or 8 below the melting point
of tin. The thermic diagram of these alloys is shown in
Fig. 28. The compound Ag 3 Sn undergoes a polymorphic
change at 232. Fig. 29 shows the structure of the 20 per
1
Zeit. Anonj. Chem., Vol. LIII. (1S07), p. 204.
140 THE PRECIOUS METALS

cent, tin alloy, by hydrogen peroxide and

after attack

ammonia, magnified twenty-four diameters. Both these

diagrams are after Petrenko.
Silver and Platinum. These alloys are also used in

dentistry. The first additions of platinum to silver raise the

melting point of the latter, and the dental alloys which
contain from 25 to 35 per cent, of platinum require a high

temperature for their preparation. They are hard but ductile,

and do not readily tarnish.
The alloy containing 33*3 per
cent, of platinum has been
used as a standard of electrical
resistance. The platinum-
silver alloys are also used in
jewellery. As the proportion
of platinum is increased the
alloys formed are less ductile
and harder. The metals do
FIG. 29. Silver, 80 per cent. not readily mix, the platinum
Tin, 20 per cent, x 24.
having a strong tendency to
settle to the bottom in molten mixtures on account of its

high density. Platinum in its alloys with silver is soluble

to some extent in nitric acid, so that it can be partly
separated from gold by the nitric acid parting process
(see below under "Assay of Platinum," Chapter XYIL).
When these alloys are attacked by boiling concentrated
sulphuric acid, the silver dissolves and the platinum
remains undissolved.
The freezing-point curves of the alloys have been
examined by Doerinckel, 1 who melted them in a carbon-
1
Zeit. Anorg. Chem., Vol. LIY. (1907), p. 338.
 SILVER AND ITS ALLOYS 141

resistance electric furnace, and measured temperatures by

a thermocouple of platinum and platinum-rhodium. The
alloys containing more than 80 per cent, platinum were

1700**

1600

1500

14-00

/- /300

1100

tOOO

900 C

f>
fl

10 20 30 4-0 50 60 70 80 30 100
Percentage weight of Platinum.
FIG. 30. Free/ing Points of Alloys of Silver and
Platinum.

not examined. The diagram of freezing points is shown

in Fig. 30. The curve ABC gives the initial freezing
points of the alloys (liquidus curve), and AED6 the points
where solidification is completed. It is uncertain whether

pure platinum or mixed crystals rich in platinum separate

142 THE PEECIOUS METALS

along the line BC. It is clear that the alloys when solidi-

fied from fusion are not homogeneous, but separate into

two constituents. When alloys containing between 30 and
45 per cent, of platinum are reheated at 1,150 to 1,180 for
six hours, they tend to become homogeneous.

Silver and Aluminium. The preparation of the alloys

rich in silver pre-
sents some difficul-

ties owing to the

low density of
aluminium and the
tendency it has to

cover itself with a

sldn of oxide which
is not easily broken

through. The best

method is to melt
aluminium and to

pour molten silver

into it. The alloys

FIG. 31. Ag, 94 per cent. Al, 6 per
containing from
;

cent, x 30.
to 4 per cent, of

aluminium and also those containing from 7*72 to 11*3 per

cent, easily receive a fine polish, because they contain only

one structural element. 1 Most of the alloys are not homo-
geneous (see Fig. 31, which is the alloy containing 6 per
cent, aluminium magnified thirty diameters). There is a
eutectic melting at 567 and containing 30 per cent, of

aluminium, and two compounds have been found to exist,

AlAg 2 and AlAg 3 .

1
Petrenko, Zeit. Ancrg. Chem., Yol. XLYI. (1905), p. 49.
 SILVER AND ITS ALLOYS 143

Silver and Bismuth form a series of alloys similar in

properties to the silver-lead alloys. There is a eutectic

melting 262 and containing 2*5 per cent, of silver. The

at

eutectic persists almost to the pure silver end of the series,

making all the alloys brittle. By annealing for two or three

hours at 260 to 270, however, the bismuth is dissolved in
the silver than 5 per cent, of bismuth is present, 1
if less

and under these circumstances these alloys become ductile.

When copper is added to the alloys, however, bismuth is no
longer soluble in the silver, and annealing is not in that
case sufficient to cure the brittleness if more than a trace
of bismuth is present. Gowland found some silver ingots
(received at the Japanese Mint from certain Japanese
2
refineries) to be brittle owing to the presence of bismuth.
Silver and Nickel form homogeneous alloys if the silver
does not exceed 4 per cent. These alloys fuse at tempera-
tures above 1,464. If more silver is present, the alloys

separate into two layers, the nickel with 4 per cent, of

silver floating on the top of nearly pure silver. Silver
does not mix at all with iron or cobalt.
Silver and Palladium mix in all proportions,
forming a
series resembling the gold-palladium alloys.
Silver and Thallium*
Theseries of alloys closely resemble
those of silver and lead and of silver and bismuth. The
eutectic alloy contains 2*5 per cent, of silver and solidifies at

287. As much as 10 per cent, of thallium, however, dis-

solves in silver, forming homogeneous ductile alloys. When
alloys rich in silver are melted, thallium is rapidly volatilised.
1
Petrenko, Zeit. A ></</. <'/>., Vol. L. (1906), p. 136.
~
2
Jour. Vol. LI. (1887), p. 410.
('/tern. Hoc.,
Petrenko, Zeit. A nor;/. C/iem., Vol. L. (1906),
*
p. 133; Poitevin,
Revue de Mcfalfun/ie, 1908, p. 159.
 CHAPTER X

COMPOUNDS AND ORES OF SILVER

Silver oxide, Ag2 0, is formed by heating finely divided

silver with cupric oxide, red lead, or manganese dioxide,
and in presence of one of these oxides is not decomposed
even at a bright red heat. When heated alone it is

decomposed and oxygen at temperatures between

into silver
250 and 300, but Troost and Hautefeuille showed that at
about 1,400silver unites with oxygen with the formation

ofAg 2At 0. 1
a red heat Ag 2 is immediately decomposed,
and cannot be formed in the absence of other metallic
oxides. Wait showed 2 that in a rich litharge containing
2*94 per cent, of silver, about 19 per cent, of the silver was
soluble in acetic acid, and must be assumed to have been
in the form of oxide, as metallic silver is not attacked by
acetic acid. Silver oxide may be prepared by adding baryta
water to a solution of silver nitrate, and washing and dry-
ing the precipitate. Thus prepared, it is a powder of a deep
brownish-olive colour, and has a density of 7'1.
Silver sulphide, Ag 2 S, is prepared in the dry way by

melting together silver and sulphur, when a dark grey

crystalline substance results, resembling the native sulphide,
argentite, in its properties. It has a feebly metallic lustre,

Comptes Rendus, Vol. LXXXIV. (1877), p. 948.

1

2
Trans. Am, InsL Mr.g. Eng., Vol. XV. p. 423.
 OF THE
UNIVERSITY
OF

COMPOUNDS AND ORES OF SILVER 145

and is opaque, easily sectile and slightly malleable. Its

density is 6'85.

It is also formed by the action of aqueous solutions of

alkaline sulphides on metallic silver, although far less
rapidly than in the dry way at high temperatures. Silver
is rapidly tarnished with the formation of Ag 2 S by the
action of sulphuretted hydrogen gas or by being rubbed
with sulphur. Silver sulphide is precipitated from silver
solutions by H 2 S or by soluble sulphides.
It is soluble in
cyanide of potassium, especially in hot
strong solutions. This property is useful in removing the
tarnish from silver articles when polishing is impossible or
undesirable. Sulphide of silver is acted on by nitric acid
with the formation of nitrate of silver and the separation of
sulphur. "Oxidised" silver is silver coated with Ag 2 S.
" "
It isprepared by dipping the articles to be oxidised into
a hot solution of potassium sulphide.
Friedrich and Leroux have shown 1 that silver sulphide
solidifies at 806, and that all mixtures of silver and sulphide
of silver containing more than about 13 per cent, of Ag2 S
solidify at 906, any excess of Ag2 S over 13 per cent.
solidifying at 806. If more than 13 per cent. Ag2 S is

present, the two substances are not miscible even in the

liquid state, but separate into layers.
Sulphide of silver readily mixes with other metallic
sulphides when they are melted together, so that in smelt-
ing, silver is readily concentrated in a matte. When fused
by itself, sulphide of silver loses a little sulphur, and the
remainder of the sulphur may be removed by fusion with
metallic iron, when sulphide of iron and metallic silver
1
Mdalhirgie, Vol. III. (1906), p. 361.
P.M. L
146 THE PEECIOUS METALS

results. Copper reduces silver from its fused sulphide only

imperfectly, but the sulphide is reduced to metallic silver
when heated in air, steam or hydrogen. "When copper
sulphide is mixed with silver sulphide and the mixture is

heated in air, sulphate of copper is produced, and at a red

heat gaseous S0 3 is
given off and silver sulphate and oxide
of copper remain. If the temperature is too
high, or if the
action is prolonged for some time, the silver sulphate is

split up and metallic silver remains.

Silver sulphide is converted into chloride by being roasted
with common salt in the presence of air, or by the aciion of
chlorine water, chlorine gas or hydrochloric acid, especially
with the aid of heat. Even in cold solutions cuprous
chloride and cupric chloride act similarly on silver sulphide,
the reactions being partly represented by the following

equations :

Ag 2 S + CuCl 2 = 2AgCl + CuS

Ag2 S + Cu 2 Cl 2 = 2AgCl + CuS + Cu.
The action is facilitated by boiling. Some metallic silver
is also produced, according to the equation
Ag2 S + CuaCla = 2Ag + CuS + CuCl 2 .

Sodium copper thiosulphate

-
(4Na 2 S 2 3 .3Cu 2 S2 3) was
shown by Eussell to decompose silver sulphide, the silver

going into solution as thiosulphate, with the formation

of copper sulphide. Ordinary sodium thiosulphate, how-
ever, does not act on silver sulphide. Alkaline cyanides
dissolve Ag 2 S, and it is decomposed by mercury, the action
being rapid in the presence of metallic iron, especially on
heating.
These properties of silver sulphide are of importance in
the extraction of silver from its ores.
 COMPOUNDS AND ORES OF SILVER 147

Silver chloride, AgCl. This compound is usually pre-

pared by adding hydrochloric acid or a soluble chloride
such as common salt to a solution of silver nitrate. It is
thrown down as a white, curdy, amorphous precipitate which
can be made to settle by shaking, or less readily by stirring

and heating. It is readily washed by decantation.

Silver chloride is also formed by the action of chlorine
gas on metallic silver and on certain compounds of silver.
This action takes place at ordinary temperatures, but is
expedited by heat, and at a red heat most of the compounds
of silver are converted into the chloride.

Silver chloride precipitate is a white anhydrous powder.

Its density is 5'5. It fuses at 360 to a reddish-yellow

liquid, which is readily absorbed by unglazed clay crucibles.

It is volatilised without decomposition at a bright red heat.
When solidified after fusion it is waxy in appearance, and
is translucent, sectile Thin plates of it are
and tough.
flexible and almost transparent, and have a horn-like

appearance. In this state it is sometimes called horn-silver.

The mineral Kerargyrite, which consists of AgCl, resembles
horn-silver in appearance.
is darkened by exposure to light owing
Silver chloride
to the formation of oxychloride of silver, Ag2 C10. The
darkening is checked by the presence of an excess of hydro-
chloric acid, and increased by an excess of silver nitrate in

solution. It is entirely prevented by the presence of

chloride of mercury.
Silver chloride is very insoluble in water. Silver can be

readily detected in watercontaining one-millionth of its

weight of silver in solution as nitrate, by the addition of
hydrochloric acid. Stas detected far smaller quantities of
L 2
148 THE PBECIOUS METALS

silver in solution than this. Silver chloride is, however,

perceptibly soluble in strong nitric acid, especially when
the acid is hot, and is more readily soluble in hydrochloric
acid and in a number of chlorides, as shown in the following
1
table :

Solution.
 COMPOUNDS AND OEES OF SILVEE 149

alkaline carbonates at a red heat. The reaction may be

represented as follows :

4AgCl + 2Na 2 C0 3 = 4Ag + 4NaCl + 2C0 2 + 2.

It is reduced by cuprous chloride in ammoniacal

also
solution or by being heated with charcoal.
Silver bromide, AgBr, is a pale yellow powder of density
6'25, melting at 434 to a red liquid. It can be prepared
by precipitating silver nitrate with a soluble bromide. Its
properties resemble those of silver chloride, but it is less
soluble in water and other solvents. Sodium bromide has
long been used instead of sodium chloride at the Utrecht

Mint the precipitation of silver in the volumetric

for

process of assay. Silver bromide is dissolved by potassium

bromide. It darkens rapidly when exposed to light.
Silver iodide, Agl, is a pale yellow powder, melting at
527 to a red liquid. On solidification a substance is
produced which is yellow, transparent and flexible above
142, and pale green, opaque, brittle and crystalline below
142. Silver iodide is almost insoluble in water and in
alkaline chlorides, only slightly soluble in ammonia
and is

(one part in 2,510 parts), but it is soluble in cyanide and

thiosulphate solutions, and in a concentrated solution of
potassium iodide.
Silver cyanide, AgCy, is obtained as a white curdy pre-
cipitate by the addition of a soluble cyanide to a solution
of silver nitrate. In its properties it resembles the haloid
silver salts, but it is not darkened by light. It is soluble in

molecular proportions in potassium cyanide, thus

AgCy + KCy = KAgCy2 .

It is decomposed, like the haloid salts, by strong sulphuric

150 THE PKECTOUS METALS

acid. It is also
decomposed by hydrochloric acid and by
nitric acid with evolution of HCy. When heated it is
decomposed, cyanogen being given off and metallic silver

remaining as a residue. Its density is 3*95.

Silver -potassium cyanide, KAgCy2
crystalline, forming ,
is

anhydrous transparent regular octahedra, soluble in 4*7

parts of water at 20 and in 25 parts of alcohol. It is

formed in solution by dissolving AgCy or AgCl in KCy,

and also by the action of cyanide solutions on metallic
silver in the presence of air or certain other oxidising

agents. The reaction may be represented as follows :

2Ag + 4KCy + +H 2 = 2KAgCy 2 + 2KOH.

Silver potassium cyanide is decomposed by acids, AgCy

being precipitated. The silver is precipitated from a
solution KAgCy2 by a current of electricity and by
of

certain metals, such as zinc, copper, etc., of which zinc is

used on the large scale in the cyanide process. The
solution of this salt is used for electro-plating.
Silver nitrate, occurs in colourless transparent
AgN0 3 ,

crystalline plates, unalterable at ordinary temperatures in

the air and having a density of 4*35. It melts at 198, and

the fused salt after solidification is the form of silver nitrate

commonly known as lunar caustic which used by surgeons.

is

At a low red heat silver nitrate is reduced to the metallic

state. It dissolves in 0*783 of its weight of water at 11,
and in half its weight of water at the boiling point.

Silver telluride, Ag2 Te, and Silver selenide, Ag2 Se, re-

semble the sulphide in their properties, and occur native.

Like the sulphide, when roasted in air they offer consider-
able resistance to oxidation.
 COMPOUNDS AND OEES OF SILVER 151

Sili't'r xiiljtJiate, Ag2 S04, is formed at a red heat by the

action of sulphates on silver sulphide. Copper sulphate is

particularly suitable for effecting this conversion. It is also

formed when silver is dissolved in sulphuric acid, and when

it is heated with acid sulphates. Silver sulphate is freely
soluble in hot concentrated sulphuric acid, but is far less
soluble in cold acid and in water. Copper, iron or zinc
precipitates silver from solutions of the sulphate.

Silver sulphate is decomposed at a bright red heat into

silver, oxygen and sulphur dioxide. It is converted into

silver chloride by being roasted with common salt at a red

heat.
Silver antimonate and Silver ar senate are formed in the
roasting of antimonial and arsenical silver ores. They are
stable at high temperatures, but can be reduced by heating
with charcoal or metallic lead. They are also convertible
into silver chloride by roasting with salt, but are not
reduced by mercury at the ordinary temperatures.

SILVER ORES.

Silver occurs native, but unlike gold it is generally

combined with other elements in its ores. Native silver,
together with horn- silver, AgCl, bromite, AgBr, etc., are
found in the oxidised portions of lodes near the surface. At
deeper levels these are replaced by argentiferous sulphides,
arsenides and antimonides.
Native Silver, Ag, occurs in cubes and octahedra, usually
in aggregates which take a filiform or dendritic shape. It

also occurs massive and in laminae. It has been found

in considerable masses at Kongsberg, in Norway, where it
 152 THE PRECIOUS METALS

is the principal ore. The largest mass recorded by

Daubreeweighed kilograms, 697 Troy. or 22,409 oz.

Large amounts of native silver were also found in the

great Comstock lode, and great quantities have been

extracted in various localities in Mexico and South America,
at Freiberg and at Broken Hill. Native silver is some-
times over 990 fine, but it generally contains such
impurities as gold, .copper, etc. The alloy of silver and
mercury (silver amalgam) occurs native.
Argentite, Ag 2 S, is one of the most important ores of
silver. It occurs in black, soft, sectile crystals and masses,
which are usually pure, containing 87'1 per cent, of silver.
Its density is 7*2 or 7'3. occurs in great quantities at
It

the Comstock lode, Nevada, in Mexico, and in many other

localities.

Other widely distributed sulphides are the ruby silver

ores, proustite, 3Ag 2 S +
As 2 S3, pyrargyrite, 3Ag 2 S Sb 2 S 3 + ,

and brittle silver ore, or stephanite, 5Ag 2 S + Sb 2 S3 .

Fahlore, or grey silver, another important ore, is a mix-

ture of antimony and arsenic sulphides with sulphides
of silver, mercury, copper, zinc and iron. Its formula is
4(Ag 2 Hg.CuFeZn)S + (SbAs) 2 S3 . It may contain any pro-

portion of silver up to 31 per cent., arsenical grey silver

being poorer than the antimonial varieties. It occurs in
Germany, the Tyrol, Bolivia, Nevada, Mexico, Colorado, etc.
Polybasite, 9(Ag 2 Cu)S + (SbAs) 2 S 3 is also a commonly ,

occurring mineral, especially in America.

sulphide minerals containing silver, but less
Other
important as ores than those given above, are castiUite,
(CuAg) 2 S.2(CuPbZnFe)S; cosalite, 2(Ag 2 Pb)S.Bi 2 S 3 dys- ;

crasite, Ag 3 Sb ; freieslebenite, 5(Ag 2 Pb)S.2Sb 2 S 3 and stro- ,

 COMPOUNDS AND OEES OF SILVER 153

mey erite, Ag2 S.Cu 2 S. Hessite, Ag 2 Te, and petzite, a tellu-

ride of gold and silver, are also of some importance as
ores.

The haloid ores, Kcraryyrite , AgCl, and embolite, Ag(ClBr)

are also very abundant in the upper oxidised zones of
lodes in South America, Mexico, etc. Many of these have
now been worked out. Bromite, AgBr, and iodite, Agl, are
not very abundant.
There are many other minerals
containing silver
described by mineralogists, but they are not important to
miners or metallurgists. Among these are Naumannite,
Ag2 Se, and other selenides, chilenite, Ag6 Bi, and a number
of complex sulphides. Several silver minerals often occur
together in the same mine.
Silver also occurs in the ores of lead, copper, zinc,
nickel, etc. Galena often contains considerable amounts
of silver (up to 5 per cent.), and zinc blende, ZnS, chalco-
pyrite, CuS.FeS.FeS 2 , bournonite and other minerals also
yield much silver. Sometimes the silver replaces one of

the base metals in these minerals, but it is often not an

essential part of the mineral, occurring in the form of thin

films which separate the crystal faces or are situated along

the cleavage planes.
 CHAPTEE XI

THE EXTRACTION OF SILVER FROM ITS ORES

SILVER is extracted from its ores by methods which may

be classified as follows :

1. Amalgamation processes.
2. Lixiviation
3. Smelting
4. Electrolytic
These are described in succession below.

1. AMALGAMATION PROCESSES.
Native silver, together with chloride, bromide, chloro-
bromide and iodide of silver can be amalgamated direct by
grinding with mercury without any special preparation.
The haloid salts are decomposed with the formation of
amalgam and haloid compounds of mercury.
silver

Owing to its lower density and its habit of being in

combination, however, silver cannot be extracted from its

ores by amalgamation so easily and completely as gold.

When crushed in stamp batteries and passed over amalga-
mated tables silver ores yield only small percentages of their
silver contents. More prolonged and intimate contact with
the mercury is required, and this is obtained in various ways.
Silver sulphide is decomposed by mercury, metallic
silver being set free and amalgamated and sulphide of
mercury being produced, but the action is very slow.
 THE EXTEACTION OP SILVER FROM ITS ORES 155

Arsenical and antimonial silver compounds are still less

easy to amalgamate, and the action is always incomplete.

Ores containing sulphides, arsenides and antimonides are
accordingly treated by roasting with salt, or by the use of
various chemicals, in order to convert the refractory silver

compounds into silver chloride or metallic silver, before

they are subjected to amalgamation.
Direct amalgamation has been used on rich ores containing
native silver and silver chloride and only small quantities
of sulphides. In the tina process, formerly employed at
Kongsberg, Norway, and in Chile and Peru, the ore was
crushed in edge-runner or Chilian mills and then amalga-
mated in tinas or wooden tubs with iron bottoms. The
charge was stirred and ground by means of iron agitators
with water and a large quantity of mercury, most of which
remained as a pool at the bottom of the tub. The agita-
tion was continued for a few hours, and the mercury was
then drawn off and filtered through canvas for the

separation of the amalgam. The ore and water was run

off into settling pits for further treatment if
necessary.
The arrastra was also used for direct amalgamation. It

consists of a shallow circular pit, with its floor constructed

of a layer of hard stones, over which other stones are
dragged. The drag-stones are attached to horizontal arms
projecting from a vertical revolving shaft in the centre.
The ore ground between the drag- stones and the pave-
is

ment with mercury and a little water. Very small

quantities of mercury were used, because liquid mercury
would escape through the pavement. This machine was
introduced into America in 1557, at the same time as the
patio process.
156 THE PEECIOUS METALS

In the patio process, which has been used in Mexico for

over 300 years, the ore is coarsely crushed in stamp-mills
or Chilian mills, and subsequently ground fine in arrastras.

Mercury is added to the pulp in the arrastra to collect the

gold and the native silver. The fine pulp is then diluted
to allow the amalgam to separate by settling, and the pulp

is subsequently run off or baled out and conveyed to a

paved courtyard called the patio. Here it is allowed to dry

partly and is then spaded over and additions made of mercury,
common salt, and either sulphate of copper or magistral, a
mixture of CuSOi, FeS04 and oxides, formed by roasting
copper carbonate ores with pyrites. After each addition
the heap of ore, or torta, is mixed with the chemicals,
usually by being trodden
over by mules. The torta
becomes hot owing to the chemical actions, and water is
sometimes sprinkled on if the mixture becomes too dry.
After the lapse of some time and much mixing, the
sulphides of silver are decomposed by the cupric chloride
formed from the interaction of CuSO* and NaCl.
The reaction may be expressed thus :

Ag2 S + CuCl 2 = 2AgCl + CuS.

The chemical action is due entirely to the difference of the
solubilities, which explains the facts that pure sulphate of
copper reacts only feebly with sulphide of silver, and that
the reaction is accelerated by sodium chloride. 1
As to complex sulphides, they are decomposed as follows,
under favourable conditions :

2Ag 3 AsS 3 + 3CuCl 2 = GAgCl + 3CuS + As S 2 3

2Ag3 SbS 3 + 3CuCl 2 = GAgCl + 3CuS + Sb S 2 3.

1
Yondracek, Revue Universelle des Mines et de Metalluryie, Vol. XIII.
(1908), 4th series, p. 105; Revue de Metallurgy, September, 1908, p. 678.
 THE EXTRACTION OF SILVER FROM ITS ORES 157

1
Generally, however, some sulphur is set free.
In a similar way cuprous chloride can decompose silver
sulphide, thus :

2Ag 2 S + Cu Cl =
2 2 2AgCl + 2CuS + 2Ag.
There is also chloride of iron present in the torta, and
this gives the equation

A g2 S + Fe Cl = S + 2AgCl + 2FeCl
2 6 2.

Finally the silver chloride is reduced by mercury with

the production of silver amalgam and calomel, thus :

2AgCl + 2Hg = H g2 Cl + 2Ag. 2

Although all the compounds of silver, when in solution,

can be reduced by mercury as well as by zinc, copper and
iron, this is not of much use in practice, because the

sulphide and sulphantimonide of silver are almost in-

soluble, and are decomposed by mercury very slowly. It is

necessary to hasten the operation by previous conversion

of these compounds into the chloride.

Some action takes place in the following manner:

2CuCl 2 + 2Hg = Cu 2 Cl 2 + Hg 2 Cl 2
2CuCl 2 + 2Ag = Cu 2 Cl 2 + 2AgCl
but Vondracek does not consider these reactions to be im-

portant, partly because the loss of mercury which they

would entail would be greater than that found in practice.
The loss of mercury, according to the equation given above,

2AgCl + 2Hg = HgaCla + 2Ag,

corresponds to 400 parts of mercury for each 216 parts of

silver recovered. In practice the loss is only about 330
1
Loc. cit.
158 THE PEECIOUS METALS
1
parts of mercury to 216 parts of silver obtained, owing
doubtless to the presence of metallic silver in the ores and
also to the fact that cuprous chloride reduces some sulphide
of silver directly to the metallic state. The action of cuprous

chloride, however, is never great.

The period of amalgamation in the torta usually lasts
from two to four weeks, samples being examined daily to
determine the progress of the reactions. The process is
suitable only to countries with a warm dry climate. Even
in Mexico cold weather sometimes delays the amalgamation,
and falls of rain dilute the pulp, so that further additions of
chemicals are required.
When the amalgamation complete, some more mercury
is

is incorporated with the torta to assist in collecting the

amalgam. The mixture is then washed in tanks or tubs
generally supplied with revolving stirring-arms. The
amalgam and heavier parts of the ore settle to the bottom
and the remainder is run off through plug-holes. The
amalgam is washed free from sulphides in bateas, and the

concentrates in later times were shipped to smelting works.

The amalgam is strained through filter-bags to separate
the excess mercury, and it is then pressed into cakes
and retorted.
The patio process is now becoming obsolete partly as
a result of the introduction of better means of communi-
cation into Mexico, but it has been of
importance in vital

that country in the past. Many ores yield about 75 per

cent, of their silver when treated by the process, without
the use of much fuel or of expensive machinery.
The cazo or caldron process is a simple method of

1
Loc. cit.
 THE EXTRACTION OF SILVER FROM ITS ORES 159

treating surface ores containing silver, either native or in

the form of chloride or bromide. was formerly in use
It

in Peru and Mexico, but has now been discarded in most

localities. The ore, after being crushed in stamp-mills and
ground fine in arrastras, was charged into amalgamating
vessels with mercury and common salt. The small vessels
originally employed were called cazos and consisted entirely
of copper, but in later times vessels of larger size with
wooden sides and copper bottoms came into use. The larger-
sized vessel was called fondon in Mexico, and was 6 or
7 feet in diameter and 2J to 3 feet deep. The vessel was
heated over a fire, and the charge of ore and water raised
to boiling. Common salt was then added with continuous

stirring and when it was dissolved several successive addi-

tions of mercury were made. In the fondon, two copper
blocks fastened to arms attached to a vertical revolving
shaft were dragged round on the copper bottom by mule
power. The
boiling and stirring was continued for about six
hours, and the charge was then run off and washed for the
recovery of the mercury and amalgam. Efforts were made
to prevent the adherence of the mercury to the copper.

In this method the silver chloride dissolved in common

salt is reduced to metallic by the copper, but
silver

sulphide of silver is not decomposed and remains in the

tailings.
It was the tendency of the amalgam to adhere to the
copper plates in this process which probably furnished the
idea of usingamalgamated copper plates in gold-mills.
The Francke-tina process is another process of amal-
gamation, the use of which is confined to South and
Central America. The ore after being crushed is roasted
160 THE PRECIOUS METALS

with salt in reverberatory furnaces, with the object

of converting sulphides of silver into chloride. The
chloridised pulp is then treated in tinas, or pans with
wooden and copper or bronze bottoms. A solution
sides
of 10 per cent, of sea salt is added to the pulp, and the

charge is stirred with wooden or iron mullers provided with

copper shoes and heated by means of steam. Mercury is
is

added at intervals, but soluble copper salts are not added,

as they are contained in the roasted ore and are dissolved
from the copper bottom or sole-plate and the copper side-
plates which are fixed to the wood inside. The silver
chloride and metallic silver
amalgamated.are The
chemical reactions which take place have the effect of
causing metallic copper to be dissolved, and the sole-plate
is worn out and requires renewal in a few months.

The Kronke process, which was introduced into Chile

in 1860, depends on the use of a hot solution of cuprous
chloride in common salt for the
decomposition of the refrac-
tory silver compounds. The silver is finally reduced to the
metallic state by means of an amalgam of zinc or lead.

The crushed ore is treated in rotating wooden barrels with

about 40 per cent, of its weight of the solution, which makes
a thick mud with the ore. The solution is prepared by
adding common salt to a strong solution of copper sulphate
untilno more is dissolved. The cupric chloride so formed
is afterwards converted into cuprous chloride by heating
it with metallic copper. The cuprous chloride remains
dissolved in the common The solution is then slightly
salt.

acidulated with sulphuric acid and used at once, being

charged-in hot.
After the barrel has been rotated for about half an hour,
 THE EXTRACTION OF SILVER FROM ITS ORES 161

mercury is added, with a small quantity of lead or zinc

amalgam dissolved in it. The amount of mercury is 10
per cent, or more of the weight of the ore. After rotating
for four to six hours longer, the barrel is filled with cold
water and rotated less rapidly for an hour or more to
collect mercury and amalgam. The barrel is then
the

emptied, and the amalgam is purified by a hot solution of

cupric chloride and finally by treatment with ammonium
carbonate. The process is suitable for treating highly
refractory ores.
The pan amalgamation process is doubtless an outcome
of the cazo and fondon processes, although it also bears
some resemblance to the obsolete Freiberg barrel amalga-
mation process, in which ores, after being roasted with
salt, were treated in rotating barrels with mercury, iron

and water. The ore is generally crushed in stamp batteries,

either wet or dry, double-discharge mortars being used in
either case. In the Washoe process the pulp is charged
into amalgamating pans (see Fig. 32) with water and
mercury. The pan is a flat-bottomed circular vessel
about 5 feet in diameter and 2 feet 6 inches deep, with a
capacity of about 1 tons of ore at a charge. The bottom
is of iron, with an iron flange reaching about to the top
of the muller, and the rest of the sides are of wood. In
the centre is a vertical revolving shaft carrying hori-
zontal iron arms or mullers. These can be lowered
so as to touch the bottom, or raised some distance. At
they are lowered, and the pulp is ground fine between
first

them and the bottom, detachable shoes and dies being used
to take up the wear.After the pulp has been completely
slimed, the mullers are raised by about half an inch, mercury
P.M. M
162 THE PEECIOUS METALS

FIG. 32. Amalgamating Pan : Eraser and Chalmers Pattern.

 THE EXTRACTION OF SILVER FROM ITS ORES 163

is added, and the charge is stirred by the mullers for some

time longer. The thickness of the pulp is such that globules
of mercury become dispersed through it without sinking to

the bottom. The amalgamating pans are heated by steam

jackets at the bottom, and amalgamation is also promoted

by the addition of chemicals, such as sulphate of copper

and common salt. These act in the same way as in the
patio process, but far more rapidly owing to the higher
temperature in the pan. The metallic iron of the pan is
also serviceable in decomposing sulphide and chloride of
silver. Caustic alkalies are sometimes used to remove

grease which may have been introduced into the ore from
the machinery.
When the amalgamation is complete that is, after about
four hours the speed of the mullers is reduced from their

ordinary rate of 60 or 80 revolutions per minute to about

30 or 40 revolutions per minute, and the pulp is diluted by
the pan to the top with water. A few minutes later
filling
the amalgam, now for the most part collected at the bottom
of the pan, is drawn off into a bowl, and the pulp is run off

through plug-holes into the settling pan. This is a vessel

similar to the amalgamating pan, but larger (8 feet or more
in diameter), and provided with wooden shoes to the mullers.
The bottom pan slopes down from the middle
of the settling

towards the periphery, where an annular groove is pro-

vided to collect the mercury. This is drawn off either at
intervals or continuously through a mercury trap.
In the settling pan the pulp is further diluted and the
pulp is stirred, the mullers revolving at the rate of 10 or
20 revolutions per minute. The lighter parts of the pulp
are thus kept in suspension, whilst the mercury, amalgam,
M 2
164 THE PEECIOUS METALS

particles of iron,and the heavier portions of the ore settle

to the bottom. The settling process is usually arranged
so that the settler can be discharged and prepared for the
next charge as soon as it is ready for transference from
the amalgamating pan. Plugs at different levels in the
side of the settler are withdrawn successively and the
muddy liquid run off, stirring being continuous. The
amalgam and mercury from the settler are mixed with
that from the pan, and the whole is treated together.
Sometimes no mercury is separated in the amalgamating
pan, but the whole charge is transferred to the settler.
In some mills the continuous process of pan amalgamation
is in use, the pulp passing automatically through a series
of pans, the first two or three of which are arranged for
grinding.
The treatment amalgam, by whichever process it
of the

is obtained, begins with the addition of more mercury, and

it is then stirred with water in a clean-up pan or rotating

barrel for some hours. The impurities in the amalgam

are thus removed and carried off suspended in the water,
and the mercury is then strained through conical bags of
canvas. These are suspended from an iron ring and kept
in a lock-up safe. The excess mercury passes through the
filter-bag, leaving theamalgam in a pasty condition, con-
taining about six or seven parts of mercury to one of silver.
By the use of pressure more mercury could be squeezed out.
The amalgam is distilled in horizontal iron cylindrical
retorts, which s-ometimes take a charge of as much as a

ton. The amalgam is usually placed in sheet-iron trays

which fit the retort. The temperature is raised gradually

to a bright red heat, and the mercury is distilled off and

 THE EXTRACTION OF SILVER FROM ITS ORES 165

condensed in water. The silver residues, still containing

about 1 per cent, of mercury, are melted in crucibles and
cast into bars.
Oxidised ores are usually treated in pans without previous
roasting, as described above, but sulphide ores are generally
roasted with salt to convert the sulphides of silver into
silver chloride before being amalgamated (Reese river pro-
cess). In this case the ore is crushed dry and roasted in

rotating tube furnaces, reverberatories, or shaft furnaces.

The rotating furnaces are large iron tubes lined with fire-
brick. The ore is tumbled over and over by the movement
of the furnace, so that hand-stirring is unnecessary. The
heat supplied by flames passing into the tube. In
is

another type of furnace the hearth is stationary, and the

ore is stirred and moved towards the discharge end by the
passage of iron ploughs attached to an iron chain. The
salt is usually added after most of the sulphur in the ore
has been oxidised, as otherwise the loss of silver by
volatilisation becomes considerable. After roasting, the
ore is allowed to cool before being charged into the

amalgamation pans.
The amalgamation of roasted ore is far more rapid and
effective than the treatment of sulphide ore before roasting.
Mercury acts slowly on sulphide of silver, forming silver

amalgam and separating the sulphur. The action is

quickened by the presence of iron, which is supplied by

the wear of the stamps in a finely divided form and also

by the inside of the pan. Amalgamation is also promoted

by heat. Nevertheless the action on metallic silver and
silver chloride, which exist in roasted ore, is far more
rapid than that on unroasted sulphides. The disadvantages
166 THE PRECIOUS METALS

in amalgamation are the extra expense

roasting before
dry as compared with wet
of roasting, the difficulties of

crushing, and the greater impurity of the amalgam, which

contains more base metals if the ore is roasted.

2. LIXIVIATION PEOCESSES.

In these processes the silver is obtained in solution, and

after filtering is precipitated from the clear liquid in the
metallic form or as a sulphide. In the Zierrogel process,
which is only applicable to argentiferous copper mattes
(see below under " Smelting "), the material is roasted in an

oxidising atmosphere, and the silver is converted into silver

sulphate, the copper being transformed into cupric oxide.

The charge is then washed in tubs with hot water, in
which the silver sulphate is dissolved, and metallic silver

is precipitated by scrap copper.

The roasting must be carefully performed, the tempera-
ture being gradually raised and not allowed to become too
high. Iron sulphate is formed at first by the oxidation of
iron sulphide. Oxides of iron and copper are also formed
simultaneously with the sulphate of iron. As the tempera-
ture rises, the sulphate of iron is decomposed, sulphurous

anhydride and oxygen are evolved, and copper sulphate is

produced. At a still higher temperature copper sulphate is

decomposed and silver sulphate formed. This is stable at
the temperature of decomposition of copper sulphate, but if
the heat becomes too intense the silver sulphate is split up
and the charge is spoilt. If the matte contains lead, bis-
muth, arsenic or antimony, some silver remains insoluble
even after the roasting, and the process is unsuitable. The
 THE EXTRACTION OF SILVER FROM ITS ORES 167

finish of the roast is denoted

by the presence of a very faint
blue colour in the liquid on treatment of a sample with
water, and the appearance of a thick, curdy precipitate of
silver chloride on the addition of common salt.

The roasted matte is extracted with hot water, and after-

wards with copper sulphate solutions resulting from the
precipitation of the silver by copper, acidulated with sul-

phuric acid. The cement or precipitated silver is washed,

pressed, dried and melted.
Another group depends on the conversion of
of processes

the silver into chloride, which is then dissolved by common

salt (Augustin process), sodium thiosulphate (Patera pro-

cess) or calcium thiosulphate (Kiss process). The silver

is converted into chloride by roasting with common salt,

as in the Reese river pan amalgamation process (see

above) .

The Augustin process has been superseded almost every-

where by other methods.
The solubility of silver chloride in brine solutions is

comparatively slight (about 0'12 part AgCl in 100 parts of

saturated solution of NaCl, see above, p. 148), and other

compounds of silver, as well as metallic silver existing in

the roasted ore, are not dissolved at all. Ores seldom yield
more than 75 or 80 per cent, of their silver to treatment

by this process. Lixiviation is commonly carried out in

wooden tubs, and the liquids are run through a layer of

granulated copper on which the silver is precipitated.

Cuprous chloride is formed and dissolved in the salt solu-
tion, and the copper is recovered by precipitation with scrap-
iron. The cement silver is washed first in hot water and
then in hydrochloric acid before being dried and melted.
168 THE PEECIOUS METALS

Sometimes this process is followed by treatment with

thiosulphate solutions.
The Patera process is more efficacious than the Augustin
process, as silver chloride is far more soluble in thiosul-

phate solutions than in brine. A soluble double thio-

sulphate of silver and sodium is formed from which the
silver isprecipitated sulphide by sodium sulphide,
as
sodium thiosulphate being regenerated. The roasted ores
are treated in large wooden vats holding fifty tons or more,
with false bottoms consisting of filter-beds. The precipita-
tion of the silver is effected in a separate vat, and the

sulphides are filter-pressed, dried carefully and charged into

a bath of molten lead, which is then usually cupelled.
The roasting of the ore with salt must be carried out with
greater care than when it is a preliminary to amalgamation,
for the reason that metallic silver, which is readily

amalgamable, isonly slowly dissolved by sodium thiosul-

phate. Undecomposed sulphide of silver is unattacked by
thiosulphate, but, unlike brine, thiosulphate solutions dis-
solve the silver from arsenate and antimonate of silver,

and accordingly the process is applicable to ores contain-

ing some arsenic and antimony. Lead and copper are
injurious, going in part into solution and being precipitated
with the silver in the form of sulphides, so that the bullion
is contaminated. The ill effects of these metals in the ore
may be in part removed by lixiviation with cold water
before the application of the thiosulphate solution. Sul-

phates and chlorides of copper, iron and zinc and chloride

of lead are dissolved by the water and removed. If hot

water is used a considerable amount of silver is dissolved

in the wash water and has to be precipitated separately.

 THE EXTKACTION OF SILVER FROM ITS ORES 169

The with thiosulphate takes place in the same

lixiviation

vat as the washing with water, the solution containing

from 0'25 to 2 per cent, of sodium thiosulphate. The
solution of silver chloride proceeds in accordance with the

following equation :

2Na 2 S 2 3 + 2AgCl = Ag2 S 2 3 .Na 2 S 2 3 + 2NaCl.

The solubility of silver chloride is reduced by the

presence of caustic alkalies and alkaline earths. Arsenate
and antimonate of silver, silver oxide and metallic silver

and gold are also dissolved, especially in hot solutions.

Sodium thiosulphate is slowly decomposed on exposure to
the air, absorbing oxygen and forming sulphate of sodium
and sulphur, with the result that the solvent power of the
solution is reduced. This effect is partly counteracted by
the method introduced by Hofman of precipitating silver

with calcium sulphide instead of sodium sulphide. This

destroys sodium sulphate, producing sodium thiosulphate,

calcium sulphate and silver sulphide. Some calcium
thiosulphate is also produced.
The is run through the bed of ore until silver
solution
cannot be detected in the effluent liquid. Water-washes
are then applied, and run with the solution into the
precipitating vats. The sodium sulphide for precipitation
is made by adding sulphur to a strong hot solution of
caustic soda. The silver is thrown down in accordance
with the equation

Ag2 S 2 3 .Na 2 S 2 3 + Na 2 S.
2
= Ag2 S + 2Na S 2 2 3 + S.
Care is taken to avoid an excess of sodium sulphide.
The excess sulphur is removed from the precipitate by
treatment with caustic soda or by distillation or careful
170 THE PRECIOUS METALS

roasting, but care is taken not to allow the temperature to

rise sufficiently for the sulphides to become ignited. The
dried sulphides, which contain from 25 to 40 per cent, of

silver, are then usually charged into a bath of lead already

molten on a cupellation hearth.
In the Kiss process, calcium thiosulphate is used instead
of thesodium salt, the methods and plant being similar to
those used in the Patera process.
In both processes, however, undecomposed sulphide of
silver as well as most of the metallic silver is left un-

dissolved, and in the Russell process a solution of sodium-

copper thiosulphate is applied after the simple thiosulphate
has been run through. This so-called extra solution con-
tains the salt 4Na2 S 2 03.3Cu 2 S 2 3 + xR<f>: The extra
solution quickly dissolves metallic silver and decomposes
silver sulphide, copper sulphide being precipitated and
silverpassing into solution. The copper- sodium salt is

not regenerated, as the copper is precipitated with the

silver when sodium sulphide is added. The Russell process
has not been uniformly successful, and has never passed
into wide use.
The cyanide process has been applied to the treatment of
silver ores in Mexico, and has proved very successful during
the last few years. The result has been that the patio

process and lixiviation by thiosulphate solutions have been

1
displaced and are now mainly of historical interest. The
principles are the same as in the treatment of gold ores,
and the methods employed do not differ essentially from

1 " Some Features of Silver Ore Treatment in Mexico," by W. A.

Caldecott, Jour. Chem. Met. and Mny. Soc. of S. A., Yol. YIIL (1908),
p. 204.
 THE EXTRACTION OF SILVER FROM ITS ORES 171

those already described on pp. 110 124. The ore is crushed

in stamp-mills and on some mines converted into slimes by
regrinding in tube-mills. The pulp is then separated into
sands and slimes by water-classification, the sands being
treated with cyanide by percolation and the slimes by

agitation, either by means of mechanical stirrers or by air-

blowing. The length of time of treatment is considerable,

owing to theslow rate of solution of silver in cyanide solu-
tions.
1
The solutions of cyanide in use are much stronger
than those employed in gold extraction. Solutions con-
taining 1 per cent. KCy are not unusual, and even in slime
treatment, according to Caldecott, solutions containing O'l
per cent, are frequently used. The values in the ores are
chiefly the form of sulphide of silver, although some
in
metallic silver, silver chloride and gold are also present.
The last-named bodies are readily dissolved, but silver

sulphide is
comparatively slowly on and requires acted
2
strong solutions. Caldecott has shown that its dissolution
depends on secondary reactions by which soluble sulphides
are removed from the cyanide solution. The primary
reaction
Ag 2 S + 4NaCy = 2NaAgCy2 + Na 2 S
is stopped by the sulphide sodium, which would
of

reprecipitate silver if present in excess, and soon establishes

an equilibrium. Of the secondary reactions the following
is the most important :

2Na 2 S + 20 2 = Na 2 S 2 3 + Na2 0.
This is followed by
Na 2 S 2 3 + Na 2 + 20 2 = 2Na 2 S0 4 .

1
Lor. ,-if.

2
Jour. <'ltein. Met. a,l M,,,j. Site, of ti. A., Vol. VIII. (1908), p. 266
172 THE PRECIOUS METALS

Like the other secondary reactions, this involves the

absorption of oxygen, so that aeration is as important in
the treatment of silver ores as it is in that of gold ores.
Soluble sulphides are also removed or rendered harmless

by the addition of lead acetate.

Largely owing to the considerable strength of the

solutions used, the consumption of cyanide is very great,
amounting to from two pounds per ton of ore. The
to five

silver is precipitated on zinc shavings which are not coated

with lead, and in cleaning-up the black silver slimes are

briquetted and melted in crucibles with fluxes without any
preliminary treatment with acid.
In the treatment of gold and silver ores alike, the

practice is
growing of using dilute cyanide solutions instead
of merely water in the stamp battery, and thus to begin to
dissolve the valuable metals from the outset. Even in the
treatment of some gold ores the amalgamated plates are
dispensed with, and reliance placed entirely on cyanide to
extract the values. After the gold and silver have been

dissolved, the liquid is now often removed through vacuum

filter frames, which are lowered into the vats and immersed
in the solution. This method is used in the Moore, Ridgway
and other filters.

3. SMELTING PEOCESSES.
These processes, which are largely carried on in North
America, and to a less extent in Germany, have for their
object the concentration of the silver in lead. The smelting
is carried out in blast furnaces of moderate size into which
a mixture of ore, fuel and fluxes are charged. The com-
position of the charge is determined beforehand, and the
 THE EXTEACTION OF SILVER FROM ITS ORES 173

quantities of ingredients required can then be calculated.

The lead usually amounts to about 10 or 12 per cent, and
the coke to from 7*5 to 15 per cent, of the charge. The
slag is a fusible silicate of iron and lime with a number of
other bodies present as impurities. It is usual to mix a
number of ores together so as to obtain a charge of the

required composition. Great heaps are formed consisting

of thin layers of sulphide ores, ferruginous ores, lead ores,

zinciferous ores, and even quartzose ores, and vertical slices

are taken for charging into the furnaces together with coke
and limestone. If gold ores require smelting, they are
mixed with the heaps. Sometimes the ores or part of them
are roasted to reduce the percentage of sulphur before they
are smelted. Sometimes poor sulphide ores are smelted
without lead and a matte (sulphides of iron, copper, etc.)
formed in which the silver is contained. The matte is then
separated from the slag and treated like a sulphide ore.
The reduced lead accumulates in the hearth at the
bottom of the blast furnace, and is run off either con-

tinuously or at intervals of time. The pigs of lead are

assayed and passed forward for desilverisation. The slags

are also assayed, and if sufficiently poor in silver and lead
are thrown away, but a portion of them is often returned
to the blast furnace.

The blast furnaces employed are either circular or rect-

angular in section. The latter shape is preferred for the

reason that much larger furnaces can be employed, the

rectangular section being as much as 12 feet by 3 feet

6 inches at the tuyer level, while the round furnaces are
not more than about 5 feet in diameter. Both types
widen from the tuyer level to the mouth. If larger round
174 THE PRECIOUS METALS

furnaces were used, the blast would not penetrate to the

middle, which would not be sufficiently heated. Cold air is
used for the blast to prevent the temperature of the furnace
from rising too high, as in that case losses of lead by
take place.
volatilisation The pressure of the blast is
usually from two to four pounds per square inch. The
number of tuyers varies with the size of the furnace, round
furnaces often having eight tuyers, and large rectangular
furnaces six on each of the long sides. The height of the
tuyer level above the hearth is 2 or 3 feet, and the
total height of the furnace from 12 to 30 feet. The walls
of the furnace in the neighbourhood of the zone of fusion
consist of water-jackets, which are constructed of iron or
steel. The hollow space in the water-jackets is about
6 inches wide, and water circulates through them.
The continuous discharge of the lead from the furnace is
effected by a syphon tap, the molten part of the charge
inside the furnace forming one leg of the syphon, the other

being an inclined passage leading up from the bottom of

the hearth to the outside of the furnace, where it terminates
in a bowl which becomes full of molten lead. The lead can
be ladled from this bowl, or tapped from it, or run off it
continuously. Cupriferous ores are not fitted for smelting
with the use of the syphon tap, as alloys of lead and copper
are formed which are liable to freeze in the comparatively
cool passages and obstruct them.
The slag may be run off continuously through slag notches
of the short sides in rectangular furnaces,
placed one in each
or tapped off at intervals. In either case it is usually run
into slag-pots, which consist of cast iron and are carried on
wheels and are cone-shaped with the apex downwards.
 THE EXTRACTION OF SILVER FROM ITS ORES 175

There a tap-hole at a few inches from the bottom.

is The
matte or speiss in the slag settles to the bottom, and after
a time the slag can be tapped off, leaving the matte in the

apex of the cone. This is saved for further treatment.

Generally the slag contains from 0*2 to 0'5 per cent, of lead
and only one or two ounces of silver per ton. The loss of

lead by volatilisation may also be considerable, and is kept

down by working at a moderate temperature with the upper
part of the furnace charge at a black heat. The escaping
gases are carried through condensing chambers, where some
of the volatilised lead is caught.

The smelting of silver ores with lead is most satisfactory

in places where fuel and limestone are cheap and large
quantities of lead ores are available. It is best adapted to

the treatment of ores at large centres where considerable

quantities of ores of all sorts require treatment. It is not

suitable for cupriferous ores and ores poor in silver and free
from lead. Such ores are better adapted for treatment by
wet processes or by amalgamation. The proportion of zinc
in a charge must not be allowed to become excessive, as
zinciferous slags are somewhat infusible.

The pig lead, which is sometimes called " base bullion,"

contains from about O'l per cent, of silver upwards. The
proportion of silver in the lead is generally kept below
2 per cent, in order to avoid undue losses in the slag. The
distribution of the silver in the lead not uniform, and
is

difficulties are encountered in obtaining trustworthy samples

for assay. The silver is concentrated towards the bottom
of the pig, and also to a limited extent towards the outside.

Samples are taken by drilling and also by means of sawcuts,

but neither method is quite satisfactory (see abpve, p. 137).

176 THE PRECIOUS METALS

TBEATMENT OF AEGENTIFEROUS LEAD.

The pig lead, if impure, is refined first and desilverised

afterwards either by Pattinsonisation or by Parke's process,
or it may be cupelled. If it is sufficiently pure, lead is

before being refined.

desilverised The impurities com-
monly contained in the pig lead are arsenic, antimony,

sulphur, copper., zinc, etc.

The Pattinson process depends on the fact that the
eutectic alloy of silver and lead, that is, the alloy of
lowest melting point, contains 2'5 per cent, of silver. If

an alloy of this composition is melted and allowed to cool,

it sets as a whole, the portions solidifying first having the
same composition as the part remaining molten. If molten

alloys containing less than 2*5 per cent, of silver are

allowed to cool slowly, the crystals first formed in the

liquid contain much less silver than the part remaining

molten. In the original Pattinson process, these crystals
are ladled by hand out of a large iron pot in which the
silver-lead alloy is slowly cooling, and an enriched silver-
lead alloy is left molten in the pot. Friedrich has shown 1
that the iron pot has no influence on the result, which
can be equally well attained in an earthenware vessel.
and that pressure up to five atmospheres has no effect
on the crystallisation of the lead. The crystals are
removed by perforated ladles, and either two-thirds or
seven-eighths of the contents of the pot are ladled out
into a neighbouring pot. The operation is repeated until
the rich lead contains about 1 per cent, of silver and the

1 " Blei und Silber."

Metallurgie, Vol. III.,
 THE EXTRACTION OF SILVER FEOM ITS OEES 177

poor lead about 0*001 or 0'002 per cent, of silver. The

limit of richness that could be reached would be lead con-

taining 2'5 per cent, of silver. This process is now seldom

used.
In an alternative method, the liquid alloy is drawn off,

leaving the solidified portion behind. The charge is stirred

by horizontal iron arms attached to a vertical rotating axle
worked by bevel gearing (mechanical Pattinson process), or
by means steam under a pressure of about three
of jets of

atmospheres (Rozan process). It is usual to tap off about

one-third of the lead, leaving two-thirds in the form of

crystals. The operation is repeated until the rich

lead contains about 2 per cent, of silver, and the

poor lead from O'OOl to 0'003 per cent. The Rozan

process is the cheapest method of Pattinsonisation, but
it is more expensive than desilverisation by means
of zinc.

Parkes process depends on the formation of compounds

of zinc and silver when these metals are melted together.
When zinc is in excess, the compound formed probably
corresponds to the formula AgZni 2 containing about 12 per
,

cent, of silver. If molten zinc is stirred with a bath

of argentiferous lead, which is afterwards allowed to

cool, a crust forms on the surface containing zinc and
lead together with the silver and any gold and copper
that may be present. The lead retains about 0*6 per
cent, of zinc dissolved in it at temperatures but little

above its melting point. At 650 lead dissolves about

3 per cent, of zinc, but as the
temperature falls zinc
is separated, and as its alloys are lighter than lead they

rise to the surface. Arsenic, antimony and tin remain

P.M. N
178 THE PRECIOUS METALS

with the lead, and if large quantities of these elements

are present, the lead must be refined before the zinc is

added.
The zinc is thrown solid on the surface of the molten
lead and the temperature is raised until it melts, or in
the alternative molten zinc is poured into the lead bath.
The mixture is then thoroughly stirred and allowed to cool
slowly. The total amount of zinc added is usually from
1J to 2 per cent, of the weight of the lead, but it is added
in two or three lots, the crusts being kept separate. The
first crust contains practically all the gold and copper, and

if the first addition of zinc is small, hardly any silver is

removed by it.
The crusts are removed from the surface as soon as they
are formed by means of perforated ladles, and the ladling
is continued until 'the lead begins to solidify. Fresh
additions of zinc are made until the lead contains only
about 0'0005 per cent, of silver (or one- sixth of an ounce
per ton). This is much poorer than can be profitably
attained by the Pattinson process.
The zinc crusts orscums contain a high percentage of
lead, for the most part mechanically mixed with the other
constituents. The lead is partly separated by liquation,
the mixture being raised to a temperature above the melt-
ing point of lead, but below that of the zinc alloys.
The lead is then allowed to flow away, leaving the en-
riched scum behind. The liquated lead must be again
desilverised.
The rich scums are usually treated by retorting, the
zinc being distilled and condensed so that it is available
for further use. The residue consists chiefly of lead, and
 THE EXTRACTION OF SILVER FROM ITS OEES 179

contains from 5 to 10 per cent, of silver. It is passed to

the cupellation furnace.
Parke's process is carried out in large cast-iron pots

holding as much as 50 tons, as against 10 or 15 tons, the

usual charge in the Pattinson process. Besides its cheap-
ness, desilverisationby zinc has the advantage of quickness,
so that the poor lead is soon available.
The cupellation of argentiferous lead is a very ancient

operation, being referred to in the book of Jeremiah

(B.C. 600). It consists in melting the lead in a reverberatory

furnace, and oxidising it by means of a blast of air directed

on to its surface. The litharge accumulates round the edge
of the molten metal and is allowed to flow away. Finally
silver only is left unoxidised.
The operation is carried out on hearths of bone-ash, marl
or magnesia, covered by a movable hood (German cupella-
tion furnace), or on smaller movable hearths, the roof

being fixed (English cupellation furnace). The use of

bone-ash in the hearth is not on account of its absorptive
properties, as in assaying, but because it resists the corro-

sive action of molten litharge. It is now seldom used, as

marl is cheaper. The litharge first formed contains such
impurities as iron, zinc, arsenic, antimony, tin, cobalt and
nickel, which oxidise readily. Towards the end of the
cupellation the litharge contains copper, bismuth and
silver. Cupelled silver may be as fine as 997 or 998,
but must often be cupelled again with fresh lead to raise
it to marketable fineness. The gold remains with it and
must be removed by parting.
The German cup filiation furnace, which is usually round,
takes from 10 to 50 tons of argentiferous lead. In the larger
N 2
180 THE PRECIOUS METALS

furnaces the work is not carried on until all the lead is

removed, as there would be a difficulty in keeping the

temperature high enough to prevent the silver from freezing
were nearly pure. The fuel used is coal, wood, or peat,
if it

and the oxidation is effected by a blast of air from two or

three tuyers directed on to the surface of the lead bath.
The first products of oxidation consist largely of arsenate
and antimonate of lead, which form a dark-coloured scum
(abstrich). This is not readily melted, and is raked off.

Then nearly pure litharge begins to form, and is drawn off

through a channel cut in the hearth. As the operation

continues, and the molten charge is diminished, the channel
is deepened. At the end of the operation the litharge
clears off from the surface of the charge, which becomes a

bright mirror reflecting the roof of the furnace. The

blicksilber thus obtained contains a few per cent, of
impurities, chiefly lead. It is broken up and recupelled in
a smaller furnace for the production of fine silver. After
each charge has been worked the hearth is broken up and
smelted for lead, and a new hearth is prepared.
The English cupeWition furnace is used for richer silver-
lead alloys than the German furnace. As a rule the first
cupellation of lead, containing, say, 10 per cent, of silver, is
partial only, the charge being enriched until it contains
60 or 70 per cent, of silver, when it is ladled out and fresh
lead charged into the furnace. The blast is usually supplied
through one tuyer, the oval test or hearth being only
3 or 4 feet long and 2 or 3 feet wide. The final cupellation
is carried out at a higher temperature, and when oxidation
is finished the silver is cast with the help of ladles. The
tests are used continuously until worn too thin for further
 THE EXTRACTION OP SILVER FROM ITS ORES 181

use. A worn-out test can be withdrawn and another one

substituted in the furnace with very little delay. The test
is contained in a movable test-frame, in which the marl is

rammed. A new test, however, requires very careful drying

before it can be used.
The English cupellation furnace is used for the produc-
tion of silver from rich lead, and the litharge produced
generally contains too much silver for it to be sold without
further treatment. In the German cupellation furnace,
poorer lead is cupelled, and quantities of marketable litharge
are usually produced. The German furnace is useful when
required to treat rich gold, or silver-bearing materials
it is

by charging them into a bath of molten lead.

4. ELECTROLYTIC PROCESSES.

The only
electrolytic process which has passed into wide
use that used for the separation of silver from argenti-
is

ferous copper obtained by smelting. The copper is cast

into slabs which form the anodes, the bath consisting of

sulphuric acid and sulphate of copper. The copper is

dissolved at the anodes together with iron and other metals

electro-positive to it. the electro-negative metals gold,

platinum, silver, etc. remaining undissolved if the current

is of the proper density. The copper is deposited at the
cathodes, and the electro-positive metals accumulate in the
solution, which must be drawn off and renewed after a time.

the current density is too great, silver will be dissolved

If

at the anode with the copper and reprecipitated at the

cathodes. Some free acid in the bath is necessary, as silver
goes into solution more readily in neutral solutions.
182 THE PBECIOU8 METALS

The anode slime may contain over 50 per cent, of silver.

It is purified in various ways, such as hy cupellation with
lead or by the Moebius electrolytic process, or by dissolving
it in
sulphuric acid and reprecipitating the silver with
copper.
In the Moebius process the electrolyte is a mixture of
nitric acid and silver and copper nitrates.About one per
cent, of free nitric acid is kept in the bath. Silver and

copper are dissolved at the anodes, leaving the gold undis-

solved. Silver is precipitated at the cathodes, leaving the

copper in solution. If too much copper accumulates in the

solution (more than 4 or 5 per cent.) it would precipitate
on the cathodes simultaneously with the silver, and part of
the liquid is in that case withdrawn and water is added to
reduce the percentage of copper present. An addition of
more nitric acid also assists in preventing the copper
from being deposited. Very large amounts of silver are
now refined by this process in the United States.
 CHAPTER XII

REFINING GOLD AND SILVER

THE gold bullion produced by the methods of ore treat-

ment described in Chapters VI. to VIII. is impure and unfit
for use in the arts until after it has been refined. Retorted
metal from the amalgamation process often contains 2
or 3 per cent, of copper and iron, besides some mercury
and smaller amounts of other base metals. Silver is

always present, in amounts varying from about one per

cent, up to more than half the total weight. Sometimes
bullion is sold to the Australian mints containing only a
few per cent, of gold, the remainder being chiefly lead,
zinc and silver. This is produced by means of the cyanide

process. The presence of 20 or 30 per cent, of iron, lead

and zinc in gold bullion gives it a worthless appearance,

bearing some resemblance to cast iron, but with an earthy
fracture. Most bullion, even if it has the outward appear-
ance of gold, is brittle before it is refined.
The bullion is often subjected to rough refining opera-
tions at the mine before being sold. For example, at some
American gold-mills the bullion is melted in a crucible and
a blast of air directed on its surface. The base metals are
oxidised and form dross which can be skimmed off. Some-
times oxidising agents such as nitre or manganese dioxide
are melted with the bullion. Care is taken, however, to
leave part of the copper, etc., unoxidised in order to avoid
184 THE PKECIOUS METALS

losses of silver in the slag or dross. In South Africa and

elsewhere the gold precipitate from the cyanide process is

often refined by cupellation with an excess of lead (Tavener

process, q.v.).

When the bullion isto London, Philadelphia,

brought
Melbourne, St.Petersburg and other great centres, it is in
the first instance melted and assayed in order that its value

may be ascertained. The assay pieces sometimes consist

of a small portion dipped out of the molten gold, and
sometimes of portions cut from the cast ingots.
After the value has been agreed upon, the bullion is at
once refined, the base metals being removed, the gold and
silver separated and fine bars of each obtained.

There are a number of different processes for parting

gold from silver, some of them of great antiquity, but only

three are now in use on a large scale. These are the
sulphuric acid process introduced in Paris in 1802, the
chlorine process, introduced at the Sydney Mint in 1870.
and the electrolytic process, introduced at Hamburg in
1888. Until recently nitric acid was used as an adjunct to
the sulphuric acid process in Philadelphia, and it formed

by far the most important parting agent during the three

centuries before the introduction of sulphuric acid. It is

still used for parting gold from silver in assaying, as in

that case the extra cost no importance. Sometimes

is of

gold is dissolved in aqua regia, and precipitated by ferrous

chloride or some other suitable agent, the impurities being
left in solution. This method does not separate tellurium
from the gold, and is too costly to be widely used.
At the present day gold to the value of more than
50,000,000 is refined annually by the sulphuric acid
 REFINING GOLD AND SILVER 185

process in London and on the Continent, an amount of about

12,000,000 is treated by the chlorine process in Australia,

and about 14,000,000 in the United States by electrolysis.

THE SULPHURIC ACID PROCESS OF BEFINING.

In this process an alloy of silver and gold
is prepared by

melting and granulating, the granulations are boiled in

sulphuric acid by which the silver is dissolved, and the
residue of gold is washed and melted. The silver is pre-

cipitated from solution by metallic copper, and the sulphate of

copper is recovered by evaporation and crystallisation. The

parting alloy is made up to contain one part of gold to about

two and a half parts of silver. The addition of some copper
isadvantageous in promoting the dissolution of the silver,
and the presence of lead is supposed to aid the solution of
the copper, but large quantities of base metals in the alloy
are disadvantageous. The copper must not exceed 10 per
cent, nor the lead 1 per cent.
When gold bullion containing a large percentage
of lead, zinc and other metals (e.g., some bullion pro-
duced by the cyanide process) is to be refined, it is sub-
"
jected to a preliminary operation called toughening."
The gold is melted, and saltpetre is thrown upon its sur-
face and stirred in. Much oxygen is evolved and vigorous
bubbling ensues. The base metals are oxidised and
skimmed off the surface of the molten metal, and more
saltpetre is added until the gold is sufficiently pure, accord-
ing to the judgment of the operator, who is guided partly by
the increasing pastiness of the metal. Some more nitre is
then stirred in and the pot covered over and its temperature
186 THE PRECIOUS METALS

raised for pouring. Toughening is a tedious operation, and

involves considerable losses of precious metal by spirting
and volatilisation. The losses are mainly temporary, as
most can be recovered by grinding and washing
of the gold

the ashes, and by the use of condensing chambers between

the furnace flue and the stack.
The mixture of nitre and oxides
the heavy metals
of

which forms on the top of the charge rapidly attacks the

pot in which the bullion is contained, corroding it in a ring
just above the surface of the metal. In order to prevent
this, bone-ash is sometimes thrown into the crucible and
pushed back from the centre of the charge towards the
periphery so as to form a ring round the inside of the pot,
and so to protect it from the action of the molten oxides.
The bone-ash does not melt but absorbs the oxides, clearing
the surface of the metal. The nitre is projected on to the
eye of metal within the ring of bone-ash.
The gold after toughening still contains copper as well
as silver, but most of the other base metals have been
removed. The metal
rarely of such a composition that
is

it can be parted by acid without any addition. If less than

about two parts of silver to one of gold are present, the

dissolution of the silver would not be complete. After the
acid had acted for some time, it would have no further
effect, the remainder of the silver being protected by the

gold. The best proportion of gold in the bullion is from 18

1
to 25 per cent., according to Dr. Bossier, but parting is still

possible with 33 per cent, of gold, and hardly any difficulties

are introduced by an excess of silver. The silver used for
mixing with the gold consists of ingots containing some
"
Metallurgy of Silver and Gold,"
1
Percy's p. 471.
 KEFINING GOLD AND SILVEE 187

gold they can be obtained, but in London it often

if

happens that such silver is scarce, and the parting silver

is used over and over again. The copper is sometimes
already contained in the silver, as when Mexican dollars
are used for parting; and sometimes old amalgamated

copper plates which contain some gold are available for

making the necessary addition of copper to the parting alloy.
The alloy for parting is melted in graphite crucibles, well
mixed and dipped out by means of ladles. The ladle is held
about three feet above the surface of a copper vessel filled
with cold water, and the molten metal is poured into it in a
thin stream with a wavy or circular motion. Under these
conditions the metal broken up and solidified in the form
is

of small beads, hollow spheres and leafy granulations.

They are caught in a perforated copper pan, which is lifted

out and drained after the pouring is completed.
The granulations are heated in cast-iron kettles with con-
centrated sulphuric acid, of S.G. 1*85, which is raised to

boiling point cautiously. Air is excluded from the kettles

as far as possible, and the sulphurous acid gas which is

given off in large quantities is carried away through leaden

pipes. The amount of acid required is four or five times

the weight of the granulations, but only about half this

amount is added at first. The reactions involved are
represented by the following equations :

2H 2 S0 4 + 2Ag = Ag S0 + S0 + 2H O
2 4 2 2

2H 2 S0 4 + Cu = CuS0 + S0 + 2H
4 2 2

The sulphates are soluble in hot concentrated acid, but tend

to be precipitated when the acid is cooled or diluted.
When the evolution of S0 2 is at an end, the liquid is
!8cS THE PRECIOUS METALS

removed to a settling pot and fresh acid added. After

three boilings the gold residue is strained and washed with
boiling water in filter-boxes until free from acid. The
" "
brown gold is then dried and melted with nitre. Unless
great care is exercised in washing and melting the gold,
some lead remains with it, and the resulting bars are
brittle.

The clear silver solution is poured into large lead-lined

tanks containing hot water and sheet or scrap copper, and
stirred until precipitation is complete. The chemical action
is one of simple replacement

Ag2 S0 4 + Cu = CuS0 4

The copper is not attacked by the dilute acid. The pre-

cipitated silver crystals are washed, dried, and strongly
compressed by hydraulic power before being melted. The
copper solution is evaporated down and allowed to cool
three times, a crop of crystals of sulphate of copper (blue-

stone) being obtained each time, and the acid mother-liquor

when sufficiently purified is used to dilute the solution of
silver sulphate before it is precipitated.
The net by sulphuric acid is stated to be
cost of refining
about one penny per ounce of gold in Europe. In the
United States the cost in the San Francisco refinery was
between four and five cents per ounce of gold in the year
1905-6, exclusive of salaries and rent, and at New York the
1
cost was still greater.
Modifications in the sulphuric acid process were intro-
duced in the United States by F. Gutzkow in 1867. They
1 "
Eeport of the Director of the United States Mint," 1906, pp. 22,
24, 40, 66.
 REFINING GOLD AND SILVEE 189

consist mainly in the removal of the silver sulphate from

the acid by crystallisation, so that the acid can be used

again, and in the reduction of silver from the crystals of

Ag 2 S04 direct by a solution of ferrous sulphate. The
removal of the silver from the acid without too great
dilution enables the acid to be used over again, so that
a greater quantity can be used in the first instance for

parting, with the result that sufficiently pure gold can be

obtained in a short time. The Gutzkow process, however,
was not adopted in Europe, and with the abandonment of
sulphuric acid for parting in America it has come to an end.
The gold produced by sulphuric acid parting is generally
from 996 to 998 fine, and is usually fairly free from lead
and other deleterious constituents. The silver is also about
996 or 998 fine if care is taken to remove all small particles
of metallic copper from the precipitate.

REFINING BY MEANS OF CHLORINE GAS.

This method is applicable especially to the refining of

gold bullion containing only small quantities of silver and
base metals, in places where the supply of silver bullion for

parting purposes is It has been used continuously

limited.

only in Australia, where

appears to be suited to the local
it

conditions. Chlorine gas is forced through clay tubes to

the bottom of clay crucibles containing molten gold, and
rises through it. The chlorine is absorbed by base metals
and silver, and chlorides are formed which rise to the

surface and are skimmed off. The gold is not attacked

perceptibly by chlorine at such high temperatures.
When very large quantities of base metals are present, it
is more economical to remove them in part either by
190 THE PBECIOUS METALS

treatment with nitre, as in the sulphuric acid process, or by

passing oxygen or air through the clay tubes before chlorine
is The oxides formed by the oxygen attack the pots
used.
and tubes, and it is necessary to add sand and a little
borax to the charge and to use thick-walled clay tubes.

When the oxygen begins to pass freely through the gold,

the preliminary operation is concluded, the slag is skimmed

off,more borax is added and the passage of chlorine is

begun. As the chlorides do not attack the clay tubes, thin-
walled tubes are used. The chlorine is generated from

manganese dioxide, sulphuric acid and common salt. A

depth of about 10 inches of gold is used in the pot, corre-
sponding to a charge of about 700 oz. A rapid stream of
chlorine is admitted at first and is entirely absorbed. After
about an hour, when the bulk of the silver has been
removed, the stream of gas is reduced, and the operation is

stopped when the gas comes through the metal freely with-
out being absorbed. The liquid chlorides are then ladled
out and cast in iron moulds, and the gold is afterwards

poured.
The chlorides contain from 5 to 10 per cent, of gold, and
are remelted on the same day in graphite pots which hold
300 oz. of silver chloride. The gold is removed from the
chlorides by means of carbonate of soda, which is thrown
upon the molten charge. Effervescence ensues and the

gold reduced, together with about an equal weight of

is

silver, and settles to the bottom of the pot. Two successive

additions of carbonate of soda are made, amounting together
to about 10 per cent, by weight of the chlorides. The
metal allowed to solidify, and the fluid chlorides are then
is

poured off and cast into slabs which contain less than O'l
 BEFINING GOLD AND SILVER 191

per 1,000 of gold and are ready for reduction. The slabs
of chloride, encased in flannel bags, are first boiled in

water for some days to remove chlorides of copper and

other base metals, and the silver is afterwards reduced by

electrolytic action by means of iron plates, fairly pure

metallic silver and chloride of iron being produced.
The gold produced by this process is about 996 fine, and
is of excellent quality, all the metals which produce brittle-
ness being removed in the early stages of the refining
operations. The silver is usually about 990 fine, but if the
cakes of chloride are not boiled in water, although the

losses of silver are reduced, the resulting bullion is some-

times very impure, containing 20 per cent, or more of

copper. The losses of gold are small, and the cost for
materials, wear and treatment of sweep, and wages is
tear,
about one farthing per ounce of gold at the Melbourne
Mint. 1
Since the introduction of the process at Sydney, in 1870,
ithas been continuously at work in Australia, where it is
favoured by the high price of sulphuric acid and the
absence of auriferous silver suitable for mixing with the

gold for parting purposes. It remains to be seen whether

parting by chlorine will be displaced by the electrolytic

process.
THE ELECTROLYTIC PARTING PROCESS.
Two electrolytic methods are in use which are both neces-
sary, neither being sufficient without the help of the other.
They are the Moebius and the Wohlwill processes, called after
the names of their inventors. In the Moebius process (see
above, p. 182), the electrolyte consists of dilute nitric acid.
"
Metallurgy of Gold," oth edition, 1906,
1
Eose's p. 422.
192 THE PKECIOUS METALS

Silver, copper and some other metals are dissolved at the

anodes, and pure silver is deposited at the cathodes. The
gold remains in the anodes undissolved, but cannot be
made sufficiently pure by this process to be fit for use in
the arts. The anode mud is accordingly melted down and
used as the anodes in the Wohlwill process, in which the
electrolyte is hydrochloric acid. In this process, when the
electric current is passed through the bath, gold, copper,
etc., are dissolved at the anodes and pure gold is deposited

at the cathodes, silver remaining undissolved in the anodes.

The Moebius process was successfully used for some
years for the treatment of dore silver, which contains only
a few parts of gold per 1,000 of silver, before it was applied
in 1904 to the treatment of impure gold bullion at the

Philadelphia Mint.
The Moebius and Wohlwill processes are now used
together, to the exclusion of the sulphuric acid process, at
the United States Mints at Philadelphia, Denver and San
Francisco, and at the New York Assay Office.
In preparing the bullion for treatment at these establish-
ments, gold and silver bullion are melted together to form
a parting alloy containing 30 per cent, of gold and 70 per
cent, of silver, copper, lead, etc. The base metals some-
times amount to 20 per cent, of the alloy. This alloy is
cast into slabs to be used as anodes, which are suspended
in earthenware tanks by means of gold hangers attached to

copper rods. At the Denver Mint there are eight earthen-

ware tanks, each 47 inches long, 26 inches wide and 22
inches deep, arranged in two sets of four tanks. 1 In each

1 " United States Mint for 1906,"

Eeport of the Director of the

p. 59.
 EEFINING GOLD AND SILYEE 193

set the liquid circulates by gravitation, being syphoned

from the uppermost cell into the next, and finally pumped

up from the lowest cell to a distributing tank above the

highest cell. The cathodes are rolled sheets of pure silver

one-hundredth an inch thick, alternating with the anodes.

of

By the addition of a small quantity of a colloid, such as

gelatine,to the electrolyte, the silver is deposited in a

coherent form on the cathodes, instead of in a granular

non-adherent form, which would fall to the bottom of the
1
cell, or form bridges between the electrodes. This
discovery obviates the necessity of the frequent scraping or
renewal of the cathodes. The electrolyte contains 2 per
cent, of free nitric acid and 3 per cent, of silver in solution
as nitrate. The current density is 7*5 amperes per square
foot of cathode surface at Philadelphia and as high as 20
2
amperes per square foot at Denver. The resistance in the
cells is from one to two volts per cell. The deposited silver
isstripped off the cathodes, which are then straightened
and used over again. The gold remains a hard spongy
mass of the original form of the anodes, and obstinately
retaining a small amount of silver, even if subjected to the
current after the point at which oxygen is freely evolved
from its surface. After being washed free from nitrate of

silver, theanodes are dried, melted, and cast into anodes

for the gold cells, if they contain over 950 parts of gold per

1,000. Sometimes the anodes are boiled in sulphuric acid

before being melted. The copper, lead, zinc, etc., accumulate
in the electrolyte, which is renewed at frequent intervals.
In the Wohlwill process the anodes consist of gold 940
"
Eeport of United States Mint for 1905,"
1
p. 62.
2 "
Eeport of United States Mint for 1906," p. 59.

P.M. O
194 THE PEECIOUS METALS

to 950 parts, silver not more than 50 parts, and a few parts
per 1,000 of base metals. The electrolyte consists of a
solution containing from 2 to 10 per cent, of hydrochloric
acid and from 2*5 to 6 per cent, of gold chloride, heated to
a temperature of 60 or 70 C. The higher strengths are
those used at the Denver Mint, where a high current

density is used. If the hydrochloric acid is insufficient

in amount, or if the baths are not hot enough, chlorine is

given off at the anodes and the gold is not dissolved. The
cathodes are sheets of pure gold, about one-hundredth of
an inch thick. Under the action of the current all the
metals in the anodes are converted into chlorides, the
silver chloride remaining undissolved for the most part, and

the other chlorides dissolving. Nearly pure gold is pre-

cipitated on the cathodes, so that the gold in solution is
gradually replaced by copper, lead, zinc, tellurium, etc.,
and additions of gold chloride are made to the electrolyte
at intervals. When the electrolyte has become charged
with base metals it renewed, and the gold in the old
is

electrolyte is precipitated with ferrous sulphate. The

anodes recommended by Dr. Wohlwill were 4 mm. thick,
but are half an inch thick at Philadelphia. The current
density employed is from 30 to 60 amperes per square foot,
and the E.M.F. required is little more than half a volt per
cell.The anodes are dissolved in from 24 to 36 hours.
The anode mud contains gold, but consists chiefly of silver
chloride. It is boiled in sulphuric acid, which converts the
AgCl into Ag 2 S0 4 The average fineness of the fine gold
.

returned from the refinery of the Denver Mint during the

year ended June 30, 1906, was 999'85. In this period

1,085,733 oz. of standard gold were refined at a total cost,
 REFINING GOLD AND SILVER 195

exclusive of salaries and rent of building, of $44,712 or 4'1

1
cents per ounce. It is probable that future costs will be
lower, and it is stated that they need not exceed 1 cent

per ounce.
The advantages claimed for the process are :

(1) The small amount of acid consumed. This is stated

to be only about 2 per cent, of that which would be

required to dissolve the gold without the aid of the

current of electricity.
(2) The absence of acid fumes. At the Philadelphia
Mint the battery of cells is placed in an ordinary room
without any special method of ventilation, but no incon-
venience is experienced. In both the sulphuric acid and
chlorine processes the fumes are disagreeable and unhealthy.
(3) Platinum, palladium, etc., are recovered instead of

being left mixed with the gold. In the Wohlwill process

these metals are dissolved with the gold, but are not

precipitated at the cathodes, and accumulate in the solution.

\Yhen the amount of platinum in solution reaches about
5 per cent., or the palladium in solution exceeds 0*5 per

cent., these metals begin to be precipitated at the cathodes.

The operation is then stopped, and the platinum is pre-

cipitated with ammonium chloride and recovered. In a

period of twelve months, in 1899-1900, the Hamburg refinery

recovered 1*5 kilograms of platinum from 2,000 kilograms
of gold. The platinum was stated to be worth more than
the cost of refining.

(4) Floor space saved and few vats are required.

is The
space required for the treatment of gold of the value of
2,500,000 per annum is put at 50 feet by 30 feet.
"
Report of United States Mint for 1906," p. 22.
1

o 2
196 THE PEECIOUS METALS

(5) The gold and silver produced are of the highest

possible quality. All the elements which make the gold-
copper alloys brittle are entirely eliminated. Since the
United States Mints,
final installation of the process in the

brittle coinage bars have become unknown, and it has

been found possible to discontinue the annealing of coinage

bars in the course of rolling.
On the other hand, there are difficulties and limitations
of the process. Bars containing less than 850 per 1,000
of gold cannot be treated by the Wohlwill process, and in
practice the successive use of the Moebius and Wohlwill
processes is found to be most satisfactory. The double
treatment makes the whole process of refining gold some-
what slow, and this is serious if a loss of interest is

occasioned. Interest at 3 per cent, per annum amounts

to one-twelfth of a penny per fine ounce of gold per day.
The exact length of time required to refine gold
by the
electrolytic process has not been published, but it appears
to take about a week, against two or three days by the

sulphuric acid process.

 CHAPTER XIII

THE ASSAY OF GOLD AND SILVER ORES

THE quantity of gold contained in ores is too minute to

allow of its being accurately determined by wet methods,
and the usual method of assay is to melt the sample with

materials yielding metallic lead, which takes up all the gold

and silver and sinks to the bottom of the molten charge.
Suitable fluxes are added to make the ore readily fusible at
a moderate temperature. After the charge has been allowed
to solidify, the lead is detached from the slag and sub-
jected to cupellation. In this operation the lead is melted
in a cupel or little cup of a porous material (bone-ash or
magnesia) and oxidised by a current of air. The litharge,
PbO, so formed is absorbed by the cupel, but the gold and
" "
silver, the noble metals, resist the fire and are not
oxidised, and, although melted, their high surface tension
prevents them from being carried into the pores of the bone-
ash. They remain as a bead of metal on the surface of the
cupel, and after cooling can be detached and weighed. The
silver is parted from the gold by boiling in nitric acid.
The process of ore assay is divisible into the following

sections :

(1) Preparation of the ore for assay by sampling and

crushing.
(2) Concentration of the precious metals in a button of
"
lead by fusion in a crucible, or by scorification."
198 THE PBECIOUS METALS

(3) Cupellation.
(4) Inquartation and parting.
(5) Weighing the gold.

Sampling the ore from a mine is partly carried out

(1)

underground by taking pieces from different parts of

the lode. In order to sample a heap of uncrushed ore it
is all broken down to a moderate size, thoroughly mixed,
and a portion set on one side either by machinery or by
"
coning and quartering," or by dividing the heap into two
halves repeatedly by taking alternate shovelfuls. In coning
and quartering the ore is heaped up into a cone, flattened
down, and divided by two vertical planes at right angles to
each other. This divides the heap into four right-angled
sectors or quarters,and the opposite quarters are removed,
mixed together, and the operation repeated as often as
necessary. Sampling machines either take a definite
portion, say one-sixteenth, of a falling or sliding stream of
ore, continuously, or take the whole stream of ore momen-
tarily at intervals, say during one second in each sixteen
seconds. The latter is considered most satisfactory,
Whenthe sample has been taken, it is crushed finer and

again sampled, and finally a weighed quantity is crushed

with pestle and mortar or otherwise to very fine powder. It
is usually considered fine enough passes through an
if all

80-mesh sieve, that is, one containing eighty holes to the

linear inch but some ores, such as those containing
;

tellurides, must be crushed through a 100-mesh or even

a 200-mesh sieve. If flakes of gold are caught on the

sieve (" metallics "), they are collected and cupelled with
lead and parted separately.
 THE ASSAY OF GOLD AND SILVER ORES 199

Finally the crushed sample is well mixed and a portion

weighed out, usually An assay-ton
one or two assay-tons.
contains 29*166 grams, or as many milligrams as there are
ounces Troy in a ton of 2,000 Ibs. Then, if the resulting
gold weighs, say, 1*5 milligrams, the result is at once
reported as 1*5 oz. of gold per ton of ore without calcula-
tion. In the assay of very poor materials, such as tailings,
as much as 12 A.T. (assay-tons) are taken, fused in several

portions, and the gold and silver added together before

parting.
Fusion or Scarification. The ore is usually mixed
(2)

with an equal weight of litharge (PbO), or red lead (Pb 3 4 ),

and enough charcoal to reduce from 25 to 30 grams of lead
to the metallic state, and to reduce all iron oxides to

the ferrous condition. The fluxes to be added depend on

the nature of the ore. Carbonate of soda is added to form

a fusible silicate with quartz or sand, and it is also useful

in taking up and removing sulphur when sulphides are
contained in the ore. Borax is needed to form fusible
borates with the oxides of the metals if these are present
in excess and is advantageous in increasing the fluidity of
almost any charge. Sand is added in the comparatively
rare cases in which the ore is deficient in silica. Sand
forms fusible with alumina, lime, oxides of iron,
silicates

etc., and protects the clay pot from corrosion, but even
when silica is added, a little carbonate of soda is added
also, as the double silicates of soda and other bases are
more fusible than silicates free from soda. Iron in the form
of nails or strips is used when sulphides are present. A cover
of common salt is sometimes added. It protects the charge
from the air and diminishes the chance of loss by spurting.
200 THE PEECIOUS METALS

The quantities of the fluxes must be judged from the

appearance of the ore and the experience of the assayer.

A quartzose ore requires about 1J A.T. of soda carbonate

and 5 or 10 grams of borax to 1 A.T. of ore. The amount
of charcoal will be about 1*25
grams. Oxidised ores require
more charcoal, up to 3 grams or even more, and basic ores
less carbonate of soda.

The charge is thoroughly mixed in a basin or on rubber

cloth and charged into a cold clay crucible which must not
be much more than half full. The crucible is put into the
furnace when the temperature is low. If coke is used, all fuel

touching the pot should be in the first instance at a black

heat. The temperature slowly raised, and after twenty
is

minutes the charge should have just reached a very low red
heat. At this time the lead is all reduced, and remains
entangled in the charge in the form of minute liquid beads.
Chemical action now begins between the carbonate of soda
and the silica. Carbonic acid gas is given off and effer-

vescence results. The temperature is gradually raised to a

full red heat, and the lead beads coalesce and sink through
the liquefying charge, which is continually stirred up by the

disengaged gas. After a period of forty or fifty minutes

from the time of charging in, the charge is in a state of
and the lead containing the gold is collected
quiet fusion,
at the bottom. The nails are then withdrawn, all adhering
lead shaken off them into the pot, and the charge poured
into a warm
iron mould, or allowed to solidify in the pot,
which broken when quite cold. The lead is found at the
is

bottom of the pot or mould, and is detached from the slag

by hammering. If the charge is melted too quickly part
of the gold remains in the slag.
 THE ASSAY OF GOLD AND SILVER ORES '201

An older method, is
largely used,
scarification, still

especially for silver ores. Finely-divided granulated lead

is used instead of litharge and charcoal, and no fluxes are

used except a little borax. The mixture of about 0*1 A.T.

1
of ore and ten to twenty times weight of lead is placed
its

in a clay dish (a scorifier) and melted in a hot muffle.

The lead is oxidised by the air, and the silica in the ore is

fluxed by the litharge. The precious metals are taken up

by the lead, which diminishes in amount as oxidation
proceeds until the slag forms a continuous layer over the
lead. The charge is then poured into an iron mould, and
the lead detached for cupellation.
(3) Cupellation. The cupels (see below for description)
are placed in a muffle furnace, which consists essentially of
a clay oven or muffle, constructed so that it can be heated
to bright redness, and can be either closed up or left open
to the air. Care must be taken to prevent the products
of combustion of the fuel from entering the muffle. Gas
muffles are now used wherever possible.
Fig. 33 is from a photograph
of one of the gas muffle

furnaces in use at the Assay Department of the Eoyal

Mint. The furnace is covered with a thick layer of
magnesia lagging. The figure shows the tube at the back,
fittedwith a regulating damper, through which the air

passes from the muffle. The supply of air to the muffle is

through the mouth, which is shown open.
cups made of bone-ash or magnesia, and
Cupels are little

are usually round as in Fig. 34, but are sometimes square.

See Rose's " Metallurgy of Gold," Chap. XIX., for full details as to
1

quantities of lead required for different ores also for composition of

;

charges in pot fusion.

202 THE PRECIOUS METALS

FIG. 33. Assay Furnace.

The bone-ash is made by calcining bones and treating the

crushed ashes with ammonium chloride. It is
ground fine,
slightly moistened with water, and compressed by the blows
of a mallet in a cupel mould, which is also shown in Fig. 34.
 THE ASSAY OF GOLD AND SILVER ORES 203

The mould is made of gun-metal or steel. The cupels are

dried very slowly to prevent cracking, and are heated for
fifteen minutes or more in the furnace before the assay-
pieces are charged in. Sometimes several cupels are made
in one piece in a large mould, as at the Royal Mint.
When the cupels are at a good red heat (about 700), the
lead buttons are placed in them by means of tongs and the
door of the muffle closed while the lead melts. Air is then
admitted, and the temperature gradually raised. The lead
bath in the cupel becomes smaller and smaller as litharge
is formed and absorbed until all the
lead has disappeared, and the button of

gold and silver remains dull and motion-

less on the cupel. It is then withdrawn
and the button or prill is detached,
cleaned, and weighed.
The losses of gold and silver in cupel-
lation are mainly due FlG 34. Cupel
to -

absorption
and Mould,
by the cupel. They amount to about
1 per cent, of the gold and 10 per cent, of the silver in gold

moderate value of
ores of containing small quantities
silver, if the lead button is free from impurities when
charged in. The presence of copper, bismuth, thallium,
and especially of tellurium and selenium in the lead increases
the losses of precious metal. 1 Care must be taken in the

pot fusion to prevent these metals from entering the lead,

which can best be done by leaving a considerable pro-
portion of the lead unreduced in the slag. If the lead is hard
or brittle in spite of this, it should be scorified with fresh
lead before being cupelled. It is also advisable to fuse the

1
Jour. ('Item. Met. and Mining Soc. of S. Africa, Jan., 1905.
204 THE PRECIOUS METALS

cupel with litharge, charcoal, and fluxes, including some

fluor-spar, and to cupel the resulting lead. Lead is more
often contaminated by copper than by other metals.

(4) Inquartation and parting. The bead of silver and

gold is weighed, flattened, and dropped into nitric acid of

specific gravity 1*16heated nearly to boiling. The silver is

dissolved, leaving the gold as a black scale of the same size
and shape as the original flattened bead. If there is too high
a proportion of gold, some silver is retained by it. The
most suitable proportions of silver for different weights of

gold are as follows :

Weight of Gold. Proportion of Silver to Gold.

Less than O'l milligram. 20 or 30 to 1

About 0'2 10 to 1
I'O 6tol
10-0 4tol
More than 200*0 2J to 1

If the silver is deficient, the prill is cupelled with more

silver after having been weighed.
This operation is called
inqtiartation, the word coming from the old view that the

gold should be a quarter of the whole. Prills containing

lessthan one milligram of gold are parted in a few seconds,

but are generally heated further for a minute or two.
With large quantities of gold and silver, a second boiling
in fresh acid may be necessary. The parting vessel is
usually either a glass test tube or a glazed porcelain

crucible.

(5) The parted gold is washed by decanta-

Weighing.
and annealed by being raised to a red heat. It
tion, dried,
then assumes the ordinary yellow colour of gold, shrinks,
 THE ASSAY OF GOLD AND SILVER ORES 205

and becomes harder. Ifc is

placed in the pan of a delicate
assay balance by the point of a pen-knife or other instru-
ment and weighed. The silver is determined by the
difference between this weight and that of the gold and

silver together. The balance should turn with ^ J^- milli-

gram or less. If 1 A.T. of ore has been taken for assay,

the unit of weight will then be equivalent to ^J^ or

2*4 grains per ton. If 12 A.T. are taken, as in the case of

the poorest tailings, the unit of weight will be 0'2 grain

per ton. This is sufficiently accurate for almost all cases.

 CHAPTEK XIV
THE ASSAY OF GOLD AND SILVER BULLION

UNREFINED gold bullion contains some silver and variable

quantities of base metals, and it is usually necessary to
determine the proportion of both gold and silver. In the
case of refined gold bullion, the gold alone is determined,
silver, if
any, being disregarded. Gold is always deter-
mined by the "parting assay," which consists in cupella-
tion with from two to three parts of silver, followed by the
removal of silver by boiling in nitric acid and the weighing
of the gold residue. The silver in refined silver bullion is

determined in one of four ways by cupellation, by conver-

sion into chloride, which is then weighed (India Mint
method), by titration with common salt (Gay-Lussac
method), or by titration with thiocyanate (Volhard method).
Silver in gold bullion is usually determined by difference,
the gold and silver being estimated together, and the gold
determined separately in the same or another estimation.
Small quantities of gold in silver bullion are determined by

parting in nitric acid and subsequent collection and weigh-

ing of the finely-divided gold.

THE PARTING ASSAY OF GOLD IN BULLION.

A clean piece is cut off the ingot, or a small quantity is

dipped out of the molten metal. The sample is then

 THE ASSAY OF GOLD AND SILVER BULLION 207

flattened by hammering, and a portion of it, usually

amounting to half a gram in weight, is weighed out on a
delicate balance, the weight of the assay piece being adjusted

by cutting with shears and filing. The

then wrapped
piece is

in lead foil together with about a gram of silver and a few

milligrams of copper, if none is contained in the gold bullion.

Four or five grams of lead are enough for bullion contain-
ing no more than 10 per cent, of base metals, the object of
the cupellation being less to remove the whole of the base
metals than to provide a clean and perfectly mixed button
and silver.
or prill of gold
The alloy for parting should not contain less than two
nor more than three parts of silver, copper, etc., to one
part of gold. After a long series of experiments at the

Koyal Mint, it has been found that the most exact results
are obtained by using a
ratio of silver to gold of about

2J to 1. If contained
silver is in the gold bullion, it is
allowed for in calculating the quantity of silver to be added.
Thus, for example, bullion containing 315 parts of gold and
685 of silver per 1,000 would not require the addition of
any more silver. An alloy containing 500 parts of gold
and 500 of silver would require the addition of about 600
parts of pure silver. Base metals such as copper can
replace part of the silver in the parting alloy, but in prac-
tice most of the copper is removed during cupellation. In
the assay of bullion of unknown
composition, the per-

centage of gold is judged by the operator, who is guided by

the colour and hardness of the alloy and by his experience
of bullion from the same source. In doubtful cases a pre-
liminary assay may be made with twice the weight of
silver. Approximately accurate results are obtained with
208 THE PRECIOUS METALS

any amount of silver between two and three times the

weight of the gold.

The assay piece of gold, silver, copper, and lead is charged

into a muffle furnace in which cupels have previously been
raised to a full red heat. The mixture at once melts on
the cupel and begins to be oxidised, fumes of litharge rising

slowly in the furnace. The molten metal is gradually

reduced in size by the oxidation and removal of the lead,
and towards the end of the operation bright spots of
litharge are seen to be formed and to move to the edge of
the lead and disappear. At last no more such spots are
formed and the metal becomes dull and motionless. A
few minutes later the cupel is withdrawn by tongs and the
bead allowed to solidify, when the setting free of the latent
heat causes it to brighten or " flash." If no copper had
been added, the disengagement of oxygen at the moment of
"
solidification would cause the button to spurt or spit,"
and the assay would be spoilt by the loss of some metal.
Spitting is more likely to occur the higher the percentage
of silver in the button. When the cupel is cool the prill is

removed, cleaned from bone-ash by brushing, and flattened

by hammering. It is then annealed by heating to redness,
rolled out into a strip about two inches long by a passage
through a pair and again annealed. The
of jewellers' rolls,

strip is coiled
up by the fingers and dropped into hot nitric
acid of S.G. 1'16 to S.G. 1'26.

If large numbers of assays are made together, each coiled

strip or cornet is placed in a little platinum cup, contained
in a platinum tray whichis plunged bodily into nitric acid

for parting.At the Royal Mint the platinum trays contain

144 cups, and 80 oz. (about 2*25 litres) of nitric acid of
 THE ASSAY OF GOLD AND SILVER BULLION 209

S.G. 1'20 are used. For a smaller number of assays a

larger amount each assay piece is required, a

of acid for

single assay requiring about 1^ oz. Assay pieces contain-

ing more than about three parts of silver to one of gold
generally break up in the acid and must therefore be boiled
"
separately in small glass vessels parting flasks." When
the temperature of the acid reaches about 95 the silver is
attacked and brown nitrous fumes are disengaged. After
about half an hour, the brown fumes having disappeared
and the acid boiling freely, the assays are removed, washed
in hot distilled water, and again boiled for half an hour in
nitric acid free from silver. At the Eoyal Mint 80 oz. of
acid of S.G. 1*20 are used for 144 assays in the second

boiling. At the end of this time most of the silver has

been dissolved, and the cornets consist of gold about 999
parts, silver about one part. They are then washed, dried,
and heated to redness.

The cornets are weighed on a delicate balance and their

exact weight recorded. It is to be observed that gold is

lost during cupellation, chiefly by being carried into the

cupel. The amount varies with the proportion of copper,
being about 0'3 or 0'4 per 1,000 when the copper in the
gold does not exceed 10 per cent., and more if large quan-
tities of copper are present. On the other hand, about
1 per 1,000 of silver is retained by the gold, which could be

reduced to about 0*5 per 1,000 by a third boiling. Conse-

quently necessary to subject pieces of fine gold (called
it is
" ""
proofs checks ") to the various operations side by
or
side and under identical conditions with the assay pieces,
" "
and thus to determine the surcharge or net sum of the
losses and gains incurred in the various operations. The
P.M. p
210 THE PRECIOUS METALS

surcharge usually positive, a proof weighing 1,000 at the

is

beginning attaining a weight of about 1,000*5 at the end.

The accuracy of the results depends on treating all the

assay pieces alike, and this can be judged by the agree-

ment between two or more proofs. At the Koyal Mint a
batch of 72 assays, including six proofs, is charged into
the furnace simultaneously by means of a charging tray
and withdrawn simultaneously. They are also annealed
and boiled in acid together. A high degree of accuracy is
attained and results can be obtained correct to within 0*05

per 1,000.

PARTING ASSAY OF GOLD AND SILVER IN BULLION.

A small quantity of gold, such as 10 per cent, or less,

does not materially interfere with the accuracy of the assay
of silver bullion described in the sequel. The gold is deter-
mined by parting, the alloy being boiled in nitric acid in
a parting flask. The gold residue breaks up into a number
of fine particles, and when the silver is completely dissolved

the gold is allowed to settle, and is washed twice by decan-

tation with hot water. It is then transferred to an unglazed
porcelain cup in the following manner. The flask is filled
to the brim with water, the cup is inverted over its mouth,
and and cup are then turned over together. A little
flask

water flows out into the cup and then forms a seal, and the
particles of gold fall into the cup. The flask is best with-
drawn after immersing the whole in a vessel of water, when
the gold particles are not disturbed. The gold is then dried
and annealed in the cup and is ready for weighing. It is
in the form of a coherent cake, which can be removed from
the cup to the balance pan by means of a pair of forceps.
 THE ASSAY OF GOLD AND SILVER BULLION 211

The presence of a small proportion of silver in gold bullion

isa source of greater difficulty, although the determination
of the gold is not interfered with. The silver is usually
determined by difference, the gold and silver being weighed
together after cupellation and the gold estimated in another
portion. An alternative method is to melt the assay piece
with two and a half times its weight of cadmium under a
cover of potassium cyanide in a porcelain cup, dissolve the
cyanide in water, and part the alloy in nitric acid. The
cadmium and silver are dissolved and the silver is precipi-

tated by hydrochloric acid and weighed as chloride. The

gold residue is also weighed. Proofs of similar composition
are required if extreme accuracy in the results is desired.

THE CUPELLATION ASSAY OF SILVER BULLION.

In this process the base metals are removed by cupella-
tion and the silver cleaned and weighed.
prill is The
weight of silver bullion taken is generally about 10 grains.
The amount of lead required varies with the composition of
the silver bullion, the minimum quantity being about six
times the weight of the silver. This is suitable for silver
not less than 925 fine (925 parts per 1,000). More lead is

required for silver of lower fineness. D'Arcet recommended

ten times the weight for silver 800 fine and twelve times
for silver 700 fine. The cupellation furnace is kept at a
lower temperature than is required for gold bullion assay-

ing, but must be raised to above the melting point of silver

J
(962 ) at the end, as otherwise the metal may set in the
furnace before all the lead has been removed. When the

cupellation is complete the door of the furnace is closed.

p 2
212 THE PEECIOUS METALS

All chinks are stopped by means of bone-ash, and the furnace

is allowed to cool down slowly and evenly. If the molten

beads were removed before solidification, spitting would

result, but as soon as they have solidified they are with-
drawn from the furnace, cleaned, and weighed.
The absorption of silver by the cupel is large, amount-
ing to about 10 per 1,000, and varies considerably, being
increased by a higher temperature of the furnace and by
the presence of a larger amount of copper or of lead. It is
therefore necessary to use several proofs in different parts
of the muffle, and to make them up of the same composition

as the assay pieces.

process is very ancient and was described by

This
Agricola in 1556. It has been largely displaced by the
wet processes introduced in the nineteenth century, but is
still in wide use. It is accurate to about 0*5 per 1000
when nearly pure silver is being assayed.

THE GAY-LUSSAC PROCESS OF SILVER BULLION ASSAYING.

This is a volumetric process in which the volume is

measured of a standard solution of common salt (NaCl) or

of sodium bromide required for the precipitation of a known

amount of silver nitrate. No indicator is used, and the end
of the operation is judged from the appearance of an
exceedingly faint cloud produced by sodium chloride in a
solution from which almost all the silver has been precipi-
tated. It is the most accurate method of silver bullion

assaying at present known, capable, with proper precautions

and the expenditure of sufficient time, of giving results
accurate to about one part in 100,000. As ordinarily
carried out it is accurate to one part in 10,000.
 THE ASSAY OF GOLD AND SILVER BULLION 213

The usual method is to make up a solution of common

salt of such a strength (5 '416 grams NaCl per litre) that
100 c.c. of it will precipitate one gram of silver. This is

/ \
100
c.c.

\ /

FIG. 36. Hand Pipette

for Decinormal Solu-
tion, Gay-Lussac
Process.

FIG. 35. Stas Pipette.

"
called the normal solution." An amount of silver bullion

containing about 1*003 grams of silver is then weighed

accurately, transferred to an 8 oz. glass bottle, and dissolved
in nitric acid on a hot plate. The composition of the
214 THE PKECIOUS METALS

bullion must be known approximately before the weight to

be taken can be determined. If the composition is, say,
900 fine, then the weight to be taken is *$$ grams, or

PIG. 37. Shaking Machine used in Gay-Lussac Process of Silver

Bullion Assay.

1*115 grams. The normal solution is then run into the

bottle from a pipette containing 100 c.c. (see Fig. 35), and
the contents are shaken until the silver chloride has curdled
and will settle to the bottom, leaving the supernatant liquid
 THE ASSAY OF GOLD AND SILVER BULLION 215

clear. One shaking machines in use in the assay

of the

laboratory of the Koyal Mint is shown in Fig. 37. It holds

twenty bottles. After the liquid has been shaken it is tested
by the addition down the side of the bottle of 1 c.c. of a
decinormal solution of salt, from a hand pipette shown
in Fig. 36. This
equivalent to the addition of one-
is

thousandth (a millieme) of the 100 c.c. of solution originally

added. If a cloud is formed near the surface of the liquid
the amount of salt solution required in the judgment of the

operator, based on the density of the cloud, to precipitate

the rest of the silver is added, and the bottle is shaken

again and tested once more. The end reaction is marked
by the gradual formation in the course of about ten minutes
of a faint cloud, the value of which in tenths of a millieme
is judged by the assayer. For example, the first addition
of salt precipitated I'OOO gram of silver, the second addition

of say 4 c.c. in all of decimal solution precipitated 0'004 gram

silver ;
then suppose that the cloud produced by the third
addition was declared to be equivalent to five ten-thousandths
or 0*0005 gram silver. It follows that the total amount of
silver present in the assay piece must have been r0045
gram, whence its composition can be calculated.
Mercury interferes with the method, as
precipitatedit is

with the silver, but itsdetected by its prevention

presence is

of the darkening of silver chloride when exposed to light.

The on the assay result caused by a small proportion

effects

of mercury are removed by the addition of acetate of soda.

necessary to use check-assays in order to
It is, of course,

test the strength of the solution. The method was invented

by Gay-Lussac in 1832, and has displaced the cupellation
process in almost all mints and in many private assay offices.
216 THE PEECIOUS METALS

THE VOLHAKD METHOD OF SILVER BULLION ASSAYING.

This method was proposed by J. Volhard in 1878, and is

particularly suitable for the assay of silver plate at the hall-

marking offices. In this work the exact composition of
the silver bullion is not required, but tests are made to

ascertain that the proportion of silver in the bullion is not

less than the standard prescribed by law. In England
there are two standards containing 925 parts per 1,000 and
958*3 parts per 1,000 respectively.
1
At the Sheffield Assay Office, where the method is in

use, a quantity of bullion is taken, such that about one

gram of silver is present. It is dissolved in 10 c.c. of dilute
nitric acid (S.G. 1*2) in a bottle of about 250 c.c. capacity
and heated in a water bath until all traces of nitrous acid

are expelled. The solution is then diluted with 50 c.c. of

water containing 2 c.c. of a saturated solution of iron alum

" "
decolorised by nitric acid, and 100 c.c. of normal
ammonium thiocyanate is added from a pipette. This
solution contains about 7'04 grams of NEUCNS per litre, and
100 c.c. is sufficient to precipitate about one gram of silver.

The bottle is then well shaken for about two minutes, when
the white precipitate of silver thiocyanate settles to the
bottom and leaves the liquid clear. If some silver is still
in solution, the liquid retains the pale green colour due to

the dissolved copper originally contained in the silver

bullion. however, the ammonium thiocyanate is in
If,

excess, the liquid is coloured red from the action on it of

the ferric salt. Proofs are used consisting of the same

by Volhard's Method." E. A. Smith.

1
"Assay of Silver Bullion
Trans. Inst. of Mining and Metallurgy, Vol. XVI. (1907), p. 154.
 THE ASSAY OF GOLD AND SILVER BULLION 217

weight of a standard silver trial plate as is taken of the

bullion to he tested. The strength of the solution of thio-
cyanate adjusted so that the liquid becomes of a pale red
is

colour in the proof bottle. Then, if the other bottles

contain green or still paler red liquids, the silver in the
bullion tested higher standard than the trial plate,
is of

and the wares are passed. If the liquid in an ordinary

assay is of a deeper red than that in the trial plate bottle,
the bullion is below standard and the wares are condemned.

By making use of a "decinormal" solution of thio-

cyanate, and adding by it little the

little, the colour in all

bottles can be made the same, and the exact composition of

the bullion ascertained correctly to about 0*2 per 1,000, or
even more closely.

THE INDIA MINT METHOD OF SILVER BULLION

ASSAY.

This method was introduced by Dr. Dodd, of the Calcutta

Mint, in 1851. It consists in adding an excess of hydro-
chloric acid to a solution of nitrate of silver and weighing
the silver chloride produced. The weight of silver bullion

taken is 18'817 grains, which is equivalent to 25 grains

of silver chloride if the bullion consists of pure silver.

After precipitation, the bottle is well shaken, and the

chloride is allowed to settle and is washed by decantation.
The chloride of silver is transferred to a Wedgwood cup
by placing the latter over the mouth of the bottle and

inverting it. Afterwards the excess water is poured off,

the chloride broken up by a glass rod, and dried. The

drying is finished on a hot plate, and the cake of chloride
218 THE PKECIOUS METALS

isweighed while still warm. A set of weights is used in

which the ''1,000" weighs 25 grains, and the smaller
weights are in proportion, so that the composition of the
alloy is indicated without calculation.
In this method the gold remains unchanged and is

weighed with the chloride of silver.

 CHAPTER XV
MINTING

MINTING is the manufacture of coins from gold, silver,

and other metals under the authority of the State. In
early times the exchange of commodities was carried on by
means of barter, but the difficulties met with under this

system caused attempts to be made to establish a medium

of exchange by which the values of all commodities might

be expressed in terms of one of them. In pastoral commu-

nities cattle and skins were used for this medium, and the
origin of theLatin word pecunia (money) may be traced in
this way from pecus (cattle). There are certain necessary
properties which money must possess, such as high
intrinsic value, indestructibility, divisibility, etc., and these
appertain to metals in a far greater degree than to other
commodities. Accordingly pieces of metal were used for
money in Egypt an early period, and subsequently, about
at

700 B.C., the manufacture of coins was begun in the island

of ^Egina and in Lydia. The art of minting spread among
the Greeks and was carried by them to Italy and other
countries around the Mediterranean, and also into Asia.

Coining of silver began in Rome in 269 B.C. and was intro-

duced by the Romans throughout a large part of Europe.
The earliest coins were made by casting in spherical or
conical moulds. The cast metal was then placed on a little

anvil and marked with a bronze die by means of heavy

220 THE PRECIOUS METALS

blows of a hammer. The lower side or reverse of the coin

was marked by the sharp edges of the anvil (ambos),

also
which was smaller than the coin. Subsequently the anvil
was engraved with various devices and was ultimately
"
replaced by a reverse die. The " blank was made red
hot and struck between cold dies which were held in

position by tongs or by a roll of lead to protect the hand

of the workman. One of the great improvements intro-
duced by the Romans was the use of iron dies, which were
sufficiently hard and durable to allow of the coins being
struck while cold. Later, strip-shaped bars of metal were
cast, and square pieces cut from them and hammered
round. The discs prepared in this way were then struck
between dies, and their edges sometimes trimmed with
shears. The smaller coins, such as the silver pennies
made in England in the Middle Ages, were made from sheet
metal reduced in thickness by hammering and were cut
" "
approximately round by shears. Such hammered money
continued to be manufactured in England until 1662 and
was in circulation until 1696.

In 1662 the "mill and screw" were finally established in

London, after having been tried for a short time in 1561.
" "
The millwas a rolling mill used to reduce the thickness
" "
of the cast bars, and the screw was a screw-press used
forapplying the power in striking the coins. A collar to
surround the coin whilst it was being struck was intro-
duced at the same time. The inside of the collar was
sometimes smooth and sometimes marked with corruga-
tions or inscriptions. The money was known as " milled
money," and the chief noticeable difference between it and
"
hammered money " was that its size was invariable and
 MINTING 221

" "
its edge marked. Hence arose the phrases milled edge
" on the edge,
and milling," referring to the graining
although the rolling mill had nothing to do with the marks
on the edge. As had been hoped, the practice of clipping
money fell into disuse on the introduction of the milled

edge. An even greater improvement than the new collar,

introduced at thesame time, was the cutting machine by
which the blanks were punched out from the flattened bars.
All these machines had been originally introduced in
Paris in 1553, and they were finally set up at the French
Mint in 1640. The power used to drive the machinery was
supplied by workmen at first, but their place was soon
taken by horses, which were finally displaced by steam in
Birmingham in 1788, in London in 1810, and in Phila-
delphia in 1835. Numerous minor improvements in mint-
ing were made in the nineteenth century, chiefly having
for their object the saving of time in manufacture rather
than the production of better-executed coins. Among the
chief improvements were the introduction of the lever press
for striking in 1839, and the invention of automatic
balances for weighing single coins in 1843.
The manufacture from gold and silver bullion
of coin

involves a number operations which may be ranged

of

under the following heads :

1. Valuation of the bullion received at the Mint.

2. Melting the metal and casting it into bars.
"
3. Boiling out the bars into strips or fillets."
4. Cutting discs or blanks from the strips.
5. Adjusting the weight of each blank.
6. Edge-rolling the blanks.
222 THE PEECIOUS METALS

7. Annealing and "blanching" the blanks.

8. Striking the coins.
9. Weighing each coin.
" "
10. Telling and packing the coin.

VALUATION OF BULLION.

Gold may be sent in to the London Mint for coinage by any

one who has some in his possession, the only restrictions
being that the average fineness must not be less than
916-6 per 1,000, the standard fineness of sovereigns and

half-sovereigns, and that the gold must be suitable for

conversion into coin without further refining. There is
no refinery in the Mint, and unrefined gold is not received.
The gold is converted into coin at the rate of ^3 17s. W^d.
per standard ounce troy without any charge or deduction
whatever. The price of standard gold in the London
market varies from B 17s. 9d. to about d63 18s., but
it is never profitable for ordinary owners of gold bullion
to send it to the Mint, because the conversion into coin

usually takes some weeks and the interest on the value of

the gold amounts to over 2rf. per oz. per month. The
result is that the Bank of England alone sends gold into
the Mint. The metal is usually in the form of refined
ingots, eachweighing about 400 oz. and containing over
99 per cent, of gold. The ingots are weighed in the
presence of a representative of the Bank of England on
scales turning easily to 0*005 oz. A piece is then cut
from the ingot and is assayed as described in Chapter XIV.

The number of standard ounces of gold in each ingot is

calculated from the weight and the assay, and a notification
 MINTING 223

of the estimated value of each consignment sent to the

Bank. Unless an objection is made, the ingot is then
passed for melting. The number of ingots usually sent
to the Mint at one time is 200, of the value of about

340,000, and four times this amount is sometimes received

in a week.
At the three Australian mints, which are situated at
Sydney, Melbourne, and Perth, unrefined gold is received
for treatment. weighed and assayed (usually after
It is

being melted), and the value is then calculated. This is

paid to the owner of the gold a short time later, less a

deduction made to cover the cost of refining and coining.
The charges vary somewhat at the different mints, and
are higher for small consignments and for bullion con-

taining considerable quantities of base metals. The actual

amount charges at the Perth Mint in 1907 was equal
of the
1
to 2*65d. per oz.of the gross weight of metal sent in.

But inasmuch as only a part of the silver contained in

the bullion paid for, the real total of the deductions
is

was equal to about 4'7Sd. per standard ounce of gold

sent in.
In the United States unrefined gold is also received at the
mints and Government assay offices. It is valued, and a
charge for treatment made to cover the cost of refining, etc.,
including the waste of metal. The charges usually amount
to about 6 or 8 cents per ounce of gold.

Silver bullion is no longer received for free coinage at

any mint. The usual practice at mints at the present day,
with currency on a gold basis, is to purchase silver bullion
with Government funds and to coin it into pieces required
"
Thirty -eighth Annual Report of the Mint,"
1
p. 137.
224 THE PEECIOUS METALS

for circulation at rates giving a large percentage of profit.

The silver coins then circulate at values which are far

greater than that of the metallic silver contained in them,

and can only be kept from depreciation by strictly limiting
the supply and by withdrawing from circulation any un-
" "
necessary coins at their full face value at the expense
of the State. In England one ounce of standard silver

is coined into 5s. 6d. in silver money, and similar rates

are in force in many other countries.
Eefined silver bullion is usually sold in the form of ingots
which weigh from 1,000 to 1,200 oz., and often contain
over 99*9 per cent, of silver.

MANUFACTURE OF COIN.

The bullion is melted in plumbago crucibles, together

with the copper required to produce bars of the correct
composition. In England standard gold consists of 916*6
parts of gold and 83*8 parts of alloy metal, and standard
silver contains 925 parts of silver and 75 parts of alloy.

Abroad the 900 standard is generally used, with some

exceptions. The crucible was formerly heated in coke
furnaces, but latterly the use of gas or oil fuel has been
spreading rapidly.
One of the disadvantages of coke for fuel is that the fire

has to be made up afresh for each charge. At the Royal

Mint, London, where coke is still in use, when the charge is
ready to be poured, some of the firebars are taken out and all
the coke is raked down into the ashpit. The crucible is then
lifted out, and as soon as possible the same or another crucible
is put in its place and the furnace is again filled with coke.
 MINTING 225

This involves a great loss of heat, and restoking with cold

fuel is also necessary in some
cases before the melting of
the charge can be completed, so that further loss of heat
takes place. The result is that the work of the furnace is
delayed. With oil or gas for fuel the heating is continuous
and more charges can be melted in a day.

The greatest disadvantage in the use of solid fuel, how-

ever, lies in the production of ashes which become enriched
with particles of the precious metals. The mint " sweep"
consists of these, together with the crucibles and floor

sweepings. They are ground up (sometimes with the

addition of mercury) in edge-runner mills resembling
mortar-mills, and washed for the recovery of the flattened
pellets of gold and silver or of the amalgam. The residues,
which still have some value, are sold to smelters, a course
which involves some loss to the mints. The greater
" "
part of the sweep results from the ashes of the fuel, and
when oil or gas is used, the inconvenience due to this

portion is avoided. A blast of air is generally used with oil

or gas, and a flame produced like that from a blowpipe.
Care must be taken to prevent the flame from impinging
directly on the crucibles or they will be rapidly burnt
through. Checker-work of brick is sometimes used to
receive the direct flame, which afterwards spreads out
round the crucible.
At the London Mint gold is melted in crucibles about
10 J inches in height, containing 1,200 or 1,300 oz. of
metal, and silver in larger crucibles containing over 5,000 oz.
In some foreign mints larger crucibles are used, holding
500, 600, and even 1,000 kilograms (32,000 oz.). When
such large pots are used, the molten metal is ladled out for
P.M. Q
226 THE PRECIOUS METALS

casting, but in the Royal Mint the crucibles are lifted out of
the furnace by basket tongs, and their contents, after being
 MINTING 227

well stirred, are poured into moulds by tilting the crucible.

The crucibles containing silver, bronze or nickel are lifted

by a travelling electric crane and placed in a tilting cradle,

shown A pot is shown in Fig. 38,
at the front in Fig. 38.
l

held in the basket tongs and suspended from the crane.

The iron moulds are shown on their wheeled truck close to
the cradle in the same figure. The dimensions of the
moulds vary somewhat, sovereign bars, for example, being 22
inches long, If inches wide, and j- inch thick. The cradle is
tilted by a hand crank working through geared wheels.
After assay, the cast bars, having been cleaned, their ends

cropped and the rough edges filed off, are reduced in thick-
"
ness by rolling. Heavy pinches between the breaking
"
down rolls, which are 15 inches in diameter, are applied at
first, and subsequently lighter pinches are given in the
"
and " finishing
" "
smaller thinning rolls. The bars
increase in length as their thickness is diminished, but their
width is hardly affected.

In Fig. 39 the breaking-down rolls in the silver and gold

rolling room at the Royal Mint are shown. The illustration
is from a photograph taken by Mr. S. W. Smith. The rolls

are driven by an electric motor placed under the floor and

connected by chain gearing. The apparatus for adjusting the
space between the rolls is shown at the top of the framework.
A truckload of bars is in front. The distance between the
faces of the rolls is reduced after each passage of the bars

through them. A number of spare rolls is shown on the wall

of the room, but the breakage of a roll is of rare occurrence.

After repeated passages through the rolls, the silver bars

1
Figs. 38, 41, 43, 44, 45, and 46 are from photographs taken by
Mr. J. E. Barnard.

Q 2
228 THE PEECIOUS METALS

become hard and are usually annealed by subjection to

a low red heat in gas furnaces before the rolling can be
completed. At the Vienna Mint the silver bars are

\ /?
>sf

PIG. 39. Eolling Eoom, Eoyal Mint.

annealed several times, in London only once, and in the

United States, since the introduction of the electrolytic
refining process, the improvement in the purity of the
metal has enabled annealing to be dispensed with.
" ''
When the rolled-out bars or fillets are of the correct
 MINTING 229

thickness, as evidenced by the use of a gauge and the weight

of a trial blank, they are passed through the cutting
230 THE PRECIOUS METALS

machines, by which one, two or more blanks are punched

out at each revolution.
 MINTING 231

The cutting room shown in Fig. 40,

at the Royal Mint is

which is from a photograph taken and kindly lent by

Mr. Edward Rigg, I.S.O., the Superintendent of the
Operative Department.
Some of the cutting machines are shown at closer

quarters in Fig. 41. The punches or cutters descend

owing to the revolution of an eccentric, and the pieces
of metal are cut from the strip between the edge of the
cutter and that of a hole in a steel plate which the cutter
fits accurately.
" "
The scissel or residual metal (see Fig. 42) is returned
to the melting house.
In Pig. 42 the
arrangement used
when two rows of

blanks are cut from a

fillet is shown. This Yio. 4 2. Scissel.

is the usual number,

but only a single row of large coins, such as dollars or
crowns, is cut from a fillet, whilst three rows of pennies
and five of farthings are cut from each strip.
" "
The blanks
are passed through an upsetting machine
(called the marking machine) for the purpose of rolling up
and thickening their edges, and are afterwards again

annealed. In the case of gold and bronze, furnaces are

used from which air is carefully excluded by a water-seal
so as to prevent the blanks from being discoloured by
oxidation. Silver blanks, however, are heated with free
access of air, and their surfaces become blackened owing to

the formation of oxide of copper. The rotary gas furnaces

in use for this purpose are shown in Fig. 43. A bag of
232 THE PKECIOUS METALS

blanks is emptied into the mouth of the furnace, and the

pieces are conveyed to the other end of the furnace by means
of an internal Archimedean screw.
They fall through an

FIG. 43. Annealing Furnaces for Silver Blanks, Royal Mint.

aperture into a copper tray placed to receive them, having

been raised to a dull red heat during their passage through
the hot zone.
The blackened silver blanks are pickled or " blanched"
by dipping in hot dilute sulphuric acid. Any oxide of
 MINTING 233

copper formed on the surface during annealing is removed

in this way and a film of pure frosted silver remains,

coating the blank. In many mints the blanks are weighed

before being struck, and those of incorrect weight are
eliminated. Light blanks are returned for melting, and
heavy ones are adjusted by filing by hand or in a machine.
In the Royal Mint the process of adjusting blanks was
abandoned in 1869, and individual pieces are not weighed
until they have been struck.
The modern coining press was invented by Uhlhorn in
Germany in 1839, and altered or independently devised by
Thonnelier in Paris immediately afterwards. The blank
isplaced by the machine on the lower die, and is surrounded
by a collar which is either plain or crenated, according to
the denomination of the coin to be struck.The upper die
is then brought down on the blank by the straightening
of a toggle-joint, which is effected by the movement of
a bent lever. A single blow is sufficient for the manu-
facture of each designs on both sides being
coin, the
made simultaneously, while the metal is forced laterally
by the enormous pressure into the corrugations of
the collar.

The press-room Royal Mint is shown in Fig. 44.

at the

There are 19 presses, and each of them can strike from

90 to 125 coins per minute. The force of the blow
varies with the coin to be struck. The pressure required
by a hydraulic press to strike an American dollar
(T529 in. diam.) at the Philadelphia Mint
was found
to be 160 tons, and in the case of a rupee (1*205 in.
diam.) it was found at the Bombay Mint to be 135*7
tons. The pressures exerted in the screw coming press
234 THE PBECIOUS METALS

were found by Capt. Bourke to be 178*84 tons for the dollar

and 145'56 tons for the rupee. The dollar was raised in
 MINTING 235

temperature about 20 C., and the rupee about 25 by the

1
blow.
After being struck, each individual coin is weighed
separately to determine whether it is within the legal
limits of weight. It is not possible to make all coins

same weight, and the maximum difference

of exactly the

from the standard weight allowed by law is called the

"remedy." It amounts to 0'2 grain above or below the
standard weight of 123*274 grains in the case of the English

sovereign, or about 1*62 parts per 1,000. The remedies on

silver coins arid on foreign coins are proportionally greater.
The coins are weighed on automatic balances which were
invented in 1843 and have undergone some changes. The
modified Cotton-Pilcher balance used at the Eoyal Mint
consists of a steel beam with steel knife-edges
supporting
"
two pans," one for the coin, the other for the counter-
poise weight. The beam is allowed to move freely for
about a second under the influence of the weight of the
coin acting against the counterpoise, and it is then seized
by forceps and held firmly in its new position. By an
ingenious arrangement of levers, the position of a shoot
down which the weighed coin falls is determined by the
final position of the beam. According to the weight of the
coin the shoot stops above one of three slots, and heavy,

light, and good coins are discharged through different orifices

and make their way into different compartments of a box.
These automatic balances are shown in Fig. 45. Each
balance weighs about 22 coins per minute, and is sensitive
to O'Ol grain. The proportion of rejected coin was formerly

1
See pamphlet by Capt. J. J. Bourke and W. G. Nichols. Privately
printed.
 THE PKECIOUS METALS

large,but has been greatly reduced of late years. It

sometimes exceeds 10 per cent, in the case of gold coin,

 MINTING 237

and is usually below 1 per cent, in that of silver coin.

This difference is due to the comparatively small remedies

allowed for gold.

The finished coin is examined for superficial defects,
"
dumb pieces are eliminated by testing the ring," and the

FIG. 46. Telling Machine, Eoyal Mint.

" "
good pieces are counted and packed in bags by telling
or counting machines, one of which deals with 3,000 pieces

per minute. This is shown in Fig. 46. The machines con-

tain a serrated wheel on an inclined table, with semicircular
notches on its
periphery. The coins are made to pass down
a narrow shoot in single file and are gripped by a pair of

india-rubber driving wheels which force the coins forward

238 THE PEECIOUS METALS

and so cause the wheel to rotate, but only

passed one coin is

forward by each notch, so that with each revolution of the

wheel a definite number of coins pass down the shoot and
fall into a bag. When the correct number of coins have
passed the wheel and entered the bag, the machine stops
automatically. For example, 1,200 pence are counted into
one bag. The bag is then removed, a new one substituted
for it, and the machine is restarted. There are four such
serrated wheels on the machine shown in the illustration.

COMPOSITION OF GOLD AND SILVER COINS.

The earliest gold alloy used for purposes of currency was

electrum, a naturally occurring silver-gold alloy containing
60 to 80 percent, of gold. In the Hellenic world, however,
the coins of gold and silver were remarkably pure, the gold
coins sometimes containing only 3 per 1,000 of silver. 1
Some of the Greek silver coins contained 30 per 1,000 of
copper, but others were from 983 to 986 fine. The Romans
in Republican times used pure gold and silver. Under the
Empire, copper was intentionally added, and in some reigns
the coinage became very debased, containing only a few parts
of precious metal per 1,000. In England after the Conquest
the gold standard was 994'8 for a long time, but the last
coins of this composition were struck in 1637. The present
standard of 916'6 was first instituted in 1525, the ostensible

object of the lowering of the

standard being to prevent the
exportation of the coin to Flanders. The view that the
copper preserves the coin from rapid wear is of far more

recent date. The silver standard of silver 925 parts, copper

1 " La Monnaie dans
Lenormant, 1'Antiquite," Vol. I., p. 187.
 MINTING 239

75 parts, has been in use since Anglo-Saxon times, except for

a brief period of debasement in the reigns of Henry VIII.
and Edward VI., when the gold coinage was also debased.
The 900 standard for both gold and silver was introduced
in France in 1794 and has been adopted in many other
countries. For subsidiary silver coins, however, lower
standards (especially 835 fine) have been introduced, in the
Latin Union and elsewhere, the alloy in use containing
most copper being that adopted by Scandinavia and Den-
mark for small coins (10 ore pieces), which contain 400
parts of silver and 600 parts of copper.

The following table gives the legal standards now in force

for gold and silver coin in the principal countries of the

world. The standards given are in parts of precious metal

per 1,000, the remainder being copper.

Gold Coin. Silver Coin.

Abyssinia .

Argentine .

Austria-Hungary
Belgium
Bolivia
Brazil
Bulgaria
Canada
Ceylon
Chile .

China .

Colombia .

Congo
( 'orea .

Costa Eica .

Crete .

Curasao
Cyprus
Denmark .
240 THE PRECIOUS METALS

Dominica .... Gold Coin. Silver Coin.

900 and 835

Ecuador ....
Dutch East Indies

....
. .

900
720
900
Egypt
Finland
Prance
....
....
875
900
900
833-3
808 and 750
900 and 835
Germany
Greece ..... . . . 900
900 900
900
and 835
Guatemala
Hayti
Holland
....
....
900
900
900
900
900
945
and
and
and
835
835
640
Honduras
Honduras
Hong Kong
....
(British)

.
.

.
.

.
925
900
800
India 916'6 916-6
Italy
Japan ....
....
900
900
900 and 835
800
Mauritius
Mexico
Morocco
....
.
800
902-7 and 800
900 and 835
Newfoundland
Nicaragua
Norway
....
....
. . . 916-6

900
925
800
800, 600, and 400
Panama
Paraguay
Persia
....
....
. . . . 900

900
900
900
900
Peru
Portugal ....
....
916-6
916-6
900
916'6
Eoumania
Russia
Salvador
....
....
900
900
900
900
900
900
and
and
and
835
500
835
Servia . . ." . 900 900 and 835
Siam 900
Spain
Sweden .... 900
900
900 and 835
800, 600, and 400
900 and 800
Switzerland ...
Straits Settlements

....
. .

900 900 and 835

830
Turkey
United States...
....
916'6
900 900
900
Uruguay
Venezuela . 900 900 and 835
 MINTING 241

Many of the countries enumerated in the table have no

actual gold coins, but have merely made provision for

them.
In addition the standards given above, others are in
to

use for coins used for trade purposes. Thus, for example,
the Austrian gold ducat, the coin of highest standard in the
world now in use, contains 986 J parts of gold and only 13|
of copper. It is not current in Austria. The Maria Theresa
thaler, containing 833*3 parts of silver per 1,000, is also
struck at Vienna, principally for use in Africa and the
Levant.

P.M.
 CHAPTEK XVI
THE MANUFACTURE OF GOLD AND SILVER WARES

COMPOSITION OF GOLD WARES.

IN England the carat system is still employed for denoting

the standard of fineness of gold wares. Under this system,
which was formerly in use for the assay
of gold at the Mint,

the troy pound of 5,760 grains is divided into 24 parts,

called carats. A carat is divided into 4 carat grains,
each of which contains 60 ordinary grains. Gold bullion,
consisting of 22 parts of pure gold mixed with 2 parts of
"
other metal or alloy," is called 22-carat gold. The stan-
dards recognised by law in England are those containing
22, 18, 15, 12 and 9 parts of gold in 24 parts. The usual
composition of wares made of these standards is as follows,
in parts per 1,000:

Standard.
 MANUFACTURE OF GOLD AND SILVER WARES 243

cutting or scraping and assayed for gold. If the fineness

is below standard the wares are broken, but if of full
standard they are hall-marked. Offices for hall-marking
are at London (Goldsmiths' Hall), Birmingham, Chester,
Sheffield, Dublin (where 20-carat gold is marked), Edin-

burgh, and Glasgow. Offices formerly existing at Exeter,

Newcastle, Norwich, and other places have now been closed.
The marks on the wares include a number giving the
standard (22 for 22-carat, etc.), a distinctive mark showing
the town where the marking was done, and a date mark,

usually a single letter of special design.

Hall-marking offices are provided by most of the govern-
ments of the civilised world, the most noteworthy exception
being that of the United States.
The following table gives the legal standards of gold plate
1
in use in some of the principal countries in the world :

Austria-Hungary . .
920, 840, 750, and 580.
Belgium 800 and 750.
Denmark .585
France .... . . . and upwards.
920, 840, 750, and 583.
583 and upwards.
Germany . . .

Italy 900, 750, and 500.

Netherlands . . .
916, 833, 750, and 583.
Portugal 916-6 and 750.
Russia . . .
<:<)(), 920, 820, 720, and 560.
Spain 916-6 and 750.
Sweden and Norway .
975'7, 847'2, and 763*9.
Switzerland 750 and 583.

1 " Return House Commons

to of as to Foreign Countries (Gold and
Silver Marking)," February 25, 1890.

E 2
244 THE PRECIOUS METALS

In England the 22-carat standard is used almost exclu-

sively for wedding rings, and was
formerly used for
mourning rings. A few watch cases are made of this
standard. The best jewellery and almost all watch cases
are made of 18-carat gold, which is manufactured into all

kinds of personal ornaments, as the beauty of its colour is

not much altered by wear. Fifteen -carat gold is also used

" "
extensively for coloured jewellery, and has greater
strength and durability than most other standards. Gee
1
states that 13-carat gold is largely used in Birmingham
for jewellery, as it is the lowest standard that can be readily
coloured to look rich and beautiful. It is not presented
for hall-marking, and very little 12-carat jewellery is

manufactured. The lowest standard, 9-carat, is now very

largely used for cheap jewellery, and Gee states that even
more jewellery is made of slightly inferior material, which
cannot be hall-marked, although it is usually called
9-carat. This standard contains some zinc, but for the

higher standards the base metal is usually copper, which

gives the wares a better and more durable colour.

COMPOSITION OF SILVER WARES.

Silver wares in the United Kingdom are almost always

of the same composition as silver coin, containing 925 parts
of silver and 75 of copper per 1,000 or II oz. 2 dwt. silver

and 18 dwt. copper in the pound troy. Another standard,

Britannia silver, containing 11 oz. 10 dwt. silver in the

pound troy, or 958'3 parts per 1,000, is occasionally used,

while in Austria-Hungary there are four standards, ranging
from 950 to 750 fine.
1 " Goldsmiths' Handbook." Fifth edition, p, 45.
 MANUFACTURE OF GOLD AND SILVER WARES 245

The
following table gives the legal standards for silver
plate in use in some of the principal countries in the
l
world :

Fine Silver per 1,000 parts.

Austria-Hungary. . .
950, 900, 800 and 750.
.900 and 800.
Belgium
Denmark ....
...
. . .

8l><>.

France
Germany .... . 950 and 800.
800 for plate. Any
fineness for jewellery, etc.
degree of

Netherlands. . . . 934 and 833.

Italy 950, 900 and 800.
Norway and Sweden
Portugal .... . . 828-1 and 812-5.
916-6 and 833. Various stan-
dards down to 800 for export.
Russia 910, 880 and 840. For silver
wire, 940 to 960.
Spain 916-6 and 750.
.875
Switzerland
Turkey
2
....
. . .

900.
and 800.

MELTING AND EOLLING GOLD AND SILVER FOR WARES.

For 22-carat gold it is usual to melt down sovereigns.

For other standards refined gold or coin is used. The gold,
silver and copper are melted together in plumbago crucibles,

and if zinc is to be added it is charged in last, after the

other metals are melted, in order to reduce loss. The
charge is covered with charcoal, and borax is frequently
added to remove dross. If it is feared that the metal will
be brittle, sal ammoniac or corrosive sublimate is added to

the charge in the crucible. The deleterious elements are

in that way removed as chlorides and the residue of flux is
" Return to House of Commons as Foreign Countries (Gold and
1
to
Silver Marking)," February 25, 1890.
2
E. A. Smith,
' '
On Silver Alloys of Industrial Importance," p. 4.
246 THE PRECIOUS METALS

sublimed. Nitre and carbonate of soda or potash are also

sometimes used to remove tin, lead, etc. The charge is

well stirred with a graphite rod, which is better than iron,

before being poured. In the United States it is customary
to add cadmium as a deoxidant to sterling silver, the mix-
ture being made up as follows :
l

Fine silver .... 925 parts.

Shot copper
Cadmium ..... . . . . 70 ,,
5 ,,

This mixture gives sound malleable castings.

The metal is cast in warm cast-iron well-oiled moulds,
no attempt being made to fill up the hollow formed by
shrinkage. Charcoal, flux, etc., are kept from entering the
mould by a piece of wood held in the hand. The bar is

cooled by plunging it into water, and it is trimmed, cropped

and cleaned before being rolled. In rolling frequent
annealing is desirable, especially of gold of 18 carats and
under.
After annealing, silver plate is plunged into a pickling
solution of dilute sulphuric acid. This is usually done
while the plate is still red hot from the furnace. Oxide of
is dissolved and the plate is left with a uniform
copper
dead-white surface. Both before and after rolling the
"
plate is carefully overlooked," and imperfections such as
pieces of embedded in the plate are
charcoal, slag, etc.,

removed by scraping, generally by a machine similar to a

2

planer, called an overlooking machine.

For the production of wire, a rolled plate after annealing
1
Jour. Franklin Institute, Vol. CLXIII. (1907), p. 118.
2
Idem, p. 119.
 MANUFACTURE OF GOLD AND SILVER WARES 247

is passed through a pair of slitting rolls, which have a

number of parallel square grooves, the projecting edges on

one roll fitting into the depressions of the other, so form-

ing a number of cutting edges. The plate is cut up into

square strips by this machine, and these, after passing

through a pair of wire-rolls, which make them approxi-

mately round, are annealed, their ends are pointed and

they are dragged through the draw-plate. The draw-plate
contains about ten conical holes of different sizes, and the
wire passed through them in succession until it is of the
is

required thickness. Square, oval, half-round and oblong

wires are also made with the aid of special draw-plates.

Oblong wires are also made by flattening square wire by

rolling. Hollow half-round wires are made from flat wire
by pulling it through half-round holes with the point of a
tapered punch held against the wire. Two oblong gold
steel

wires placed one on each side of a piece of steel wire and

drawn through a plate produce hollow-square or hollow-
round wires. The wires are used for the manufacture of
gold and silver chains and other purposes. The pieces of
wire cut to length are put together in innumerable ways
and soldered by hand.

MANUFACTUKE OF WARES FROM PLATE.

The ancient method of making spoons and forks was to

hammer them out from a rod by hand, finishing them by
sawing and filing. This has now almost entirely given
place to machine processes, in which the metal is stamped
out from plates, and shaped by striking between dies. The

general method of cutting by means of punches and strik-

ing between dies has been applied to the manufacture of
248 THE PRECIOUS METALS

numberless articles. The presses for striking are generally

of the falling-weight type, the weight being raised by steam

or electric power and released by hand, the feeding of the

pieces of plate on to the dies and their removal being also

done by hand. Much work is also done by turning, and by
" "
spinning or burnishing into the desired shape on a lathe

by means of special burnishing tools. The metal must of

course be soft and malleable for spinning.

Decorative work in the case of "wrought" wares is

carried out by hand hammering. Other decorative

methods are engraving and chasing, the raised pattern in
the latter case being applied by means of punches having
their surface in intaglio.

SOLDERING.

The pieces of gold or silver, whether cut out or stamped

or otherwise prepared, are fitted together by hand, usually
by means of soldering. The best solders for " coloured "
gold articles, that is for gold of 13 carats and upwards, are
1
given by Gee as follows :
 MANUFACTUKE OF GOLD AND SILVER WARES 249

but tin and arsenic are also used with some solders. The
solder should fuse at a lower temperature than the metal-
work to which it is applied, but the less difference there is
between the two fusing points the more tenacious the joint
will be.

The compositions of a large number of solders are given by

E. A. Smith and by Gee, 2 among which are the following
l
:
250 THE PKECIOUS METALS

removed. If the required colour is not obtained in the first

dip, they are dried and dipped again. Several colouring

mixtures are in use, the composition of one of them being as
follows

.....
:

Nitre 2 parts.
...
.

Common salt . 1 part.

Alum . . 1

The pickling solution usually consists of hot nitric acid,

and after immersion in this the work is rinsed in a very
weak solution of potash to remove the acid, and then in
boiling water. After drying in warm sawdust, the wares
are often burnished, i.e., rubbed with steel or agate imple-
ments of various shapes.

Dry colouring is said by Gee not to be suitable for gold

of lower standard than 18 carats.
In wet colouring, a small quantity of water is added to
the other ingredients of the colouring pot. The wares are
prepared by careful cleaning and polishing, followed by
annealing or heating to redness on an iron or copper plate.
The blackened wares are boiled in dilute nitric or sulphuric
acid and well washed, and are then ready for the colour-
pot. The ingredients of the colouring mixture vary, the
composition of one of them being given by Gee as
l
follows :

Nitrate of potassium . . . 14 oz.

Common salt . . . . 7
Alum 7

Spirits of salts (HCl) . . 2

30 oz.
1 " Goldsmiths'
Handbook," p. 168.
 MANUFACTUEE OF GOLD AND SILVER WARES 251

The salts are crushed mixed, heated in a

fine, well
crucible and the acid added when the mixture begins to
effervesce. The wares are then immersed in the mixture

for a few minutes, well rinsed in boiling water, and again

immersed in the colouring-pot, the operation being repeated

until the desired colour is obtained. Water is added to the

colouring-pot every now and then to make up for the loss

by evaporation.
The colour of silver wares is also improved by various
methods, the simplest being the same as that employed in
the blanching of coins (q.v.).

IMITATION GOLD AND SILVER WARES.

" " "
The chief of these are gold plate," rolled gold," gold
filled," and electroplate.
Gold plate can be made from any quality of gold. A little
bar of gold of the required standard is united by soldering
to a bar of base metal, and the combined bar can be rolled

out, stamped, drawn into wire and otherwise manipulated,

the gold still covering the surface of the alloy. Even
the relative thickness of, say, one of gold to 25 of alloy

may be retained in great part. The value of gold plate is

of course very small, 9-carat gold plate costing less than

2s. per ounce. Old Sheffield silver plate was made similarly.
Electro-plating is a comparatively recent invention, not
having been used until 1840. Before that time the

ordinary method of gilding metals was to apply gold amal-

gam to the surface by means of a brush previously dipped
in a solution of nitrate of mercury. The amalgamated
article was then rinsed and dried and the mercury expelled

by evaporation over a charcoal fire. Afterwards the article

252 THE PEECIOUS METALS

was scratch-brushed and burnished, or coloured or treated by

one of the ormolu processes, 1 which consisted in heating the
articles, after covering them with a paste of alum, oxide of

iron, saltpetre, common salt, and other ingredients.

Electro-gilding is almost always carried out in cyanide

baths. The bath is usually made by the battery process.
For this Philip recommends dissolving one pound of pure
2

cyanide of potassium in a gallon of hot distilled water.

Electrodes consisting of plates of pure gold are immersed
in the liquid and a current of electricity from two Daniel
then passed through the solution until one ounce of
cells is

gold has been dissolved, a point determined by weighing

the anode. The condition of the bath may be ascertained
at any time by trying if a clean cathode of German silver
can be gilded promptly, when cathode and anode are of
about the same size. Gold chloride may be used to
strengthen the solution in gold if required. The bath is
used hot, at about 130 F., and is frequently stirred. The
hotter the solution the deeper becomes the colour of the

deposit. The
articles to be gilded are previously cleaned

by scrubbing with soap and water and dipping into potash

or ammonia solution followed by rinsing and scratch-
brushing. If the deposit is at all uneven or patchy the

wares are taken out, rinsed, and again scratch-brushed.

The anode is kept of about, the same dimensions as the
article to be gilded, the size of the anode being varied by
keeping partly immersed and raising or lowering it.
it If

the anode is too small, the deposit may become too pale ;
if

1
See "Electro-plating and Electro-refining of Metals," by A. Philip,
for full details of these processes,
2
Idem, p, 179.
 MANUFACTURE OF GOLD AND SILVER WARES 253

it is too large, the deposit becomes dark coloured. Im-

purities in the bath are to be avoided, but a very little
organic matter, such as grease, is not disadvantageous, as
it gives a deeper and richer colour to the deposit. Hot
distilled water is added from time to time to make up for
the loss by evaporation. When the deposit is thick enough,
the article is taken out and well rinsed and scratch-brushed.

Silver-plating is also almost exclusively carried out by

means of cyanide baths, which be prepared similarly
may
to gold-plating baths. About three ounces of potassium
cyanide are used to one gallon of water, and one ounce of
silver is dissolved in this. The deposit from such a bath
will be dead-white, and requires to be scratch-brushed.

By adding a little carbon bisulphide, however (about one

part of 82 to 50,000 parts of the bath) a bright deposit of
silver may be obtained. Fresh bisulphide must be added
every day.
Manufacture of Gold Leaf. Gold leaf is of various
qualities, containing from 90 to 98 per cent, of gold and
the rest and copper. The metal is cast into flat
silver

ingots, weighing 2 oz. each, and these are rolled out with
frequent annealings until about 0*0013 in. thick. The
riband of gold then cut into inch-square pieces, a packet
is

of which interleaved with vellum is beaten with a 16-lb.

convex-faced hammer until each gold leaf is 4 inches

square. They are again cut into inch-square pieces and

beaten between gold-beaters' skins with a 10-lb. hammer
until about 3J in. square,when they are put up in small
books between paper leaves. Each book of 25 leaves con-
tains from 4 to 10 grains of gold. Gold leaf is commonly
from 12 ifeocj t(>
soouoo in tnick
- -
 CHAPTEE XVII

PLATINUM

History. Platinum occurs in so many alluvial beds mixed

with grains of gold, that it would be surprising if it had
never been observed by the ancients. There is little doubt
"
that Pliny refers to it
l
under the name " aluta as occur-
ring in gold mines, and presenting itself in the form of
rounded black concretions varied with white specks, of the
same weight as gold, and remaining with it after the

washing has been completed.

It attracted the attention of Don Antonio de Ulloa when
he was at Choco, in Colombia, in 1735, with a French expe-
dition. It was brought to Europe in 1740, from Jamaica,

and was described by Watson in the Philosophical Trans-

actions for 1750. It had apparently been long known in

South America, and was called by the Spaniards " platina

del Pinto," "platina" being the diminutive of "plata,"
silver, and the Pinto the river where it was first discovered.
The Spanish Government had forbidden its export and

ordered it be thrown into the sea, to prevent it from

to

being fraudulently used to alloy with gold. In 1788, how-

ever, the Spanish Government offered Ss. per pound for

platinum (perhaps for the same purpose of mixing it with

gold), and according to Don Vincente Eestrepo of Bogota,
1
Book XXXIV., c. 47.
 PLATINUM 255

1
3,280 were obtained in that year.
Ib. All platinum came

from South America until 1823, when it was recognised in

certain white metallic grains from the Urals which had been
found in 1819. Production in Kussia began in 1824, and
since that time most of the platinum in the world has

been derived from these "placers," which at present show

no signs of exhaustion.

PROPERTIES.

Platinum is not so white as silver and has a greyish

tinge, but its lustre is not much less brilliant than that
of silver in polished specimens. Finely divided platinum
is black. Its hardness is 4*3, or greater than that of copper,
silver, or gold, and slightly more than that of phosphor
bronze. Its tenacity is between those of silver and copper.
It is malleable and ductile, but the presence of small

quantities of certain impurities greatly impairs these

qualities. Thus the presence of 0*3 per 1,000 of silicon

makes it hard and brittle, and small quantities of the other
metals of the platinum group (palladium, rhodium, iridium,
etc.) reduce its ductility.
The density of cast platinum is given by Deville and
Debray as 21'50 at 17*6 after melting in hydrogen, and by
Tilden as 21 '323 at 18. The density of platinum foil and
wire varies from 21'2 to 21"7, and that of platinum sponge
from 16-32 to 21'24 (Landolt-Bornstein Physikalisch-
Chemische Tabellen, 1905, p. 227). The addition of
iridium increases the specific gravity of platinum. Pure

platinum, according to G. Matthey, has a density of 21*46.

The coefficient of expansion of platinum is given by
1
"Mineral Industry for 1892," p. 373.
256 THE PRECIOUS METALS

Fizeau as 0'00000907 at 50, and by Le Chatelier as

Q'00001130 at 1,000. This is less than that of other
metals and is about the same as some kinds of glass, such
as Crown glass. In consequence, itcan be placed in molten
glass without causing the latter to crack on cooling, and

platinum wire is therefore used to lead electric currents

through glass, without allowing the passage of air.

Platinum melts at about 1,710, according to Harker
(Proc. Eoy. Soc., Vol. LXXVI.A, 1905), and at 1,730 accord-

ing to the Bureau of Standards at Washington (1907). An

older determination by Violle gave it as 1,775. The melting
point of platinum is lowered by arsenic, carbon, and most
other impurities. In the electric furnace Moissan found
that platinum volatilises readily, but its boiling point is
unknown, although obviously above those of gold,

copper, etc.

The specific heat of platinum is 0*0323 between and

100. Its conductivity for heat is 19, taking that of silver
as 100, but according to Dewar and Fleming its electric

conductivity is only about 13'4 at when silver 100. =

Platinum is not affected by heating in dry or moist air.
When molten
it absorbs oxygen which is given off on
cooling, sometimes with sufficient rapidity to cause the
metal to spit like silver. Oxygen is also occluded by
platinum foil, which can absorb as much as 80 volumes of
the gas at a moderate temperature. Hydrogen is occluded
in still greater quantity, from 80 to 200 volumes being
absorbed under varying conditions. The whole of these
gases can be removed by strongly heating the metal
in vacua. At a bright red heat, hydrogen rapidly passes
through platinum. At 180 platinum foil quickly absorbs
 PLATINUM 257

oxygen and hydrogen and enables them to combine, so that

it causes electrolytic gas (H 2 + 0) to explode.

Platinum black, which is platinum in the form of a finely

divided black powder produced by precipitation from

alkaline platinum solutions by means of certain substances
such as formic acid, absorbs still larger quantities of gases,
especially with great rapidity. When charged
of oxygen,

with oxygen, platinum black acts as an oxidiser, and by its

action hydrogen, CO, vapours of alcohol, ether, etc., are

rapidly oxidised and usually ignited. This is one of the

best known examples of catalytic action. The property is
used in making machines for the spontaneous ignition of
gas, alcohol vapour, etc. Platinum black charged with
hydrogen is a powerful reducing agent. Platinum black
becomes lustrous and greyish-white when burnished. Its
density varies from 15*8 to 17'6.
Platinum is not acted on by either pure hydrochloric,
nitric or sulphuric acid. It is dissolved (with the formation

of PtCLi) by aqua regia and other mixtures which evolve

chlorine, but far less readilythan gold, so that gold after
being fused so as to be firmly adherent to a platinum
crucible can be slowly dissolved by dilute aqua regia
at moderate temperatures without much injury to the
crucible. Platinum is dissolved more rapidly by aqua regia
dissolutions of chlorine under pressure. Dry chlorine is
without action on it at the ordinary temperature, but at
230 to 250 PtCl 2 is formed.
When
platinum alloyed with silver, copper, lead, zinc
is

and some other metals, it is attacked and partly dissolved

by nitric^fcid,with the formation of platinum nitrate.
When platinum black is heated with sulphur, PtS is

P.M. S
 258 THE PEECIOUS METALS

formed in the absence of air, and platinum is heated with

if

sulphur and an alkali (dry), PtS 2 is formed. Both

sulphides are decomposed at a red heat, even in the
absence of air.
When platinum heated strongly in phosphorus vapour,
is

phosphide of platinum is formed, and the phosphorus is not

entirely driven off even at a white heat in the absence of
air. Arsenic attacks platinum at a red heat, forming a
fusible arsenite. Platinum is oxidised at high temperatures

by molten alkali in the presence of air, and by nitrates even

if air is excluded. It is also attacked at high temperatures

by cyanides (forming platinocyanides), by carbon (forming

a carbide), and by silicon, or carbon and silica acting

together (forming silicides of platinum).

The atomic weight of 194*8 taking oxygen
platinum is

as 16. It is closely allied chemically with iridium and less

closely with osmium. Platinum forms two sets of com-

pounds corresponding with the oxides PtO and Pt0 2 In .

each case the oxide of platinum can act either as an acidic

or basic radicle, forming salts such as Pt(S0 4 ) 2 and
platinates such as Na 2 0.3Pt0 2 . The compounds in
which platinum is tetravalent have been studied most.
Platinum also forms a large number of complex compounds
such as H 2 PtCl 6 H 2 PtCy 4 etc.
, ,

COMPOUNDS OF PLATINUM.

Platinum dichloride or platinous chloride, PtCl 2 is formed ,

by heating platinum black in chlorine until no more of the

gas is absorbed. It can also be obtained by heating 2 PtCle H
to about 230, or by heating PtCl 4 in a stream of dry HC1.
 PLATINUM 259

It is a greenish-brown powder, insoluble in water but

soluble in hydrochloric acid in the absence of air. It

is decomposed at a red heat into platinum and chlorine,

and is readily reduced to metallic platinum by FeS(>4 and
other reducing agents. Platino-chlorides are formed by

dissolving PtCl2 in solutions of metallic chlorides. They

are highly soluble compounds of the general formula,
R 2 PtCl 4, obtained as dark red crystals by crystallising from
concentrated solutions. Platino-chlorhydric acid, H PtCl4,
2

isprobably obtained by dissolving PtCl2 in HC1, but has not

been isolated.
Platinum tetrachloride, PtCLt, is formed by heating
H PtCl
2 6 to 165 in a stream of dry HC1. It forms H
2 PtCl 6

by dissolving in hydrochloric acid, and platini-chlorides

by dissolving in solutions of metallic chlorides. Bromides
and iodides of platinum are known corresponding to these

chlorides.

Platini-chlorhydric acid, H PtCl6,

2 is formed by dissolving
platinum in aqua regia. It be crystallised out after
may
expelling the nitric acid by evaporation with HC1. It
forms red-brown deliquescent crystals, soluble in alcohol,
and decomposed by heating to about 230. Its solution is
the starting point for the formation of most of the platinum

compounds. forms platini-chlorides of many metals

It

when their chlorides are added to it in solution. Most of

these platini-chlorides are yellow-red crystalline substances,

readily soluble in water and alcohol, but the salts of

potassium and ammonium are only slightly soluble in
water and are insoluble in alcohol. These properties are
made use of for the estimation of potassium and ammonium
and also for the separation of platinum from its ores and
s 2
260 THE PKECIOUS METALS

residues. passing S0 2 through a solution of

By 2 PtCl 6 H ,

some reduction to H
2 PtCl 4 takes place, but there is no

precipitation of Pt or PtCl 2 . Platinum is precipitated by

long continued boiling with FeS0 4 and by formic acid and
,

soda carbonate acting together.

Ammonium platini-chloride, (NH 4)2 PtCl 6 occurs in yellow
,

octahedral At 20,
crystals. gram and at 100
0'665
1*25 gram dissolves in 100 c.c. of water. The substance is
nearly insoluble in alcohol. The solubility of potassium

platini-chloride, K 2 PtCl 6 is greater, T12 gram dissolving at

20 and 5'13 grams at 100 in 100 c.c. of water. When
these two salts are heated to redness they are reduced and

platinum sponge remains, mixed, in the case of potassium

platini-chloride, with potassium chloride.
When platini-chlorides are electrolysed the platinum
moves towards the anode and the solution around the
cathode becomes weaker in platinum. Nevertheless

platinum is deposited at the cathode by being reduced by

hydrogen which is liberated there.

Carbonyls of platinous chloride are formed by passing

carbon monoxide over PtCl 2 . Of these, PtC^.CO volatilises
at 250 without decomposition, thus contrasting strongly
with almost all other compounds of platinum, which
decompose when heated without volatilising. However,
PtCl 4 sublimes to some extent when heated in a current of
chlorine, like AuCla.

ALLOYS OF PLATINUM.

Platinum and Gold. See above, p. 58.

Platinum and Silver. See above, p. 140.
 PLATINUM 261

Platinum and Copper. These metals mix in all propor-

tions, but require a high temperature for their production,
like all alloys, which are usually prepared in an
platinum
oxy-hydrogen or oxygen-coal gas furnace, or in an electric
furnace. As the proportion of platinum increases, the
colour of the alloy passes from reddish-yellow to golden-

yellow and finally to white. The alloys also become

harder, and less ductile. They are less easily tarnished by
the atmosphere than ordinary brasses or bronzes. The
alloy containing platinum 18*75 per cent., copper 81'25 per
cent. (Cooper's gold) is of a golden-yellow colour and is
malleable and ductile, although hard. It is susceptible of
a high polish and closely resembles 18-carat gold. When
the metals are present in equal parts the alloy is still

yellow, but is somewhat brittle. The alloy of 3 parts

platinum to 2 parts copper is nearly white and is very hard
and brittle, without ductility.
About 4 per cent, of zinc is sometimes added to alloys of
platinum and copper, for use in jewellery, ornaments,
mathematical instruments, chronometer wheels, etc.
Platinum and Nickel. Platinum unites with nickel,

forming malleable alloys susceptible of a high polish and

not readily tarnished. Alloys of nickel, platinum and tin
have been manufactured under the name of platinum-
bronze, but generally contain very small percentages of
platinum. Equal parts of platinum and nickel yield a
yellowish-white malleable magnetic alloy, as fusible as

copper.
Platinum and Iridium. Alloys containing these metals
are very hard and elastic. They are unalterable in air and
take a high polish. They fuse at very high temperatures,
262 THE PRECIOUS METALS

much above the melting point of platinum. When not

more than about 20 per cent, of iridium is present the
alloys are malleable and ductile, but the 25 per cent, alloy
"can only with great difficulty and waste be worked into sheet
and wire when heated at low temperatures." 1 Alloys con-
taining more than 25 per cent, of iridium seem to be of very
little value. The alloy of iridium 10, platinum 90, prepared
by Messrs. Johnson and Matthey in 1870 for the Comite
International des Poids et Mesures, was chosen after exhaus-
tive trials as the material for standard weights and measures.
2
Subsequently G. Matthey recommended the alloy contain-
ing 15 per cent, of iridium for standard measures of length,
made in the tubular form, and the 20 per cent, alloy for
standard weights. The densities of platin-um- iridium
alloys, according to Deville and Debray, are as follows :

Platinum.
 PLATINUM 263

The alloys containing from 15 to 20 per cent, of platinum

were shown by Daubr6e to exhibit magnetic polarity.
Some native grains of platinum containing similar propor-
tions of iron are also natural magnets.
The alloys containing a large proportion of iron are
attracted by the magnet, but show no polarity. Stodart
and Faraday found that equal parts of steel and platinum
gave an alloy of density 9'862, taking a beautiful polish, but
not very malleable.
Platinum and Zinc alloys when treated with hydrochloric
acid sometimes leave an explosive platinum residue.
Cohen and Strengers have shown that the metal is heavily
charged with occluded hydrogen and oxygen, and that the
explosion is caused by the union of these elements.
Platinum and Lead form readily fusible grey, brittle
alloys, from which the platinum can be extracted by

cupellation. In ordinary furnaces, however, the platinum

freezes before the whole of the lead has been removed, and
it is necessary to finish the operation with the aid of the
oxy-hydrogen flame.
Platinum and Tin unite in all proportions, forming brittle
alloys, of which the melting point is higher, as in the other

alloys, in proportion as the percentage of platinum is

increased. A small quantity of platinum interferes with
the malleability of tin. Ten per cent, of platinum yields
a white ductile alloy which takes a beautiful polish.

Equal parts of tin and platinum yield a brittle, dark grey

alloy, with a coarse-grained fracture.

Platinum and Antimony unite with incandescence and
yield brittle, hard, dark grey alloys with fine-grained
fracture. A trace of antimony makes platinum brittle.
2fi4 THE PEECIOUS METALS

Platinum and Arsenic also unite with incandescence.

l
According to Friedrich and Leroux they combine to form
the compound Pt 2 As 3 which contains 36'7 per cent, of
, arsenic,
and melts at about 1,400. The alloy of lowest melting-

point contains 86'8per cent, of platinum and 13*2 per cent,

of arsenic. This eutectic alloy melts at 597 and can be
detected in all the alloys containing from 97 to 63 per cent,
of platinum. The remarkably low melting point of this
alloy accounts for the readiness with which platinum
vessels are destroyed when heated with substances contain-

ing arsenic. At high temperatures arsenic is volatilised

and spongy platinum left behind in a fit state for working.

The alloy was used for a long time after 1784 as the only
means of preparing platinum in a suitable condition to be
worked.
Platinum and Bismuth form brittle alloys with a foliated
fracture, taking a purple or violet colour when exposed to
the air. They can be partly separated by liquation at a
moderate heat.
Platinum and Cadmium are said to unite to form the
compound PtCd 2 from which any excess of cadmium can be
,

removed by volatilisation. It is a brittle, fine-grained

substance.
Platinum and Mercury do not unite directly.

OCCURRENCE OF PLATINUM IN NATURE.

Platinum occurs in flattened or angular grains in beds of

gravel or sand which resemble gold placers, and have been
formed in the same way by the erosion of older deposits.

1
Metallurgie, February, 1908, p. 148,
 PLATINUM 265

The and most extensive platinum placers occur in

richest
or near the Ural Mountains in gravels about 3 or 4 feet
thick and buried below thicker layers of barren material.

Usually, but not invariably, the gravels are also auriferous,

and other minerals occurring with the platinum are zircon,
spinel, corundum, magnetite, and osmiridium. The deposits,
besides quartz grains, contain fragments of basic magnesian
volcanic and metamorphic rocks, such as serpentine, olivine

rock, porphyries, etc.

Platinum alluvial deposits also occur in Brazil, New
Granada, California, British Columbia, New South Wales,
and in many other localities, but the quantities derived from
these places are small.
Platinum occurs in situ in serpentine in the Urals and
elsewhere, and in other metamorphic rocks in various parts
of the world, but the quantities so found are insignificant.
It is found in the form of Sperrylite, arsenide of
also

platinum, PtAs 2 associated with sulphide of nickel at

,

Sudbury in Ontario. Platinum sometimes occurs in

fahlerz, zinc blendes, lead ores, etc.
The amount of platinum occurring in placers
usually is

only a few grains per cubic yard. Sometimes nuggets are

found, but according to Schnabel and Louis the largest
" "
lump of platinum found up to the present in the alluvials
of the Ural Mountains is in the Demidoff Museum at
St. Petersburg and weighs 21*64 Ibs.
The platinum grains occurring native are alloys con-
taining iridium, rhodium, palladium, osmium, iron and
copper. The chief impurities are iridium, which may form
more than half the alloy, and iron, which has been
known to amount to 19 per cent, of the mass. Usually,
266 THE PEECIOUS METALS

however, the grains contain from 70 to 80 per cent, of

platinum.
The following analyses of native platinum concentrates
are quoted by Thorpe (" Diet, of Applied Chemistry," Vol. II.,

pp. 253, 254) from Deville and Debray.

 PLATINUM 267

The residue consists of platinum grains (which are not

attacked by mercury) together with some iridosmine, sand,
etc. These are treated for the production of commercially
pure platinum, very occasionally by cupellation with lead,
but usually by wet methods.
The Wollaston process, which came into use in London
about the year 1808, is as follows The crude platinum is
:

heated with 10 to 15 times its weight of aqua regia for

several hours. In this way a solution is obtained of most of
the metals of the platinum group, but osmiridium remains
undissolved with the sand, etc. The solution is evaporated
to a syrupy consistency to remove the greater part of the
excess of acid, and is then diluted with water and syphoned
off from the insoluble residue. The clear red solution is

treated with a solution of ammonium chloride, and the

double chloride of ammonium and platinum precipitated.
The precipitate is washed, dried and heated to dull red-
ness to drive off the ammonium chloride and chlorine,
and spongy platinum is obtained. This is broken up by
rubbing, sieved, and the finer particles made into a paste
with water and strongly compressed. The solid cake is

heated strongly and then hammered while still red hot.

The metal thus obtained can be worked by heating and
hammering or rolling, but always contains about 2 per
cent, of iridium, which is precipitated with the platinum and
can be removed only by a second refining.
The mother liquor, after precipitation of the ammonium
platmi-chloride, contains some platinum.
still The metals
are precipitated by means of iron, and the precipitate,
after digestion with hydrochloric acid, is redissolved in

aqua regia for further treatment.

268 THE PBECIOUS METALS

Some modifications have been introduced in this method.

Heraeus of Hanan evaporates the aqua regia solution to
dryness and heats to 125 in order to convert the chlorides
of iridium, rhodium and palladium to sesquichlorides and
so to prevent their precipitation by ammonium chloride.
Deville and Debray about the year 1856 introduced the
method of melting the spongy platinum in a lime furnace
by means of the oxy-hydrogen blowpipe. The furnace
consists of a hollow block of lime, made in two pieces so
that the upper half acts as a lid. The nozzle of the blow-

pipe, in which coal gas now

generally used instead of
is

hydrogen, passes through the lid, and the flame impinges

directly upon the metal. An oxidising flame is used, and
some refining effect takes place, oxides of iron and copper
being formed and absorbed by the lime, and osmium being
volatilised as oxide. The products
combustion pass out
of

through an aperture in the furnace, which also serves as an

outlet for the molten platinum.
Another method introduced by Deville and Debray has
for its object the preparation of platinum direct from its

ores in a furnace. A small reverberatory furnace with a

hollowed-out bed of firebrick linedwith marl is raised
to a bright red heat and a charge of about 200 Ibs.
of crude platinum with an equal weight of galena is
added little by little. The iron in the metal is sulphurised
and metallic lead is set free and forms a fusible alloy with
the platinum. About 200 Ibs. of litharge and some glass
are then added, the matte is oxidised, and a fusible slag
results. The osmiridium is not taken up by the lead but
sinks to the bottom and is separated. The platinum-lead
alloy is cupelled and the platinum melted in the oxy-
 PLATINUM 269

hydrogen furnace. Thus prepared it contains iridium and

rhodium, but is suitable for making vessels for chemical
purposes.
G. Matthey adopted the following process for the prepara-
tion of pure platinum :

Commercial platinum is melted with six times its weight

of pure lead, and after granulation it is dissolved slowly in
nitric acid diluted with eight times its volume of water.

The solution contains most of the lead, copper, iron,

palladium and rhodium.

The black amorphous residue is digested with weak aqua
regia, by which the platinum and lead are dissolved and
the iridium left unattacked. After evaporation the lead is

precipitated by sulphuric acid and the platinum chloride is

dissolved in distilled water and treated with an excess of

chloride of ammonium and sodium. The whole is then
heated to about 80, and in the course of a few days the
ammonio-chloride of platinum settles down as a firm deposit
at the bottom of the vessel.

The precipitate is washed first with a saturated solution

of chloride of ammonium and afterwards with hydrochloric
acid. Ehodium is separated by mixing the dried precipitate
with bisulphate of potash together with a small proportion
of bisulphate of ammonia and subjecting the mixture to a

gradual heat brought by degrees to a dull red in a platinum

basin. The platinum is reduced and the rhodium remains
as bisulphate ofrhodium and potash, which is dissolved out
in boiling water. The platinum, a black spongy residue,
is now in a high state of
purity and merely requires to be
washed and melted.
1
Fror. Roy. ,SV, Vol. XXVIII. (1879), p. 464.
270 THE PEECIOUS METALS

USES OF PLATINUM.
In 1828 the Eussian Government began the coinage of
platinum, striking 3, 6 and 12 rouble pieces. These were
demonetised and called in in 1845, as it was found that the
cost of making the coins was very great owing to the

difficulty of melting the metal. Moreover, great changes

in the value of platinum occurred rapidly owing to the

extremely limited supply. The unsuitability of platinum

for coinage purposes is due to these causes.
Owing to its high temperature of fusion and its resistance
to attack at a red heat by used largely
acids, salts, etc., it is

for the manufacture of crucibles, basins, foils, wire gauze,

etc., for chemical laboratory purposes. It is also used for

the manufacture of weights for scientific purposes, and to

serve as standards.
A far larger amount is used annually in the manufacture
of stills for the concentration of sulphuric acid. The first

still forpurpose was made in 1809, and Messrs.

this

Johnson, Matthey & Co. have used as much as 40,000

oz. ofplatinum in a single year in the manufacture of
1
improved forms of such apparatus.
The dental industry requires even more platinum, which
is used for the manufacture of pins for attaching artificial
teeth to the plate, and for other purposes.
Platinum is also used in the electrical industry to connect
the inside filament of incandescent lamps with the outside

copper wire through the glass. Its suitability for this

purpose is due to the fact that its coefficient of expansion

is nearly the same as that of glass.

1 "
Thorpe' 8 Diet, of Applied Chein.," Vol. III., p. 257.
 PLATINUM 271

Among other uses may be mentioned the manufacture of

platinised incandescent gas mantles, platinised paper for

photographic purposes, and platinised asbestos, etc., for

igniting gases. It is also employed in obtaining a silver

colour on porcelain, in watch-making, in jewellery, in

stylographic pens, in surgical and philosophical instruments,

and for "oxidising" silver. Of late years the demand for
platinum for the electrical, incandescent light, and dental
industries has increased largely whilst the new supplies
have not increased. The high price has caused very large
supplies of old material to become available.

ASSAY OF PLATINUM.

The determination of platinum may be made by weighing

it as the platini-chloride of ammonium or potassium or as
metallic platinum. The aqua regia solution containing
platinum is evaporated to dryness cautiously and then redis-

solved with the smallest possible quantity of water. Twice

the volume of absolute alcohol is then added and a large
excess of pure crystallised ammonium chloride. The liquid is
warmed, stirred, and left for twenty-four hours and the pre-
cipitate of Am 2 PtCl6 is collected on a weighed filter paper and
washed with alcohol. The washings must be evaporated
to dryness and the residue cupelled, as they contain platinum.
The precipitate may be ignited and weighed as platinum
instead of being weighed as the platini-chloride, but during

ignition part of the platinum is carried away with the

volatilising ammonium chloride.
A
small quantity of iridium usually remains with the
platinum. It may be removed by redissolving in aqua

regia and reprecipitating the platinum.

272 THE PRECIOUS METALS

The use of alloys of platinum in jewellery makes it

necessary for assays of such alloys to be often made.
The methods in use have been worked out by Chaudet,
Biche and Forest.
The alloys platinum with base metals present no
of

difficulty. They in a suitable acid which

are dissolved
removes the base metal and leaves the platinum as a residue.
If the platinum exceeds 30 per cent, of the alloy it is not
completely freed from base metal in this way, and must be
cupelled with silver and parted by means of sulphuric acid.
Platinum, Gold, Silver, Copper. The most complicated
case for assay an alloy containing platinum, gold, silver,
is

and copper or other base metal. Here it is necessary to

determine the approximate composition of the
first of all

alloy and then to make an exact assay.

In the approximate assay, the amount taken is usually 100
" "
half-milligrams, or 100 milliemes, where 1,000 equal to
is

0*5 gram. cupelled with one gram

This is of lead at a very

high temperature, and if the button is flat it is again

cupelled with more lead. When a rounded button has

been produced it is weighed and the weight accepted as
that of the gold, silver and platinum together.
The button is then inquarted by cupellation with twice
its weight of silver and is attacked by boiling in concen-
trated sulphuric acid, which dissolves the silver and leaves
the gold and platinum as a residue.
The separation of gold from platinum is effected by nitric

acid, but it is necessary for the amount of gold to be at

least ten times that of the platinum, in order that the result

may be satisfactory. This proportion of gold is accordingly

added with 2j times its weight of silver, and the cupelled
 PLATINUM 273

button is parted in nitric acid as in the ordinary gold bullion

assay. The cornet represents the original gold contents
(after allowing for the amount added) and the platinum is
taken by difference.
Thus, to take an example:

Weight of alloy taken for assay, 100 half-milligrams.

,, after cupellation (Au-f-

Pt+Ag) . 85
To this is added 170 of silver, and
the whole is cupelled. The weight
after second cupellation and parting
in sulphuric acid (Au+Pt) is . 75

To this 750 half-milligrams of gold and 2J X 825 or 2,062'5

half-milligrams of silver are added, and the whole is cupelled.
Weight after third cupellation and parting in nitric acid
800 half-milligrams, of which 750 had been added in the
form of pure gold. The approximate composition of this
alloy is thus determined to be as follows :

Gold 50
Platinum 25
Silver 10
Copper 15

100

The final assay cannot conveniently be made on a single

assay piece, because in the sulphuric acid parting it is

desirable for the assay piece to contain one and a half to
two parts of gold to one part of platinum. In the nitric
acid parting, on the other hand, ten parts of gold are
required to each part of platinum.
P.M. T
274 THE PRECIOUS METALS

Kiche and Forest accordingly recommend that 0*5 gram

("1,000" in gold bullion assay weights) of the alloy be
taken for the sulphuric acid parting if the platinum does
not exceed 300 milliemes, and that a little less weight
be taken if a greater proportion of platinum is present.
Gold is added, if necessary, to give a maximum of two parts
of gold to one of platinum.
The amount of lead required for cupellation is about
double the amount of lead which would be required in a
gold assay, or in the case under consideration, as the
copper amounts to 150 per 1,000, about 15 grams of lead
would be used for 0'5 gram of the alloy. In cases where
the platinum exceeds 200 per 1,000 of the alloy, d'Arcet
and Chaudet recommend a second cupellation with one or
two grams of lead.

The weight of the cupelled button is, of course, compared

with a check assay-piece made up in accordance with the
approximate composition of the alloy, previously deter-
mined, and this is done step by step through all the

operations. If there isa surcharge l in the check of more

than one or two parts per 1,000 in the first cupellation,

showing retention of copper, the prills are re-cupelled with

fresh lead. The corrected weight of Au+Pt+Ag is then,
say 853.
The cupelled button or prill in this case does not require
the addition of gold, and is re-cupelled with 2| parts of silver to
one part of gold and platinum taken together, that is to say-
Gold + =platinum 750 approximately
Silver already present = 100 ,,

Silver required to be added = 1,775, or 0'877 gram.

1
See above, p. 209.
 PLATINUM 275

The amount of lead required for this second cupellation is

5 grams.
The button is worked as described above in the assay of
gold bullion and the cornet boiled in concentrated sulphuric
acid for ten minutes. This is a delicate operation, as the

boiling point of sulphuric acid (325C.) is so high that glass

vessels are liable to be cracked by cold draughts. More-

over, the acid boils explosively with heavy bumping, and
boiling sulphuric acid is terribly corrosiveand inflicts
dangerous wounds. It is advisable to use a platinum

boiler, if possible, but if glass is used it must be protected

from the air by being wrapped in asbestos, and heated at
the sides more than at the bottom. Sometimes pieces of

carbon are added to assist the acid to boil quietly. After

boiling, the vessel allowed to cool almost completely, as

is

otherwise it would crack during decantation. The liquid is

then decanted into a dry vessel, or into a considerable bulk
of cold water, to avoid danger from the heat generated by
mixing sulphuric acid and water. The decantation must
be as complete as possible, to avoid heating on the addition
of the washing water, which should be warm in order to

dissolve the sulphate of silver. After washing twice, the

cornet is again boiled for ten minutes in sulphuric acid
and washed, dried, and annealed. Its weight gives the
amount of gold -f- platinum, in this case, say 752*5.
For the determination of the gold, 0'25 gram or 500

half-milligrams of the original alloy is taken, containing

250 gold
125 platinum
50 silver

75 copper.
T 2
276 THE PKECTOUS METALS

This is cupelled with 0'5

gram of pure gold (making ten
parts of gold to one part of platinum) and 1'412 grams of
silver and 5 grams of lead. The resulting button is worked
and parted in nitric acid as described in the assay of gold

bullion. If, as generally happens, there is a surcharge in

the check assays of more than one or two parts per 1,000,

showing the retention of platinum, the inquartation and

parting is repeated. The finalthen, say 1,248
weight is

half-milligrams, and after deducting 1,000 and doubling the

remainder, the result 496 is obtained for the amount of
gold present in 1,000 parts of the original alloy.
The final result of the assay will then be reported as
follows :

Gold
Platinum
Silver
.... 496-0
256'5
100'5

Copper (by difference) . . 147 '0

Total 1000-0

If necessary as small a quantity as 01 gram may be

taken for assay. This is a convenient amount for alloys
containing 900 to 950 platinum and the remainder gold.
It is, of course, obvious that alloys of platinum and silver

only can be assayed by sulphuric acid parting without the

other methods, and platinum-gold alloys do not require the

sulphuric acid parting. The cupellation requires a higher

temperature in proportion as the percentage of platinum
is greater, and alloys containing over 50 per cent, of

platinum cannot be freed from lead without the help of the

oxygen-gas blowpipe.
 CHAPTEE XVIII

PRODUCTION AND CONSUMPTION OF THE PRECIOUS

METALS

PRODUCTION OF GOLD IN THE WORLD.

No means amount of gold pro-

exist of estimating the

duced in ancient times. Doubtless it was small compared

with the output during the last half-century. The con-

sumption, however, was much less, and considerable stocks

accumulated in the hands of certain monarchs. Thus the
treasure Pytheus, 470 B.C., has been estimated at
of

3,600,000, derived from the sands of the Pactolus in

Phrygia; and Jacob states that Alexander acquired in Persia
gold probably amounting in value to over 50,000,000.
Production of gold in the Middle Ages was very small,
most of the old goldfields in
Egypt, Spain, Phrygia, India,
etc., having been exhausted and no new ones discovered.
After the discovery of America in 1492, gold reached Europe
from the West in a steady and increasing stream. The
following table gives Soetbeer's estimates of the annual
production between 1492 and 1885, and the estimates of
the Bureau of the United States Mint for the years 1886
to 1906.

The total amount of gold produced in the period is given

as 603,781,000 fine ounces of the value of 2,412,750,000.
278 THE PRECIOUS METALS
PRODUCTION OF GOLD IN THE WORLD.

Period.
PEODUCTION, ETC., OF THE PBECIOUS METALS 279

Country.
280 THE PKECIOUS METALS

The production of Canada in 1906 came chiefly from the

Yukon (250,000 oz.) and from British Columbia (296,000 oz.).
A little is produced in Ontario and Nova Scotia.
The output of Africa in 1906 and 1907 is divided as
1
follows :
 PRODUCTION, ETC., OF THE PRECIOUS METALS 281

is smuggled out of the country. It is probable that this

gold is not reckoned in the output of any country.

PRODUCTION OF SILVER IN THE WORLD.

The following table is from the Annual Report of the

Director of the United States Mint for 1907, p. 142. It

gives the annual average production of silver in the world

since the discovery of America in 1492 :

Annual average.
Period.
Ounces (fine).

14931520 1,510,000
15211580 7,040,000
15811640 13,240,000
16411700 11,200,000
17011760 14,140,000
17611820 24,100,000
18211830 14,807,000
18311840 19,175,000
18411850 25,090,000
18511860 28,791,000
18611870 39,226,000
18711880 71,046,000
18811890 100,457,000
18911895 157,581,000
18961900 165,693,000
19011905 167,450,000
1906 165,755,000

The production "

of silver is given by the Mineral
Industry for 1907," p. 474, as 184,552,343 fine ounces for
1906, and 193,542,381 fine ounces for 1907.
2S2 THE PEECIOUS METALS

The production of silver for the year 1906 for different

countries given by the Director of the United
is States
Mint as follows :

Country.
 PRODUCTION, ETC., OF THE PEECIOUS METALS 283

Country. Ounces (fine).

Brought forward . .
162,895,464
South America continued
Colomhia .... .

763,335
Chile

Argentine
Ecuador
.... 397,853
14,440
13,592
Central America . . .
1,670,159

Total . . .
165,754,843

The production of silver in the United States is divided

between the following States and Territories :

Ounces (fine).

Montana 12,540,300
Colorado
Utah
Idaho
...... 12,447,400
11,508,000
8,836,200
Nevada 5,207,600
Arizona 2,969,200
California 1,517,500
New Mexico 453,400
Texas 277,400
Alaska 203,500
Michigan 186,100
South Dakota 155,200
Oregon 90,700
Other States and Territories 125,400

Total . . .
56,517,900
284 THE PEECIOUS METALS

CONSUMPTION OF THE PKECIOUS METALS.

One of the difficulties in estimating the amounts of gold

and silver used in the Arts is that variable proportions of
old material (jewellery, etc.) are used with new material.
In 1880 Dr. Adolf Soetbeer estimated the consumption of
the precious metals by the civilised nations of the world
at 2,689,000 oz. of gold and 15,142,000 oz. of silver,

reckoning new material only. The following are estimates

made by the Director of the United States Mint :

Year.
 PEODUCTION, ETC., OF THE PRECIOUS METALS 285

remelting old coin. In the same year 120,339,500 oz. of

silver were used in coinage, of which 34,200,000 oz. con-

sisted of old coin.

World's Stock of Precious Metals. The stock of gold and
silver in the world has increased largely in recent years, but
was formerly very small. Jacob estimated the amount of
gold and silver in circulation in Europe in 1492 at only
34,000,000, and MacCulloch estimated that this had risen
to 400,000,000 in 1850. In 1894 the amount of gold coin
only was put at 813, 500,000.* According to the Director
of the United States Mint, the world's stock of gold on

December 31, 1906, was of the value of $6,888,900,000, or

about 1,425,000,000, of which $3,764,900,000, or about
779,000,000, was in banks and public treasuries. It is pro-
bable that the estimate of the amount in circulation

(646,000,000) is too high. The total stock of silver at the

same date is put at $3,260,200,000, or about 674,000,000

at coining value. The stock of gold coin in the United
Kingdom is estimated to be about 100,000,000, and that
of silver coin to be about 25,000,000.

PRODUCTION OF PLATINUM.
The greater part of the annual production of platinum is
obtained from Kussia. It is estimated that the total
amount produced in that country from its discovery in 1819
down to 1907 was 7,500,000 oz., including 1,500,000 oz.,
stolenby the miners, which has secretly found its way to the
market and is not accounted for by the Government
2
statistics. The production, at first small, rose to 110,000 oz.

"
Eeport of the Director of the United States Mint for 1894."
1

" Mineral
Industry for 1892," p. 388, and subsequent years.
2
286 THE PRECIOUS METALS

per annum in 1846, but the coining of platinum

after

by the Kussian Government had been discontinued, the

output fell in 1852 to 6,400 oz. 1
Subsequently the amount
gradually increased and reached an average of 53,000 oz.
per annum in 1876, 139,000 oz. in 1886, and 200,000 oz.
2

in 1896. The output has been stationary for some years,

and amounted to 210,318 oz. in 1906. 3 In that year 120
platinum placers were being exploited, employing 6,200
workmen who washed 1,886,000 tons of alluvium and
obtained an average of 0'09 oz. crude platinum per ton of
1
gravel.
The production in other countries is insignificant.
Colombia produces about 4,000 oz. per annum, the United
States a few hundred ounces, and Australia, Canada,

Borneo, etc., smaller quantities.

The price of platinum has risen in an extraordinary
manner in the past few years. After being between 30s.
and 40s. per oz. for many years prior to 1894, it rose to
i'4 10s. per oz. in 1905, and reached a maximum of 8 per
oz. in 1906. This great increase in price led to no unusual
activity on the part of the producers in the Urals, because

the entire output for a number

years to come is stated to
of

have been bought at fixed prices by the dealers in platinum.

The consumption of platinum is put at 325,000 oz. per
annum, or much more than the production. The deficiency
is made up by old platinum which comes into the market,

partly owing to the high price of the metal.

1 " Mineral
Industry for 1906," p. 656.
2 " Ore
Deposits," J. A. Phillips, 1884, p. 402.
3 " "
Mining Journal," May 4, 1907, Mineral Industry for 1907, "p. 784.
 INDEX
A. Arrastra, 154
Ar sen ate of silver, 151
ABSTRICH, 179 Arsenic-gold alloys, 51
Agricola, G., 7, 8, 104, 212 Arsenide of gold, ;-!6, 51
Alchemy, 6 Assay furnace, 202
Alloy, parting, 180 of gold bullion, 206
Alloys of gold, 37 of ores, 197
platinum, 260 of platinum, 271
assay of, 272 of silver bullion, 149, 211
silver, 132 -ton, 199
Aluminium-gold alloys, 60 Augustin process, 167
-silver alloys, 142 Aurates, 33
Amalgam, 44, 72, 137 Auric bromide, 26
purification of, 164 chloride, 22
retorting, 164 oxide, 33
Amalgamated plates, 9, 93 Auricyanides, 31
dressing of, 97 Aurocyanides, 28
electro-plated, 94 Aurosulphites, 34
scaling and sweating of, 98 Aurous bromide, 26
shaking, 104 chloride, 21
Amalgamation, invention of, 8 cyanide, 28
of gold, 72, 85 oxide, 33
of silver, 154, 161 Automatic balance, 235
Ammonium platino-chloride, 260
Ancient coinages, 5
furnaces, 2, 3 B.
methods of gold extrac-
tion, 2, 3 BAGRATION, 29
methods of silver extrac- Balance, automatic, 235
tion, 128 Ball mills, 101
names of gold, 2 Barba, 9
Annealing in mints, 228, 231 Barnard, J. E., 227
Antimonate of silver, 151 Barrel amalgamation, 160, 161
Autimonide of gold, 36 chlorination, 12(5

Antimony-gold alloys, 50 Base bullion, 175

Arbor Dianae, 138 Batea, 75
Argentiferous lead, 136 Bay, Dr., 31
desilverisation of, 175 Bismuth-gold alloys, 51
sampling of, 175 -silver alloys, 143
Argentite, 152 Black sand, 79
288 INDEX

Blanching silver, 232 Cleaning-up in cyanide process,

Blast furnaces for smelting ores, 113
172, 173 placer operations,
Blicksilber, 180 80
Bluestone, manufacture of, 188 stamp mills, 98
Bodlander, G., 29 Clean-up barrel, 98
Bourke, J. J., 235 Cobalt-gold alloys, 57
Bromide of gold, 25 Cohen, 263
silver, 149 Coin, manufacture of, 224
Bromite, 153 Coinage in early times, 5
Brown gold, 188 world, 284
Bullion, assay of, 206 Coins, composition of, 238
base, 175 Colloidal gold, 16
melting of, 224 Coloured gold wares, 244
valuation of, 222 manufacture of, 249
Colouring silver wares, 251
Composition of coins, 238
gold wares, 242,
243
CADMiUM-gold alloys, 48 silver wares, 245
-silver alloys, 135 solders, 248
use of in assaying, 211 Comstock lode, silver at, 152
silver plate, Concentrates, treatment of, 108,
246 124
Calaverite, 67 Concentration of gold ores, 106
Caldecott, W. A., 100, 101, 117, Coning and quartering, 198
170, 171 Consumption of gold and silver,
Caldron process, 158 284
Cams in stamp battery, 90 platinum, 286
Carat system, 242 Cooper's gold, 261
Carbon bisulphide, use of in Copper amalgamated plates, 9
silver plating, 253 -gold alloys, 40
Carbonyls of platinum, 260 -silver alloys, 132
Cassius, purple of, 17 sulphate, manufacture of,
Castillite, 152 188
Cazo process, 158 Cornets in gold bullion assay,
Cement gravels, 72 209
mill, 77 Cosaiite, 152
silver, 167 Cradle, 75
Chasing wares, 248 Crushing by ball mills, 101
Chaudet, 272, 274 roller mills, 103
Chile nite, 153 rolls, 123
Chloride of gold, 20 stamps, 85
silver, 147 dry, 123
solubility of, 148 Crystals of gold, 12, 65
Chlorination process, 124 Cupellation in assaying, 201,
Chlorine gas, refining by, 189 211
Classification of pulp, 108, 110 losses of gold and silver in,
Claudet, A. C., 137 203, 209, 212
 INDEX 289

Cupellation in English furnace, Electrolytic process of gold

180 refining, 192
German furnace, of silver extraction,
179 181
of argentiferous lead, refining,
178 191
Cupels, 201 Electrum, 5, 40, 238
Cyanide of gold, 28 Eisner, 29
and potassium, Einbolite, 153
29 Engraving wares, 248
silver, 149
and potassium, P.
150
process, 110, 170 FAHLORE, 152
solution in stamp battery, Faraday, 263
172 Faraday's gold, 16
Feeders to stamp battery, 87
D. Filter pressing, 115
vacuum frame, 172
D'ARCET, 211, 274 Flashing of prills, 208
Daubree, 263 Fleming, 256
Debray, 255, 262, 266, 268 Fluxes in ore assay, 199
Decantation cyanide process, 114 Fondon. process of silver amalga-
Denver mint, refining at, 192, 194 mation, 159
Deville, 255, 262, 266, 268 Forest, M., 272, 274
Dewar, 256 Francke-tina process, 159
Diehl process, 122 Freieslebenite, 152
Dies, in coining, 233 Friedrich, 145, 176, 264
in manufacture of wares, Frue vanner, 107
247 Fulminating gold, 34
Diodorus Siculus, 3, 5 Furnaces, assay, 202
Distribution of gold, geographi- cupellation, 179
cal, 71 melting, 2, 224
Dodd, Dr., 216 roasting, 123
Doerinckel, 58, 140 Fusion in ore assay, 199
Don, Dr., 66, 70
Dredges, 78, 81 G.
Dry crushing, 123
Dyscrasite, 152 GAY-LUSSAC, 215
process, 212
effect of
E. mercury
in, 215
EGYPT, alchemy in, 6 Geber, 9
gold in, 1, 2, 3, 5 Gee, 244, 248, 249, 250
Egyptian furnace, 3 Gilding, electro, 251
Electro-plate, 251 wet, 251
precipitation of gold in Gold, alloys with aluminium, 60
cyanide process, 115 antimony, 50
P.M. U
290 INDEX

Gold, alloys with arsenic, 51 Gold wares, soldering, 248

cadmium, 48 Gold-filled wares, 251
cobalt, 57 Golden fleece, 3
copper, 40 Gowland, W., 1, 3, 143
iridium, 60 Granulating alloys, 187
iron, 55 Gravels, auriferous, 70
lead, 52 treatment of,
manganese, 58 2, 72
nickel, 57 Green gold, 37
palladium, 60 Grey silver, 131
platinum, 58, ore, 152
272 Grizzly, 79
rhodium, 60 Gutzkow process, 189
silver, 37
tellurium, 62
thallium, 63 H.
tin, 49
zinc, 46 HALL-MARKING offices, 243
antimonide, 36 Hammered money, 220
arsenide, 36 Harker, 256
bromides of, 25 Hatchett, 57
bullion, assay of, 206 Hautefeuille, 144
chemical properties of, 12 Heraeus, 268
chlorides of, 20 Hessite, 153
compounds of, 20 Heycock, C. T., 48, 132, 133
cyanide of, 28 History of coining, 219
discovery of, 1 gold, 1

geographical distribution of, 128

silver,
71 Hoefer, F., 128
6,
history of, 1 Herman, H. O., 169
identification of, 15 Hoitsema, 38, 41
iodides of, 27 Hornsilver, 147
nitrate of, 33
Huntingdon mill, 103
oxides of, 32 Hydraulic classifiers, 108
phosphide of, 35 elevator, 77
plate, 251 mining, 77
production of, 277
properties of, 11
pure, preparation of, 18 I.

selenide, 35
silicates, 35 INDIA mint, assay method, 217
solvents of, 14 Industrial arts, gold and silver
sulphides, 35 used in, 284
telluride of, 35 platinum used
thiosulphate of, 34 in, 286
wares, 242 Inquartation, 204
colouring, 249 Iodides of gold, 27
composition of, 243 silver, 149
manufacture of, 247 lodite, 153
 INDEX 291

Iridium in gold, 60 Losses of gold and silver in assay-

Iron-gold alloys, 55 ing, 203, 209, 212
Isaac, oo Louis, H., 265
Lunar caustic, 150

J.
M.
JACOB, 277, 285
Jars, M., 7 MACCULLOCH, 285
Jewellery, alloys used in, 40, 44, Magistral, 156
55,140,242,261,272 Manganese gold alloys, 58
manufacture of, 247 Manufacture of gold and silver
wares, 247
Maria Theresa thaler, 241
K. Markes, E. S., 78, 84
G. N., 78, 84
KASENTSOF, 44 Marking machine, 231
Kerargyrite, 147, 153 Marriner process, 122
Kiss process, 170 Matthey, E., 58
Kongsberg, silver ores at, 151 G., 255, 262, 269
Kronke, 160 Matthiessen, 38, 42
Krupp ball mill, 101 Melbourne Mint, 191, 223
Kurnakoff, 42 Melting bullion, 99, 224, 225
gold for wares, 245
silver for wares, 246
L. Mercury, action of, on AgCl, 148
-gold alloys, 44
LAZZLO amalgamator, 104 in Gay-Lussac process,
Le Chatelier, 256 215
Lead, argentiferous, sampling of, in gilding, 251
175 in gold extraction, 8,
treatment of, 92, 94
175 loss of, in gold mills, 1 05
-gold alloys, 52 in patio pro-
-silver alloys, 136 cess, 157
-zinc alloys, liquation of, in silver ex-
178 traction, 154,
Lenormant, 23S 156
Leroux, 1-15, 2(54 purification of, 106
Levin, 57, (>3 -silver alloys, 137
Levol, 137 traps, 96
Level's alloy, 132 wells, 97
Lindborn, 30 Metallics, 198
Litharge, manufacture 179
of, Milled edge of coins, 221
silver in, 179, ISO Mint, Melbourne, 191, 223
Liversedge, 69 Eoyal, 222
Lixiviation of gold ores, 110 sweep, 225
silver ores, !(>(}
Sydney, 184, 223
LOFS of mercury, 105 United States, 192
292 INDEX

Minting, 219 P.
Moebius process, 182, 191, 192
Moisson, H., 18, 50 PALLADiUM-gold alloy s, 60
Moore filter, 172 -silver alloys, 143
Morison, D. B., 91 Pan amalgamation, 161, 164
Mortar in stamp battery, 86, 88 miners, 74
Moulds for coinage bars, 227 Parke's process, 48, 137, 177
Mount Morgan mine, 126 Parting alloy, 186
Mullers, 161 in assaying, 204,
207/209, 272,
275
assay, 204, 206
N. flasks, 209
process, chlorine, 189
NAGYAGITE, 67 electrolytic, 192
Native gold, 64 nitric acid, 184
Native silver, 151 sulphuric acid,
Naumannite, 153 185
Neville, 48, 132, 133 processes, history of, 9
Nichols, W. G., 235 Patera process, 168
Nickel-gold alloys, 57 Patio process, 156
-silver alloys, 143 Pattinson process, 175, 176
Nitrate of gold, 33 Percy, Dr., 137, 138, 186
silver, 150 Petrenko, 133, 136, 139, 142, 143
Noble metals, 12, 54, 197 Petzite, 67, 153
Nuggets, gold, 64 Philip, A., 252
Phillips, J. A., 286
Phosphide of gold, 35
Pickling silver plate, 246
0. Placers, gold, 64
platinum, 264
OBRUSSA (ancient process), 6 Plates, amalgamated, 93
Ores of gold, 64 Platinichlorides, 259
assay of, 197 Platinum, 254
origin of, 69 alloys of, 260
treatment of, 74 assay of, 271
platinum, 265 black, 257
treatment of, chlorides, 258
266 compounds of, 258
silver, 151 gases absorbed by, 256,
assay of, 197 257
sampling of, 198 -gold alloys, 58
treatment of, 154 native, 264
Ormolu process, 252 ores, 26'J
Osmium-gold alloys, 60 price of, 286
" "
285
Overlooking machine, 246 production of,
Oxide of gold, 32 properties of, 255
silver, 130, 144 pure, preparation of,
" Oxidised" 145 269
silver,
 INDEX 293

Platinum-silver alloys, 140 Rock-breaker, S(>

trays in parting assay, Rocker, 75
208 Rolled gold, 251
uses of, 270 Roller crushing mills, 103
Plattner process, 125 Rolling coinage bars, 227, 228
Pliny, 6, 8, 80, 128, 254 Rolls, crushing, 123
Poitevin, 143 Romans, refining by, 5
Polybasite, 152 treatment of silver ores
Posepny, 70 by, 128
Precipitation of gold by zinc, 30 Roskelley, 94, 97
Presses, coining, 233 Rossler, Dr., 186
falling weight, 248 Royal Mint, 224
Prill, 203 Rozan process, 176
Prister, 31 Ruby silver ores, 152
Production of gold, 277 Ruer, 60
platinum, 285 Russell, 146
silver, 281 process, 170
Proof gold, preparation of, 18
Proofs in bullion assay, 209
Properties of gold, 11 S.
platinum, 255
silver, 129 SAMPLING ores, 198
Proustite, 152 Scheinnitz mill, 104
Pyrargyrite, 152 Schlieinann, Dr., 5
Schnabel, 265
B. Scorification, 201
Screens in stamp battery, 88
KEESE RIVER process, 165 Sea, gold in, 69
Kefming gold, 5, 183 Selenide of gold, 35
by chlorine, 189 silver, 150
electrolysis, 191 Shaking amalgamated plates, 104
oxygen, 190 Siemens-Halske process, 115
sulphuric acid, Sieves, invention of, 5
185 Silicates of gold, 35
silver, 183 Silver alloj's with aluminium, 142
Re-grinding, 117 bismuth, 143
Remedies in coinage, 235 cadmium, 135
Retorting amalgam, 99, 164 copper, 132
Rhodium-gold alloy, 60 gold, 37
Richards, R. H., 89, 91, 1)2 lead, 136, 175
Riche, 272, 274 inercurv, 137
Rickard, T. A., 70, 89 nickel, "l43
Rid g way filter, 172 palladium, 143
Riffles, 77 platinum, 140
Rigg, E., 231 thallium, 143
Roasting furnaces, 123, 165 tin, 138
silver ores, 165 zinc, 133
Roberts- Austen, W. C., 37, 40, 41, antimonate, 151
43, 132 arsenate, 151
294 INDEX
Silver assay of, 211 Stas, 147
bromide, 149 pipette, 213
bullion, assay of, 211 Stephanite, 152
chloride, 147 Stock of gold in the world, 285
cyanide, 149 silver in the world, 285
history of, 128 Stodart, 263
iodide, 149 Strabo, 128
native, 151 Strengers, 263
nitrate, 150 Striking coin, 233
ores, 151 Stromeyerite, 152
amalgamation of, 154 Sulphide of gold, 35
cyaniding, 170 silver, 130, 144, 156
lixiviation of, 167 Sulphocyanide of gold, 32
smelting of, 172 Sulphuric acid, parting by, 185
oxide, 130, 144 Surcharge in bullion assay, 209
"
oxidised," 145 Sweep, mint, 225
plating, 253 Sydney mint, 184
properties of, 129 Sylvanite, 67
pure, preparation of, 130
selenide, 150
spitting of, 129 T.
sulphate, 151
sulphide, 130, 144 TAILINGS, treatment of, 106, 110
decomposition of, Tammann, 55
15(5 Tavener process, 113
preparation of, Telling coins, 237
169 Telluride of gold, 35
telluride, 150 silver, 150
Slimed ore, cyaniding, 114, 115 ores, treatment of, 122
Sluice, 76 Tellurium-gold alloys, 62
Smelting silver ores, 172 Thallium-gold alloys, 63
Smith, E. A., 138, 216, 245, 249 -silver alloys, 143
F. C., 67 Thiosulphate of gold, 34
S. W., 227 silver, 168
Soetbeer, A., 277, 284 sodium, 168
Soldering gold and silver, 248 sodium and cop-
Sonstadt, 69 per, 146
Sperrylite, 265 Thonnelier, 233
Spinning gold and silver, 248 Thorpe, 266
Spitting of silver, 129, 208 Tilden, W. A., 255
Spitzkasten, 120 Tina process, 155
Spurr, J. E., 70 Tin -gold alloys, 49
Stamp-battery, 5, 7, 85 -silver alloys, 138
Stamps, duty of, 120 Torta, 156
order of drop, 90 Toughening gold bullion, 185
parts of, 85 Troost, 144
Standards of gold wares, 242, 243 Tube mills, 117
Standards of silver wares, 245 Tyrol, ancient methods of amalga-
Standard silver, 133 mation in, 9, 104
 INDEX 295

U. Wares, soldering, 248

standards of, 243
UHLHORX coining press, 233 Washing gold, 72, 74
Undercurrent, 80 Washoe process, 161
Uses of platinum, 270 Weed, 70
Utrecht mint, silver assay at, 149 Weighing in assaying, 204
Wet gilding, 251
processes of gold extraction,
V. 110
silver extrac-
VACUUM filter frames, 172 tion, 166
Violle, 256 White precipitate, 31
Vogel,47, 48, 49, 51, 52 Wilfley table, 107
Volhard, J., 216 Wire drawing, 247
process, 216 Wohlwill process, 192, 193
Vondracek, 148, 156, 157 Wollaston, 267
Wrought wares, 248
w.

WAGONER, L., 66, 68, 70

Wait, 144 ZIERVOGEL process, 166
Wares, colouring, 249 Zinc, desilverisation of lead by,
composition of, 242, 2-15 177
gold, 242 -gold alloys, 46
imitation, 251 in lead smelting, 1 75
manufacture of, 247 -silver alloys, 133
melting and rolling, 245

BRADBUKY, AUNKW, & CO. LD.. PRINTERS, LONDON AND TONBKJDCJE.

(D
RETURN CIRCULATION DEPARTMENT
202 Main Library
LOAN PERIOD 1

HOME USE
 YC 19165

THE UNIVERSITY OF CALIFORNIA LIBRARY