anxa
2713
-974
A TEXT BOOK
OF
FIRE ASSAYING
EDWARD E. BUGBEE
© Raymond Pettibon
RESEARCH LIBRARY
THE GETTY RESEARCH INSTITUTE
JOHN MOORE ANDREAS COLOR CHEMISTRY LIBRARY FOUNDATION
‘
ex LEX T- BOOK
OF
RE ASSAYING
.
a
A TEXT-BOOK
OF
FIRE ASSAYING
EDWARD E. BUGBEE
Assistant Professor of Mining Engineering and Metallurgy,
Massachusetts Institute of Technology.
Printed for use in Gofinection with
THE COURSE IN FIRE ASSAYING AT THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY.
¢
fy rs wh rm Dy oa \
‘ ! 4 . Me hee va
4
Copyright 1915, rele
. By EDWARD E. BUGBEE.
PRESS OF
BOSTON, MASS. Ne
¢
THE GETTY RESEARCH
INS UW |BRARY Sem
CONTENTS
CHAPTER I
AssAY REAGENTS AND FUSION PRODUCTS .......2.+eee Silos
. Definitions. Reagents. Chemical Reactions of Reagents. Fusion
Products.
é
| CHAPTER II
. FURNACES AND MrrauuurcicaL Cuay Goops ........4.4...
Crucible Furnaces. Muffle Furnaces. Fuels. Coal Furnace. Coke
Furnace. Gasoline Furnace. Gas Furnace. Crude Oil Furnace.
Furnace Repairs. Muffles. Crucibles. Scorifiers. Roasting Dishes,
etc. Testing Crucibles.
CHAPTER III
en eg ig oe pence as amo e wow he Sh wt ewe
Definitions. Moisture Sample. Sampling Operations. Theory of
Ore Sampling. Duplicate Sampling. Sampling Ore Containing
Malleable Minerals.
SHAPLER.LY
UMIIMMERRCTOMVVCRSIPER ¢ 4 . 6 4k ke we ee ke we 8s
Flux Balance. Pulp Balance. Button Balance. Theory of the
Balance. Directions for use of Balance. Weighing by Equal Swings.
Weighing by No Deflection. Weighing by Substitution. Check
Weighing. Adjusting and Testing an Assay Balance. Weights.
Calibration of Weights. Riders. Testing Riders.
CHAPTER V.
Cupellation. Assay of Lead Bullion. Loss of Silver in Cupelling.
Effect of Temperature. Protective Action of Silver on Gold. Influence
of Impurities. Indications of Metals Present. Testing Cupels. Re-
tention of Base Metals. Portland Cement and Magnesia Cupels.
9-20
21-36
37-50
51-67
iv
CHAPTER VI
PARTING °°) 2 soe See ee ee eee
General Statement. Parting in Porcelain Capsules. Inquartation.
Parting in Flasks.
CHAPTER VII
Tue ScoriFICATION ASSAY
@ @ @) -@. “re. Sree, 8 Ry i ie ee
General Statement. Solubility of Metallic Oxides in Litharge. Heat
of Formation of Metallic Oxides. Ignition Temperature of Metallic
Sulphides. Assay Procedure for Ores. Chemical Reactions. Color
of Scorifiers. Assaying Granulated Lead. Scorification Assay for
Gold. Scorification Assay of Copper Matte. Table of Scorification
Charges for Different Ores.
CHAPTER VIII
Tar Cructpir Assay
Theory of the Crucible Assay. Classification of Ores. Crucible
Slags.‘ Classification of Silicates. Fluidity of Slags. Acid and Basic
Slags. Mixed Silicates. Notes on the Fusibility of Silicates. Slags
for Class 1 Siliceous Ores. Slags for Class 1 Basic Ores. Reducing and
Oxidizing. Reducing Power of Minerals. Slags for Class 2 Ores.
Action of Borax in Slags. The Cover. Testing Reagents. Assay of
Class 1 Ores. Methods for Assay of Class 2 Ores. The Niter Method.
Preliminary Fusion. Estimating Reducing Power. Regular Fusion.
The Iron Method. The Roasting Method. Assay of Class 3 Ores.
Determining the Oxidizing Power of Ores.
CHAPTER IX
Sprciat MetHops of AssAy .....2.... | ee
The Assay of Telluride Ores. Crucible Methods for the Assay of
Ores and Products High in Copper. The Assay of Antimonial Gold
Ores. The Assay of Auriferouis Tinstone. The Corrected Assay.
CHAPTER X
THe Assay OF BULLION o o e . cde a Pee 4 . e . é * . . * . .
Definition. * Weights. Sampling Bullion. The Assay of Lead
Bullion. The Assay of Copper Bullion. Atl Fire Method. Crucible
Method. Mercury—Stulphuric Acid Combination Method. Nitric
Acid Combination Method. The Assay of Doré Bullion. U.S. Mint
Assay of Gold Bullion. |
68-72
73-82
83-111
. 112-118
119=132
PS ee
ne et a
CHAPTER XI
OC ROL UTIONS we es ek wee
_ Evaporation in Lead Tray. Evaporation with Litharge. Precipita-
tion by Zinc and Lead Acetate. Precipitation as Sulphide. Precipi-
tation by Cement Copper. Precipitation by Silver Nitrate. Precipi-
tation by a Copper Salt. Electrolytic Precipitation. Colorimetric
~ Method.
CHAPTER XII
a BO) rc
_ General Statement. Lead Ores. Accuracy and Limitations of
Method. Quantity of Ore and Reagents Used. Manipulation of
Assay. Influence of Other Metals. Procedure for Assay. Assay of
~Slags. ;
e e e ° ° ° ° e e e ° e ° e e ° ° e e e e e e e .
133-139
140-145
147-150
= 4
7 ~
py
an A y,
eye
h i Fi - S
4 s
: A
:
‘
4
c
it
5
by)
3
E
4
%
4
J
A TEXT-BOOK OF FIRE ASSAYING.
GHAPTER. I.
ASSAY REAGENTS AND FUSION PRODUCTS.
Assaying is a branch of analytical chemistry, generally defined as
the quantitative estimation of the metals in ores and furnace prod-
ucts. Inthe Western part of the United States, the term is employed
to include the determination of all the constituents, both metallic
and non-metallic, of ores and metallurgical products.
Fire Assaying is the quantitative determination of metals in ores
and metallurgical products by means of heat and dry reagents. This
involves the separation of the metal from the other constituents of
the ore and its weighing in a state of purity.
The reagents used in fire assaying may be classified as fluxes; acid,
basic or neutral, and as oxidizing, reducing, sulphurizing or desulphur-
izing agents. Some reagents have simply one property, as for in-
stance silica, an acid flux, others have several different properties, as
litharge, a basic flux but also an oxidizing and desulphurizing agent.
A flux is something, which, if added to a body infusible by itself or
fusible only with difficulty, will cause it to fuse at a lower temperature
than it would have done alone. For instance, quartz by itself is’
fusible only at a very high temperature, but by adding some sodium
carbonate, to the pulverized quartz it can be fused at a temperature
easily obtained in the assay furnace.
The student should remember that to aid in the fusion of an acid
substance, a basic flux such as litharge, sodium carbonate, limestone,
or iron oxide should be added, for a basic substance an acid flux such
as silica or borax should be used.
The principal fluxes used in assaying follow:—
Silica, SiO», is an acid flux and the strongest one we have. It
combines with the metal oxides to form silicates which are the founda-
tion of almost all of our slags. It is used as a flux when the ore is
deficient in silica and serves to protect the crucibles and scorifiers
from the corrosive action of litharge. Care must be taken to avoid
2
an excess of silica, as too much of it will cause trouble and losses of
precious metals by slagging or by the formation of a matte. Silica
melts at about 1625° C. to an extremely viscous liquid. (Day «&
Shepherd). It should be obtained in the pulverized form.
Glass is used by some in place of silica. Ordinary window glass,
a silicate of lime and the alkalies with the silica in excess, is best. Its
acid excess is always doubtful and so is not commonly used. If used,
a blank assay should be run on each new lot to insure against intro-
ducing precious metals into the assay in this way. Its chief advan-
tage is that 5 or 10 grams too much glass will ordinarily do no harm
in a fusion whereas 5 or 10 grams of silica in excess might spoil it.
Borax, Na:B,0;,10H:O, acts as an acid flux. It contains a large
amount (47 per cent) of water which is given off on heating. During
the heating, the borax swells to more than twice its original bulk,
and if an excess of it is used, and especially if not thoroughly mixed
with the charge, it may force part of the charge out of the crucible.
It should never be used for scorification work.
Borax Glass, Na2,B,O;, is an active, readily fusible, acid flux. It
is made by fusing borax in a crucible and pouring the fused mass on
a clean iron or brick surface. It should be crushed to pass a 16 or
20 mesh screen before using. Crystaline borax glass melts at 742° C.
Finely divided amorphous borax glass begins to sinter at 490°-500° C.
It is extremely viscous when melted.
Its rational formula Na,O, 2B.03 indicates an excess of acid and
experiment proves this to be the case. It is an excellent flux for all
the metallic oxides, and fusing as it does at a low temperature, it
helps to facilitate the slagging of the ore. Either borax or borax
glass is used in almost-every flux mixture. It should be used in prefer-
ence to silica as a flux for ime, magnesia, iron, manganese and zinc
oxides.
- Borax glass cannot always be used in preference to borax, as it is
more violent in its action and causes much boiling of the charge.
The fine material especially takes on moisture from the air and tends
to cake. It is sometimes used as a cover on top of a crucible charge, —
especially in muffle fusions. Here it melts before the rest of the charge
and prevents loss of the fine ore by “dusting.”
Sodium bicarbonate, NaHCOs, is the most common of the alkali
carbonate fluxes. It is decomposed on heating to 276° C. forming the
normal carbonate as shown by the following reaction :—
2NaHCO; = NaeCO; + H,O + COz
Pre
3
It acts as a basic flux, a desulphurizing agent and in some cases
as an oxidizing agent. It is the cheapest of the alkali carbonates,
and, as it is readily obtained pure, and does not deliquesce, it is pre-
ferred by many to the normal carbonate.
Sodium carbonate, Na.COs, (normal carbonate) melts at 852° C.
: It occurs crystallized, (sal-soda) containing more than half its weight
of water, and in this form, is not at all suited for a flux. In the
' anhydrous form, it is somewhat stronger, weight for weight, than the
bicarbonate, but unfortunately it tends to absorb water from the air
| and is therefore unsatisfactory for use in some climates. The com-
mercial normal carbonate of this country is made by the Solvay pro-
cess from the bicarbonate and should be very pure. The variety
known by the trade as, 58% dense, soda-ash has been found entirely
satisfactory for assay purposes and is but little affected by atmospheric
moisture.
It acts as a basic flux, a desulphurizing agent, and in some cases, an
oxidizing agent. Heated in the presence of silica, it breaks up giving
off CO, and the Na,O combines with the silica forming silicates, as
for example :— |
NasCO; a SiOs = NasSi0; a CO.
| Many of these silicates are readily fusible and for this reason soda
is used in nearly every fusion.
Potassium carbonate, K,COs, fused at 894° C. and is also a basie
flux. Its action is in all ways similar to sodium carbonate.
A mixture of sodium and potassium carbonates fuses at a lower
temperature than either one alone, due to the formation of a double
salt and therefore the mixture is used whenever it is desired to main-
tain a low temperature during the assay. The lead assay is a case in
point. 7
Potassium carbonate should be kept in an air tight receptacle as
otherwise it takes on water from the air and forms a hard cake.
Litharge, PbO, (92.83% lead) is a readily fusible basic flux. It
~ acts also as an oxidizing and desulphurizing agent and on being re-
duced it supplies the lead necessary for the collection of the gold and
silver. It melts at 883°C. (Mostowitsch).
- Litharge begins to combine with silica at 600° C. forming lead
silicates which are pasty at this temperature.
2PbO + $10. = Pb.SiO, (monosilicate). This silicate is fusible
Fa ipl
—
+
at a low temperature 746° C. and is as fluid as water. If the pro-
portion of silica be raised above that of the tri-silicate, the mixture
becomes less easily fusible and is decidedly viscous when fused.
Litharge has such a strong affinity for silica that if enough is not sup-
plied to it in the crucible charge, it will attack the acid material of the
crucible itself and if left long enough will eat a hole through it.
Litharge readily gives up its oxygen if heated with carbon, carbonic
oxide, hydrogen, sulphur, metallic sulphides, iron, ete. The reaction
with carbon begins at about 550°C. It thus acts as an oxidizing,
and in the presence of sulphur, as a desulphurizing agent :—
2PbO + C = CO, + 2Pb (oxidizing)
3PbO + ZnS = ZnO + SO. + 3Pb (desulphurizing and oxidiz-
ing) .
The liberated lead is then available for the collection of the gold and
silver.
Lead silicates do not readily give up their lead to carbonaceous and
sulphurous reducing agents. The higher the proportion of silica, the
less readily is the silicate broken up. In order to extract all the lead
it must first be set free by the use of a stronger basic flux. Thus
“metallic iron decomposes all fusible lead silicates at a bright red heat,
provided enough is added to form a singulo-silicate.”” (Hofman).
Ordinary commercial litharge contains a small amount of silver,
varying from 0.2 oz. to 1.0 oz. or over per ton. A practically silver
free variety is made from Missouri lead by giving a zine treatment, as
for the Parkes process and then cupelling. It is never safe to assume,
however, that litharge is silver free until it has been proven so by
assay. Each new lot received should therefore be carefully mixed to
make it uniform and assayed. See page 89.
Lead in the granulated form (test lead) is used in the scorification
assay as a collector of the precious metals and as a flux. When oxid-
ized by the air of the muffle it becomes a basic flux. Ordinary test
lead usually contains more or less silver and every new lot should be
assayed before being used. Lead melts at 326° C.
Test lead may be made by pouring molten lead just above its freez-
ing point into a wooden box and shaking it violently in a horizontal
direction just as it becomes pasty and continuing until it becomes
solid. The fine material is sifted out, the coarse is re-melted.
Argols is a reducing agent and basic flux. It is a crude bitartrate
of potassium obtained from wine barrels. It is one of the best re-
ducing agents. .
a
~
: i 5
a a
5
Cream of tartar, KHC,H,Osg, is refined bitartrate of potassium.
Being free from sulphur it is used as a reducing agent in the copper
assay. Both argols and cream of tartar break up on heating as fol-
lows :-—
2KHC,H,O;, + heat = K,0 + 5H.O + 6CO + 2C
The K;O thus liberated being available as a flux.
Charcoal, sugar, flour etc. are also reducing agents because of
the carbon that they contain. Flour is very commonly used in flux
mixtures and is satisfactory in every respect.
Sulphides, arsenides, antimonides etc. in ores all have a re-
ducing effect.
To determine the reducing power of the different minerals and
reducing agents see chapter on the crucible assay.
Iron is a desulphurizing and reducing agent. When heated with
the sulphides of lead, silver, mercury, bismuth and antimony the
_ sulphides are decomposed yielding a more or less pure metal and iron
sulphide. Copper, nickel and cobalt sulphides are partly reduced by
iron as would be expected by a study of the heat of formation of the
same.
It also reduces most of these metals and some others from their
oxide combinations, as for example :—
PbO + Fe = Pb + FeO
The iron oxide formed acts as a basic flux. Iron decomposes all
fusible lead silicates by replacing the lead, thus:—
; 2PbO.Si0, + 2Fe = 2FeO.S8i0. + 2Pb
It should therefore always be used in the lead assay.
It is used in the form of spikes or nails and sometimes, especially
in Europe, an iron crucible is employed.
Potassium nitrate, KNO;, commonly known as niter is a power-
_ ful oxidizing agent and also a basic flux. It melts at 339° C. and fuses
without alteration at a low temperature, but ata higher temperature
breaks up giving off oxygen which oxidizes sulphur and many of the
' metals, notably lead and copper.
_ It is used in the fire assay especially to oxidize sulphides, arsenides,
antimonides, etc.
In a charge containing a large excess of soda and litharge the reac-
tion with pyrite is as follows:-—
| 6KNO; + 2FeS, + 3Na,CO; = FeO; + 3K.S80O. + 3NaSO. +
= 3CO. + 3Ne
4 In this case one gram of niter would oxidize 0.39 grams of pyrite, or
6
its oxidizing power would be 4.75 taking the reducing power of pyrite
in this type of charge as 12.20.
In a charge containing considerable silica the reaction with pyrite
is about as follows:—
1OKNOs + 4FeS, + 2810, = 2FeSi104 + 5K SO, + 3580, + 5Ne
In this case one gram of niter would oxidize 0.475 grams of pyrite.
Taking the reducing power of pyrite as 9 in this type of charge, the
oxidizing power of niter in terms of lead is 4.27. This latter is more
nearly the figure obtained in practice. As a little oxygen usually
escapes unused and as the commercial article is never absolutely
pure, a figure as low as 4.0 is often found about right for practice.
Niter also reacts with carbon and silica as follows:—
or 1 gram of niter oxidizes 0.15 grams of carbon,
2KNOs; + SiO. = KeSiO; + 50 + Ne
This action begins at about 450° C.
If finely divided lead is fused with niter, the lead is found to be
directly oxidized by the niter. Fulton found the oxidizing power of
niter used in this way to be 2.37. The following reaction shows ap-
proximately what happens:—
7Pb + 6KNO; = 7PbO + 3K,0 + 3Ne + 40,
It should be noted that in this case a considerable portion of the oxygen
escapes unused.
Many assayers object to the use of niter because of its oxidizing
effect on silver. Large amounts of niter cause violent boiling of the
crucible charge and necessitate careful heating to prevent loss. It
is found to give less trouble when the crucible is uniformly heated, as
in themuffle, than when the charge begins to melt first at the bottom,
a3 in the pot furnace. The student should select an extra large cru-
cible and carefully watch the fusion when using more than 20 or 30
grams of niter in any charge.
Potassium cyanide, KCN, is a powerful reducing and desulphur-
izing agent. It combines with oxygen forming potassium cyanate,
thus :—
PbO + KCN = Pb+ KCNO (reducing action)
and also with sulphur, forming sulphocyanide, as follows:—
PbS + KCN = Pb + KSCN
It is sometimes used in the lead assay and usually in the tin and
bismuth assays. It is extremely poisonous, and should be handled
with great care. It fuses at 526° C.
Salt, (sodium chloride) NaCl, melts at 819° C. and is used as a
=
(
cover to exclude the air, and to wash the sides of the crucible, and
prevent small particles of lead from adhering thereto.
It does not enter the slag, but floats on the top of it.
It is often
colored by the different metallic oxides of the charge and sometimes
helps to distinguish assays which have become mixed in pouring.
Fluorspar, CaF», is occasionally used as a flux.
It fuses at a high
temperature, 1400° C., but when melted, it is very fluid and assists in
liquifying the charge.
Cryolite, AlNa;F, is not commonly used by assayers, but it is
sometimes used in melting bullion.
and has the property of dissolving alumina.
Name
Silica
Glass
Borax
Borax glass
Sodium bicarbonate
Sodium carbonate
Potassium carbonate
Litharge
Potassium nitrate
Argols
Cream or tartar
Flour
Charcoal
Lead
Iron
Potassium cyanide
Salt
Fluorspar
Cryolite
TABLE 1.
Formula
S102
xNasO.yCaO.zS8iO2
NazgBsO7.10H2O
KNOs
KHC4sH4O¢ + C
KHC4H40¢
CaFe
AlNasF¢
ae
pears:
Fusion Products.
Every gold, silver or lead assay fusion, if the charge is properly
proportioned and manipulated, should show two products, a lead
button and above it a slag.
It fuses at a low temperature,
ASSAY REAGENTS.
Properties in order of their importance.
Acid flux
Acid flux
Acid flux
Acid flux
Basic flux, desulphurizing
Basic flux, desulphurizing
Basic flux, desulphurizing
Basie flux, desulphurizing, oxidizing
Oxidizing, desulphurizing
Reducing agent, basic flux
Reducing agent, basic flux
Reducing agent
Reducing agent
Collecting agent
Desulphurizing and reducing agent
Reducing and desulphurizing agent
Cover and wash
Neutral flux
Neutral flux
Two undesirable products, matte and
speiss are occasionally also obtained.
The lead button should be bright, soft and malleable and should
separate easily from the slag.
The slag is usually a silicate or borate of the metallic oxides of the
ore and fluxes used.
of undecomposed ore.
It should be homogeneous and free from particles
A good slag should usually be more or less
8
glassy and brittle. When poured, the slag should be thin and fluid
and free from shots of lead. If too acid, it will be quite viscous and
stringy, and the last drops will form a thread in pouring. If too
basic, it will be lumpy and break off short in pouring. When cold,
the neutral or acid slag is glassy and brittle, the basic one is dull and
stony looking.
The slags should never be allowed to get mixed with the fuel, as they
quickly destroy the furnace lining.
Matte, is an artificial sulphide of one or more of the metals, formed
in the dry way. In as aying it is most often encountered in the niter
fusion of sulphide ores when the charge is too acid. It is found lying
just above the lead button. It is usually blue gray in color, approach- -
ing galena in composition and is very brittle. It may be in a layer
of considerable thickness, or may appear simply as a granular coating
on the upper surface of the lead button. This matte always carries
some of the gold and silver and as it is brittle, it is usually broken off
and lost in the slag, in the cleaning of the lead button. The student
should examine the lead button as soon as it is broken from the slag,
and if any matte is found, he may be certain that his charge or furnace
manipulations are wrong.
Speiss, is an artificial, metallic arsenide or antimonide formed in
smelting operations. As obtained in the fire assay, it is usually an
arsenide of iron approaching the composition of Fe;As. Occasionally
the iron may be replaced by nickel or cobalt. The antimony speiss
is very rare. In assaying, speiss is obtained when the iron method is
used on ores containing arsenic. It is a hard, fairly tough, tin white
substance found directly on top of the lead and need adhering
tenaciously to it.
If only a small amount of arsenic is present in the ore, “ake Speiss.
will appear as a little button lying on top of the lead, if much arsenic
is present, the speiss will form a layer entirely covering the lead. It
carries some gold and silver. If only a gram or so in weight, it may be
put into the cupel with the lead and will be oxidized there, giving up
its precious metal values to the lead bath. A large amount of speiss
is very hard to deal with as it is difficult to scorify. The best way is
to repeat the assay using some other method.
e
a a
CHAPTER II.
FURNACES AND ASSAY SUPPLIES.
Furnaces for assaying may be divided into the two following
classes :—
1. Crucible or Pot-Furnaces. These are furnaces used solely
for melting purposes in which the crucible is in direct contact with the
fuel or flame and the contents therefore more or less subject to the
action of the products of combustion.
2. Muffle-Furnaces. These: are furnaces in which the charge
to be heated is in a space (the muffle) apart from the fuel or products
of combustion. The muffle is a semi-cylindrical receptacle of fire-clay
or other refractory material set horizontally and so arranged that the
fuel or products of combustion pass around and under it. Thus the
material to be heated is entirely separated from the products of
combustion.
As muffle furnaces may be used for melting purposes as well as for
scorification and cupellation, many assayers in America use this type
of furnace exclusively, especially in connection with soft-coal fuel.
The advantages of muffle furnaces for melting are the greater ease and
saving of time in charging and pouring, the better control of temper- |
ature and the better distribution of heat for melting purposes. Cru-
cibles also seem to stand more heats in a muffle furnace than they will
in pot furnaces, due no doubt to the slower and more uniform heating.
Pot-furnaces have the advantage of size, so that for instance in
dealing with low-grade ores a larger charge and crucible may be used
than in the ordinary size muffle furnaces. A higher temperature may
be obtained in pot-furnaces than in muffles and this occasionally is
an advantage of the pot-furnace.
The furnaces themselves are made of fire-brick or fire-clay tile and
“may be set in an iron jacket or surrounded by common red brick.
Fire-brick is best laid in a mortar made from a mixture of two parts
ground fire-brick and one part fire-clay. Sometimes a small amount
of Portland cement is added. In any event the brick and tiles should
be thoroughly wet previous to applying the mortar. Finally, as little °
mortar as possible should be used since the bricks are much harder
than the solidified mortar.
10
Assay furnaces are made to burn practically all kinds of gaseous,
liquid and solid fuels. Those most commonly used are natural and
artificial gas, gasoline, kerosene, crude-oil, wood, charcoal, coke,
bituminous and anthracite coal. |
Gas is the cleanest, most easily controlled, most efficient in com-
bustion and except in the case of a natural supply the most expensive
fuel. A blower is usually required to supply air under a low pressure
with this method of firing.
Oil is nearly as clean and as convenient to use as gas, the efficiency”
of combustion is high and in localities near the oil-fields it may be
very cheaply obtained. The calorific power of the hydro-carbon
fuel oils is high, about 50% more than the best coals, which makes
them particularly suited for use in isolated localities where freight
charges are high. Gasoline is forced under pressure through a heated
burner where it is vaporized and the gas injected into the furnace
carries with it a sufficient supply of air for combustion. Crude-oil
requires steam or air under pressure to aid in atomizing the oil pre-
liminary to proper combustion. Gasoline, kerosene and crude
petroleum all have a heating value of about 21,000 B.T.U. per pound.
Solid fuels are usually the cheapest and are therefore more ex-
tensively used than any of the others. In isolated districts where
coal or coke is not available wood is occasionally used as fuel in assay
furnaces. For this purpose it should be felled in winter and thor-
oughly air dried for at least six months or longer according to the
climate. The air-dried wood will still retain from 20 to 25 per cent
of water and in this condition has a heating value of about 6000
B.T.U. per pound. Charcoal is seldom used in this country for assay
purposes on account of the abundant supply of other fuels.
Bituminous-coal is the most satisfactory solid fuel for muffle
furnace firing and coke for pot furnaces. Good soft-coal has a calor-
ific power of about 14,500 B.T.U. per pound, should be low in sulphur
and the ash must not be too readily fusible. Coke should be hard and
strong, low in sulphur and the ash should be infusible at the temper-
ature of the furnace. That is to say it should be high in silica and
alumina and low in iron, calcium, magnesium and the alkalies to
prevent clinkering of the walls of the furnace.
Gas and Oil vs. Solid Fuel. Gaseous and liquid fuels have many
advantages over solid fuels for assay purposes, some of which are as
follows :—
;
ne
:
:
3
ua
1. The fire is kindled in an instant and the furnace may be quickly
heated to the desired temperature for work.
_ 2. The temperature is readily controlled and may be quickly
varied to suit the requirements of the work.
3.4A high efficiency of combustion is possible in properly designed
furnaces and as soon as the work is completed the fuel supply may be
shut off and fuel consumption stopped.
4. The avoidance of labor in firing gives the assayer more time for
other duties. |
5. The cleanliness in operation due to absence of solid fuel and ash
is obviously a great advantage in any analytical laboratory.
On account of the expense however coal is much more generally
used than either oil or gas. It is easy to make a comparison of the
costs of any of the fuels by considering the heat units. For instance,
with soft-coal at $5.00 per ton and gasoline at 15c¢ per gallon we may
say that one cent invested in soft-coal will buy us 4 x 14,500 =
58,000 B.T.U. and that the same amount invested in gasoline will
bring approximately + x 5.4 X 21,000 = 7560 B.T.U. That is to
say the gasoline is over seven times as expensive as the coal on the
basis of heat units and for steady running this may be taken to be
approximately correct. However for a small amount of work a
gasoline furnace may be cheaper to run even with the cost of fuel as
above assumed, for the small furnace is quickly heated and as soon as
the work is completed the oil supply may be shut off and the expense
stopped, while a coal furnace takes much longer to heat and then must
be allowed to burn out after the work is completed.
Coal Furnaces. This type of furnace is used in most of the large
custom and smelter assay offices in this country.
The furnace may be built either with a tile or fire-brick lining. The
- tile lining is more easily set up but whether or not it is as durable as
a properly constructed fire-brick lining is open to question. The
outside of the furnace is usually laid up with common hard-burned
red brick. If the furnace is to be lined with fire-brick several rows
of headers should be left to hold the lining securely in place. The
furnace is held together with angle-irons, stays and tie-rods.
In the furnace as ordinarily constructed the muffles are supported
by “jamb” bricks projecting from the sides. When these are used
it is well to leave a hole or lose brick on the outside of the furnace
to facilitate removing the stubs when these bricks become broken
off and the ends slagged in. Fulton recommends using long tiles
here which meet in the center, thus giving better support for the
12
muffle. He claims a prolonged life for the muffle with this arrange-
ment. The writer has found the Scotch Gartcraig brick to outlast
3 or 4 best American fire-brick for muffle supports. Another method
of supporting muffles in furnaces of this type is by the use of iron pipes
or castings extending directly across the furnace and through which
cooling water is circluated.
These furnaces occupy a floor space of approximately three and
one-half by four feet. They are built in a variety of sizes, those
taking NN, QQ and UU muffles are the sizes; most commonly used.
The NN muffle is 103 X 19 X 63 inches outside, similarly the QQ
is 124 x 19 x 72 and the UU is 14 X 19 X 7% inches outside,
Each NN muffle will hold twelve—-20 gram or eight-—30 gram cru-
cibles allowing in each case for a row of empty crucibles in front to
act as warmers, while the QQ muffle will hold fifteen—20 gram .or
twelve—30 gram crucibles also allowing for a row of empty crucibles in
front.
The furnaces are best arranged to be fired from the rear although
they may be arranged to be fired from the front or sides. The flue
makes off from near the front of the furnace thus tending to heat the
muffle uniformly throughout its entire length. It should be from
one-sixth to one-eighth the grate area.
The stack for one of the furnaces will need to be at least 20 feet
high and possibly higher depending largely on the character of the
coal. It should not be built directly on the furnace but may be placed
directly over the furnace if supported by arches and cast-iron columns,
or it may be put to one side of the furnace and in this case will extend
down to the ground. When the stack is supported independently
of the furnace it allows the furnace to expand and contract with less
danger of cracking and also permits of tearing down and rebuilding
the furnace without interfering with the stack.
With long-flame coal these furnaces are best fired with a rather
thin bed of fuel say 8 inches. The sequence of firing will consist
of first running the slice bar along the entire length of the grate in one
or two places and lifting up the fire to break up any large cakes and
thus allow free passage of air through the fire, second to push the
well coked coal forward with the hoe and third to add 2 or 3 shovels
of fresh coal near the firing door. As this coal is heated it begins to
coke and the gas given off passes over the white-hot coal of the fire
and is there mixed with heated air. This results in a free draft and
good volume of hot flame. If instead of adding the fresh coal near
the firing door it is spread all over the fire it will quickly cake and
tend to smother the fire by shutting off the draft.
13
The temperature of the muffle may be regulated at will by man-
ipulating the draft and firing doors. For instance, after a batch of
cupels have started the draft may be closed and the firing door opened
thus admitting cold air above the fire which quickly cools the muffles
to any required degree. :
Wood Furnaces. Wood-burning furnaces are made with single
and double-muffles and are much like the soft-coal furnaces except
that a larger fire-box and grate are used. Wood is usually sawed
in half-cord lengths and with dry wood the muffle may be easily
heated sufficiently for assaying. Hard wood is much to be preferred
as it does not burn out as rapidly, but almost any kind of dry wood
may be used.
The large fire-box and the grate which is set about 8 inches below
the bottom of the fire-door are the principal distinguishing character-
istics of a wood burning assay furnace.
Coke Furnaces. Coke is still used to a considerable extent in
pot furnaces but for muffle furnace fuel it is fast falling into disuse,
at least in this country.
Comparing the coke and the soft-coal muffle furnace the coke furnace
has the advantage that it can be more quickly heated to a cupelling
temperature and that it requires less frequent stoking. On the other
hand it is harder to regulate the temperature, especially to cool it
off quickly when cupelling, the stoking is harder work and the fuel
cost per assay is higher in most localities.
The great advantage of the coke pot-furnace is the very high
temperature which may be obtained and the fact that even though
the crucibles boil over or eat through no harm is done to the furnace.
Coke furnaces especially should be supplied with a good quality of
fuel. If the ash tends to melt the walls become quickly covered with
elinkers and are bound to be more or less damaged when these are
removed.
Gasoline Furnaces. A gasoline furnace outfit consists of a
furnace which may be either a muffle, crucible or combination of
the two, a burner with piping etc. and a gasoline tank with pump.
The latter for a small assay office consists of an ordinary tinned-
steel pressure tank equipped with a hand pump, pressure-gauge and
the necessary piping connections. These range from 2 to 15 gallon
capacity.
The burners are usually constructed of special bronze alloys cap-
able of withstanding oxidation at high temperatures. They consist
14
of a filtering chamber for purifying the gasoline, a generating chamber
where the gasoline is vaporized, a generating pan and valve for the
initial heating of the burner, a spraying nozzle and valve through
which the gasoline vapor is injected into the furnace and a,mixing
chamber where the proper amount of air for combustion is mixed
with the gas. From the filter the gasoline passes around the interior
of the burner face (the generating chamber), where it is heated by
the radiated heat from the furnace and vaporized so that once the fur-
nace is under way the generating burner may be shut off. Gasoline
is supplied to the burner under a pressure of from 20 to 50 pounds per
square inch.
The great feature to be sought and one of the hardest to attain
in any gasoline furnace is an even distribution of heat. Another
feature found wanting in many gas and gasoline furnaces is the poor
draft through the muffle. Owing to the fact that the pressure in-
side the furnace is slightly greater than the outside pressure there
is a great tendency for the products of combustion to work back
through the hole in the rear of the muffle, thus to a large extent ex-
cluding the air and unduly prolonging cupellation or scorification.
In operating a gasoline burner care should be taken to see that
combustion takes place only in the furnace. All burners have more
or less tendency to back-fire, that is for the flame to jump back and
continue in the mixing chamber. If this is allowed to continue the
burner gets so hot that the metal oxidizes and then it is only a matter
of a short time before it is entirely destroyed. Every furnace should
be provided with a shut-off valve between the burner and the gasoline
tank. When it is desired to shut off the furnace, close this shut-off
valve letting the burner continue as long as any pressure is left and
do not ever entirely close the burner valves. The valve stem or
needle is of steel and the seat is of bronze and owing to the different
rates of expansion of these metals the valve is injured if these are
left in close contact when the burner is cooling. This precaution is
especially to be observed when the burner is provided with the ordin-
ary needle valve, as when this valve is once enlarged the whole effi-
ciency of the burner is destroyed.
Gas Furnaces. Gas furnaces are used in some assay Offices,
especially where a natural gas supply is available. Where artificial
gas has to be used this type of furnace proves decidedly expensive
if used for any considerable amount of work. As the gas is usually
not under sufficient pressure to carry in its own supply of air for
combustion, these furnaces are customarily supplied with air from @
15
blower, which adds to the expense and difficulty of the furnace
operation.
Crude Oil Furnaces. When a cheap oil supply is available
crude oil is frequently used as an assay fuel. The furnaces themselves
are built exactly as for gasoline firing.
Crude-oil and kerosene cannot be vaporized in the burner as they
deposit carbon when heated and thus clog the tubes. Consequently
to insure complete combustion the oil must be thrown into the furnace
in as fine a state of mechanical subdivision as possible. This is
accomplished by atomizing the oil with a jet of steam or air. The
steam may be from an outside source or may be generated in the face
of the burner from the heat reflected: from the furnace. The Marvel
burner made by the Braun Corporation of Los Angeles uses the latter
method of supplying steam. This burner requires to be heated by |
a torch or alcohol lamp to start and the necessary outfit includes two
pressure tanks one for oil and one for water. Oil of any gravity up
to 60°, Beaume may be used with this burner.
Furnace Repairs.
Fire-clay usually forms the basis of mortars used in furnace con-
struction and repairs, as lime mortar and hydraulic cement are not
suited for use with masonary exposed to high temperatures. Fire-
clay is a clay containing only very small amounts of iron, lime, mag-
nesia and the alkali oxides. It forms a more or less plastic and sticky
mortar and on heating loses its moisture and plasticity and the mortar
hardens.
All clays shrink more or less on drying and burning and to prevent
this as far as possible in the mortar as well as to make it strong a
certain amount of crushed fire-brick or sand should be added.
Crushed fire-brick is better than sand owing to its porous and ir-
regular shaped. grains as these give a better mixture with the clay
and a stronger cement.
A good mortar for general use around assay furnaces is made with
a mixture of two parts ground fire-brick through 12 mesh and one
part fire-clay.. A small amount of Portland cement or molding clay,
say not over one-third part, will make the mixture adhere better and
the mortar will be harder when set. For work at very high tempera-
a tures the Portland cement must be omitted as it acts as a flux for the
other materials and causes the whole to melt.
All mortars should be made up dry and thoroughly mixed before
4 - the required amount of water is added. The water should be thor-
16
oughly mixed in and the mortar should be sticky and of the right
consistency. It is well to mix the mortar several hours before using.
When laying bricks or making repairs about a furnace the bricks
and-brick work should be thoroughly wet before applying the mortar
as otherwise the bricks absorb so much water that the mortar does not
form a good bond with them.
In laying fire-bricks as little mortar as possible should be used as
the bricks are always harder than even the best of mortar. The mor-
tar should be made to fill every crevice. The best way to attain this
is to put an extra amount of fairly thin mortar on the wet brick and
then drive or force it firmly into place, allowing the excess mortar to
squeeze out.
The ash from many coals is quite readily fusible and results in the
formation of clinkers and accretions on the sides of the furnace,
especially just above the grate. When the furnace is cold these
adhere very tenaciously to the walls of the furnace and in breaking
them off, pieces of the brick are removed with them. To remove these
accretions with the least damage to the furnace they should be cut
off with a chisel bar just after a hot fire has been drawn.
In putting in a new muffle, first remove the old one with the mortar
that held it, also any clinkers which would interfere with the working
of the furnace. Patch the lining of the furnace if it requires it and
see that the bricks or other supports for the muffle are in place and
in good condition. After trying the muffle to see that it rests properly
on the supports, remove it, sponge over the brick work where the mor-
tar is to come in contact with it, place some rather thick mortar on
each of the supports and replace the muffle. See that it rests evenly
on the different supports and on the front wall of the furnace. The
muffle should be level horizontally and slope slightly toward the
front end. Fill up the space between the muffle and the front wall
of the furnace with some rather thick mortar, working from both
inside and outside of the furnace. This outside joint should be finished
up neatly with the aid of a trowel. It is best to allow the furnace to
dry for a day or two if possible, but if necessary it may be used as
soon as finished by heating up slowly.
For patching the linings of furnaces use the mixture recommended
for general use or try the following which is recommended by Lodge.
Fire-brick through 12 mesh 7 parts, Portland cement 2 parts, fire-
clay 1 part. Put this on as dry as possible and it will make a patch
almost as hard as the original brick.
Cracked and broken muffles may be made to last much longer if
patched with some of the following mixtures.
ee ee a ea ee ey
17
~ When the bottom is almost gone Lodge recommends a mixture of
2 parts Portland cement, 1 part ground fire-brick, 4 to $ part fire-
clay. For patching holes he recommends a mixture of glass, sand and
clay to which a little litharge has been added. After heating this
becomes as hard as the muffle. A mixture of short fibered asbestos
and silicate of soda is also recommended.
Metallurgical Clay Goods.
Under the caption metallurgical clay goods are included muffles,
crucibles, scorifiers, roasting dishes, annealing cups ete. These
embrace many of the most important utensils of the assayer and
upon their good properties much of his success depends. Fire-clay
is the only material which answers the double purpose of satisfactory
service and inexpensive construction. Refractory clay or fire-clay
as it is commonly called is a clay which will stand exposure to a high
temperature without melting or becoming in a sensible degree soft or
plastic.
All clays contract both upon drying and burning and this leads to
more or less warping and cracking of the finished product. ‘To pre-
vent this shrinkage as far as possible and also to add strength to the
finished article it is customary to add a certain amount of sand or
well burned clay to the mixture. Burned clay is usually preferred
to sand for this purpose, not only because its rough porous grains
give a better bond with the fire-clay and make a stronger cement, but
it also makes an article which is less readily corroded by assay slags
and fusion products. The intermixture of coarse grains of burned
clay helps also in that it makes a product better able to withstand
sudden changes in temperature.
The exact proportions of raw and burned clay used by any manu-
facturer are carefully guarded trade secrets and depend of course
very much on the clay used as well as upon the article to be manu-
factured. The larger the article the more is the care which must be
taken to prevent warping and cracking. Usually however, the pro-
portion of raw to burned clay will lie between the limits of one to one
and one to two.
Muffles. Muffles may be made of a variety of materials but for
assay purposes fire-clay muffles are used exclusively. They are
made in a great variety of sizes and shapes.
Muffles as well as other fire-clay ware should be stored in a warm,
dry place and should be heated and cooled slowly and uniformly for
18
maximum service. The life of a muffle is also much influenced by
the manner of supporting.
Crucibles. Assay crucibles are made either of a mixture of raw
and burned clay or of a mixture of sand and clay, the first being known
as clay or fluxing crucibles and the second as sand crucibles. The
raw clay is finely ground, mixed with the right proportion of coarser
particles of sand or burned clay and water and the whole well kneaded
and compressed in molds of the proper shape.
Good crucibles should have the following properties:—
. Ability to withstand a high temperature without softening.
. Strength to stand handling and shipping without breaking.
. Ability to stand sudden changes of temperature without cracking.
. Ability to withstand the chemical action of the substances fused in
them.
5. Impermeability to the substances fused in them and to the products
of combustion.
me Wh Ke
Of course it is impossible to get any one crucible which will possess
all of the above good properties to a high degree. For instance if a
crucible is to be made as nearly impermeable as possible it will be
made of very fine grained material and tightly compressed. Such
a crucible however will not stand handling or sudden changes of tem-
perature as well as one made with a skeleton of coarser material.
Furthermore the manner and temperature of burning has much to
do with the way that crucibles will stand handling and shipping.
A fairly hard burned crucible will be stronger and less likely to be
broken in handling but on the other hand it will not stand sudden
changes of temperature as well as a soft-burned crucible. Crucibles
made of elay containing little uncombined silica and of burned elay
of the same nature will stand a high temperature and chemical cor-
rosion much better than those made of sand and clay or of elay eon-
taining much free silica.
Crucibles are tested for resistance to chemical corrosion by actual
service and also by fusing litharge in them and noting the time it
takes to eat through. To make a test of this sort which is of any value
care must be taken to see that the temperature, the quantity of
litharge and all other conditions are the same for the crucibles being
tested. A crucible may be tested for its permeability to liquids by
filling it with water and noting the time it takes before it becomes
moist on the outside.
Crucibles come in a great variety of shapes and sizes. Those most
19
commonly used for assaying may be classified into two groups as —
follows :—
Pot Furnace Crucibles. These are comparatively slim, heavy
walled crucibles with practically no limit as to height. The base is
small so that they may be forced down into the fuel and for this reason
they are easily tipped over and are not suitable for muffle work.
The sizes most used are the E, F, G, H, J and K. Crucibles of the
same designation made by different manufacturers vary considerably
in capacity. The approximate capacity of some of the pot-furnace
crucibles is shown in the following table:—
TABLE II. CAPACITIES OF POT FURNACE CRUCIBLES.
Crucible designation E F G H I J kK
1 Battersea 180cc | 210ce | 300ce | 420cec - 600cce | 750ce
2 Denver 180ce | 240ce | 400cec —% 530ce 685ee | 950ce
\ Made by the Morgan Crucible Co. London, England.
2 Made by the Denver Fire Clay Co. Denver, Colorado.
Muffle Crucibles. These are made with a broader base so that
they may stand securely on the floor of the muffle and are usually
not more than four inches high. Muffle crucibles are designated by
gram capacity, the 10, 15, 20 and 30 gram sizes being most frequently
used. The intention of the system is that the numbers indicated
the grams of ore charge which the crucibles will take. They are
usually generously proportioned so that often an assay ton of ore
(29.166 grams) may be treated in a 20 gram crucible.
The approximate capacity of the more important muffle crucibles is
shown in the. following table:—
TABLE III. CAPACITY OF MUFFLE FURNACE CRUCIBLES.
Crucible Designation Sem.) 10 ems) 12. gm. | 15 gm. | 20 gm. | 30 gm.
2) | ee, ee en
Denver 70cec | 100ce | 140ce | 160cce | 190ce | 260ce
Battersea 70ce. | 100ce - 135ee | 190ce | 260ec
Scorifiers. These are shallow fire-clay dishes used in the scorifi-
cation assay of gold and silver ores. They should be smooth on the
inside, dense and impermeable to lead and slag and should be composed
so as to withstand as much as possible the corrosive action of litharge.
Scorifiers are designated by their outside diameters. Of the large
20
number of sizes made the following are the most commonly used:
24/7, 24/", 22", 3’’, 34’".. The Bartlett scorifier is shallower than the
regular one and was designed for the treatment of heavy sulphide
ores containing considerable metallic impurities. Scorifiers particu-
larly should be made of clay containing a minimum of uncombined
silica, as the scorifier slags are usually very basic. Particularly
when they contain copper they attack the silica of a scorifier with
avidity and one with a siliceous skeleton may become perforated
and allow its contents to escape onto the floor of the muffle, thus
spoiling the assay and injuring the muffle.
2. + oe ee ee
yore ae
Ee ee Ee ae ee ee Le ee me ee
CHAPTER III.
SAMPLING.
Definition. A sample is a small amount which contains all the
components in the same proportions as they occur in the original lot.
The object of sampling an ore is to obtain for chemical or mechan-
ical tests a small amount which shall contain all the minerals in the
same proportion as they occur in the original lot. In the subsequent-
discussion the word “sample” will be taken to mean that fraction
which is taken to represent the whole, whether or not it does so. The
compound words correct-sample, representative-sample, true-sample,
will be used to represent the ideal conditions.
In the intelligent operation of a mine or metallurgical plant, it 1s
necessary to sample and assay continually. In most mines, the differ-
ent faces of ore are sampled every week, sometimes every day. In
concentrating plants, it is customary to sample the products of every
machine to ascertain whether the machine is doing the work expected
of it. In smelters, every lot of ore, as well as fluxes and fuels, have
to be sampled and assayed in order to calculate a charge which will
run properly in the furnace. The slag, flue dust and metallic products
must also be sampled and assayed in order to maintain control of
the operations. In lixiviation plants, the ore and tailings as well
as the solutions must be sampled in order to control and check the
daily work of the plant. In fact, careful sampling and assaying can
not be disregarded, and is becoming more and more important every
day as the grade of ore decreases and the margin of profit becomes
less.
The assayer will usually have the major part of the sampling done
for him, but he is expected to know how to do it when called upon.
He will usually have only to prepare the final sample, but will oc-
casionally receive lots of 10 to 100 or more pounds to assay in which
case he will have do to his own sampling. The following discussion
will deal principally with the assay laboratory problems of sampling
and the questions of mine and mill methods will be omitted.
22
Labelling Samples. Every lot of ore coming into an assay office,
laboratory, custom mill or smelter should be given a lot number which
should never be repeated, and should be immediately labelled with
this number. A record book should be kept for this purpose, should
show the number of the sample, date of receipt, name of mine, com-~
pany or individual from whom received, the gross and net weight, as
well as notes on the general mineral character, ete., ete.
Moisture Sample. Assays and chemical determinations are
always made on dry samples and the value of a lot of ore is always
figured on the moisture free basis. Exeept in cases when the entire
lot may be dried, it is necessary to take a sample from which to de-
termine the moisture. This sample must be taken as quickly as
possible after the ore is weighed. If the ore may be quickly crushed
and sampled to a small amount of 12 or 14 mesh material, the mois-
ture sample should be taken from this and put in a closely covered
pail or box. Duplicate samples of one or more kilograms of this are
weighed out into a porcelain or enamelled iron dish and dried at 110° C.
The loss of weight being called moisture. As the sample is handled
more than the reject, it loses some moist ire enroute and a constant
should be added to compensate for this difference. Brunton’ finds
10 per cent in summer and 7 per cent in winter a fair average figure.
For instance, if the sample showed 5 per cent moisture for a lot of
ore shipped during the summer months a fair figure for the actual
moisture content would be 5.5 per cent. |
Operations.
Ore sampling may usually be considered to consist of three distinct
operations repeated as many times as necessary. These operations
are Ist, crushing; 2nd, mixing; 3rd, cutting. After the cutting we
have a sample and a reject. The sample may be further reduced by
a repetition of the three operations until it has reached the desired
bulk.
The whole science of ore sampling depends primarily on a eorrect
knowledge of the proper relation between the maximum size of the
ore particles and the weight of the sample taken. The problem to
be solved in each case is something as follows:—having crushed a
particular ore to a certain size (say 10 mesh), how small a sample is
it safe to take from this and still keep within the limit of error allow-
'T. A.M. EH, XL) pg: 5672 (1909)
.
;
“4
4
able? It is necessary to know thé ore, the limit of error allowable,
and the mathematical principles involved.
Sampling is classed as hand sampling when the mixing and cutting
down is done by men with shovels and as machine sampling when
done by some form of automatic machine.
Crushing. All the of ore, unless already fine enough, is broken
or crushed to pass some limiting size screen. This size depends upon
the value of the ore and other factors to be considered later. The
finer the pulp is crushéd the more uniform in size are the particles
and more thorough mixing and better sampling is possible. If the
ore is to be smelted, most of it should be left in the coarse state as
fine ore is undesirable. If it is to be roasted or leached, on the other
hand, fine ore is not objectionable, and the first crushing may be
carried further. Asa rule, however, the aim is to minimize the crush-
ing, thus saving in cost and keeping down the dust.
Machines for crushing should be rapid in action and capable of
easy cleaning. Jaw breakers and rolls fulfil these requirements, ball
mills and pebble mills do not.
Mixing. This step in the process of sampling is often omitted or
allowed to be taken care of itself. It is a necessary forerunner of
quartering and channeling, but is usually omitted before the other
methods of cutting. Especially in the handling of small lots of ore
in the laboratory, it is best to be over careful in this particular rather
than the reverse, and, as it adds but little labor, to give each lot of
crushed ore a thorough mixing before cutting.
The four following methods are used in assay office sampling, some
being better suited for large lots and some for small lots.
1. Coning. The sample is shovelled into a conical pile, each
shovelful being thrown upon the apex of the cone so that it will run
down evenly all around. In mixing a large lot of ore by coning, it
is first dumped in a circle and then coned by one or more men who walk
slowly around between the cone and the circle of ore. The best
results are obtained by coning around a rod, as by this means the
center of the cone is kept in a vertical line. Coning does not thor-
oughly mix an oré, but rather sorts it into fine material which les
near the center and coarser which rolls down the sides of the cone.
If the ore is practically uniform in size, and specific gravity, the mix-
ing may be more thorough. A slight dampening of the ore is said to
allow of better mixing by coning. The floor for this and other hand
24
sampling operations should be smooth and free from cracks which -
would make good cleaning difficult or impossible. A floor made from
sheet iron or steel plates is preferable.
2. Rolling. For lots of 200 pounds or less the method of mixing
whereby the ore is rolled on canvas, rubber sheeting or paper is often
used. When the ore particles are fairly uniform in size and specific
gravity, this method is satisfactory, but for ordinary ores in the coarse
state, it should be avoided. For ore crushed so fine that it has little
or no tendency to stratify, as for example the assay pulp ground to
100 or 120 mesh, the method has been found satisfactory when the
operation is properly performed. ‘This method is almost universally
used by assayers for mixing the final lot of pulverized ore just before
taking out the assay portion. —
3. Pouring. For small samples the method of pouring from one
pan into another is sometimes employed, especially as a preliminary
to rife cutting. Like the two above, it is imperfect when performed
on ordinary coarse and fine ore mixed.
4. Sifting. For mixing small lots of ore or fluxes the method of
sifting is particularly good. The apertures in the sieve should be
two or three times as large as the largest particles. The ore should
be placed on the sieve a little at a time and allowed to fall undisturbed
into a flat receiving pan until all the ore has passed the sieve. Two
or three siftings are equivalent to 100 rollings. Sifting has the further
advantage over all the other methods that all lumps are broken up
and the ore composing them distributed.
Cutting. The final step in the sequence of sampling operations
consists in taking out a fraction of the whole, say one-quarter or one-
half, in some systematic impartial manner. The part taken out is
called the sample and the operation of taking it is the cutting.
The four following methods of hand cutting are used considerably,
but some of them are giving way to machine sampling methods.
1. Fractional Selection. This is a rough starting method suited
only to large lots of low grade or fairly uniform ore. When the ore
is being taken away from the crusher or shovelled out of cars as the
case may be, every second, third, fifth, tenth shovelful, depending
on the value and uniformity of the ore is taken and placed in a sep-
arate pile which is afterwards cut down by some of the later described
|
:
:
|
:
25
methods. When shovelling the ore, care must be taken that each
shovelful is taken from the floor. In case the ore contains lumps too
large for the shovel, they should be broken and put back on the pile.
The method is open to the serious objection that it is a very simple
matter for a prejudiced party to make the sample either higher or
lower in grade than the average by selection of his shovelful samples.
_ 2. Channeling. This consists in spreading out the crushed and
mixed ore in a flat layer a few inches thick and then taking the sample
out in parallel grooves or channels across the pile’ Often two sets
of channels are made one set at right angles to the other. Channeling
is a slow method, requiring much labor and floor space, and owing to
the coarse pieces which fall into the channels from the sides it is in-
accurate. ‘The method is fast falling into disuse.
3. Quartering. This is the method of cutting which accompanies
coning. It presupposes a thorough mixing by coning, as the two
always go together.
When the cone is completed, it is worked down into the form of a
flat truncated cone by men who walk around and around drawing
their shovels from center to perifery, or by starting at the apex and
working the shovel up and down in the path of a spiral. The point
to be observed here is not to disturb the radial distribution of the
coarse and fine ore. After flattening, the cone is divided into four
90 degrees sectors or quarters by means of a sharp edged board, or
better by a steel bladed quarterer. These quarters should of course
radiate from the position of the center of the original cone. Two
opposite quarters are taken out and rejected and the two others are
then taken for the sample. This sample may be again mixed by
coning and quartered, or crushed, coned and quartered as the case
may require.
When properly carried out the method may be made to yield fairly
accurate results, but at best it is a slow and tedious process, and re-
quires the most conscientious work on the part of the laborers to
insure correct results. It is open to the objection that it affords
opportunity for manipulation of the sample by dishonest operators.
Coning and quartering is the old Cornish method of ore sampling
and was almost universally used 30 years ago. It is still somewhat
used as a finishing method at sampling works and by engineers in
the field where no machinery is available.
4. Riffle Cutting. Riffle cutting or splitting is the most accurate
laboratory method available. The riffle, splitter or split-shovel
26
consists of a number of parallel troughs with open spaces between
them, the spaces being usually the same width as the troughs. These
troughs are rigidly fastened and the whole is made into the form of a
shovel, called a split shovel. The ore is taken up on a flat shovel or
special pan and spread over the troughs, care being taken to prevent
heaping the ore above the troughs. Either the ore which falls in
the troughs or that which falls between them may be taken as the
sample. The cutting may be repeated as many times as is deemed
desirable. For the best results in cutting any sample of ore by this
method, care should be taken to have only a thin stream of ore fall-
ing from the pouring pan and to move this pouring pan back and
forth over the split shovel in a horizontal direction perpendicular
to the riffles, so that every part of the stream of ore is being directed
alternately and rapidly first into the sample and then into the reject.
The more irregular in size, specific gravity and value are the minerals,
the greater the care which should be taken in this particular. The
sample should be mixed before re-cutting.
A modification of the riffle or split-shovel known as the Jones
Sampler. or simply as a ‘“‘splitter’’ has recently come into use. It is
a riffle sampler in which the bottoms of ‘he riffles are steeply inclined,
first in one direction and then in the o.her. The ore is spread over
the riffles in the Jones Sampler exactly as over the split-shovel. The
ore 1s caught in two pans placed underneath the splitter.
Riffle cutting is the most rapid method of hand sampling and is
also the most accurate. It is used as a finishing method in most
modern sampling works. One objection to the Jones Sampler and
other similar models is the considerable amount of fine ore-dust which
may be lost due to the greater length of fall of the ore before coming
to rest. One way to obviate this would be to slightly moisten the
thoroughly mixed ore before cutting.
In selecting a split-shovel or riffle cutter for any particular sampling
operation, care should be taken that the distance between the riffles
be at least three times the diameter of the maximum particle of ore.
It is found that a slight bridging action may occur if this precaution
is not observed.
Machine Cutting. A large number of machines have been de-
vised to take the place of the slow laborious methods of hand sampling.
They all depend on taking the sample from a stream of falling ore.
All these devices for sampling fall either under the head of continuous
or intermittent samplers. The continuous samplers take part of the
stream all the time by placing a partition in the falling stream of ore
27
to separate sample from reject. The intermittent samplers, as the
name implies, deflect the entire stream at intervals to make the
sample.
The continuous method of sampling is open to the objection that
it is impossible to get a stream of falling ore containing coarse and
fine particles which is uniform across its entire section. Therefore,
any continuously taken sample (except possibly one-half of the stream)
will be either richer or poorer than the average. Because of these
conditions this type of sampler will not give uniformly reliable re-
sults, and is now but little used.
The intermittent method of sampling gives better results. The
machine should be so designed that it takes equal portions all across
the stream at regular intervals. While it is not possible to produce a
stream of ore which is uniform in value throughout its entire length,
yet by taking a large number of small samples entirely across the
stream the average thus obtained gives a good representative sample
of the entire lot. It is essential that the percentage of sample taken
from all parts of the delivery pipe be the same, in other words that the
vertical sample section, taken in a direction parallel to the motion of
the intake-spout should be a rhomboid.
Three machines of this type have come into general use; these
are the Brunton, the Vezin and the Charles Snyder. The Brunton
oscillates in a vertical plane through an arc of 120 degrees and by a
change of gears any proportion of the stream from 5 per cent to 20
per cent may be taken. The Vezin and Charles Snyder machines
have sector shaped sample cutters radiating from a shaft around which
they revolve.
Hand and Machine Sampling Compared. In comparing hand
and machine sampling it may be said that machine sampling is gen-
erally cheaper and with a properly designed machine it is more ac-
curate than coning or fractional selection. Perhaps the most im-
portant advantage of all is that being strictly mechanical in operation
it affords less opportunity for manipulation of the sample.
Precaution to be Observed. Besides the danger of ‘‘salting”’
from crushing machines, elevators, sampling machines etc. special
attention must be paid to the disposition of the fine ore dust. Asa
rule the rich minerals in the ore are more brittle than the gangue
with the result that the ore dust is far higher in grade than the average
of the ore. Whence is seen the necessity of preserving all of the ore
dust and of taking pains to see that the sample contains its proper
proportion of the same J
28
Theoretical Considerations. The most certain method of
obtaining a representative sample of a lot of ore would be to crush
the whole to 100, 120 mesh or finer, mix it thoroughly and then cut
down by one of the methods just described. This method can be
followed for small amounts of a pound or so, but in the ease of large
lots, it would entail too much labor and would usually unfit the ore
for future treatment. The method generally adopted is a compromise
and consists in crushing the whole lot to a certain predetermined
maximum size and then taking out a certain fraction as a sample.
This sample is again crushed to a smaller size and cut down as
before and this process repeated until finally the assay sample is
obtained.
The care which is required in sampling as well as the size to which
a lot of ore or other material must be crushed before a sample is taken
depends upon the value and uniformity of composition of the material.
The more uniform it is, the smaller may be the sample taken after crush-
ing to any particular size. For instance, if we have a solid piece of
galena containing silver uniformly distributed as an isomorphous silver
sulphide, we can break off a piece anywhere, crush it and have for
assay a lot of ore which is truly a sample of the piece. If, however,
our specimen is not solid galena, but is made up of galena and lime-
stone, the silver value still being contained in the galena, we will
have to crush the whole lot to a uniformly fine size before taking out
a fractional part for a sample. Furthermore it will readily be seen
that the greater the difference in the grade of the different minerals
in the ore, the finer must a lot of ore be crushed before a certain sized
sample should be taken from it.
Since ores are never perfectly uniform in composition a certain
amount of crushing is evidently necessary in every case. To determine
the amount of crushing we must first consider the commercial side
of the question, that 1s we must determine how far it will pay to go
with the process. Evidently a mistake of 1 per cent in the iron
contents of a car load of iron ore worth say $3.00 a ton is less serious
than the same percentage error in the copper contents of a car of cop-
per ore worth say $50.00 a ton. ‘Therefore it may be seen that it
will pay the seller or buyer of the copper ore to go to more pains and
expense in the sampling of the ore than if he were dealing with the
less expensive iron ore. |
Varying Relation of Size of Sample to Maximum Particle.
The variation of any portion of a lot of ore from the average com-
position of the whole is due to the excess or deficit of one or more
4
:
;
)
;
|
‘
;
7
| :
29
particles. The effect upon the results will be greatest when the piece
or pieces which are in excess or deficit are of the largest size, greatest
specific gravity and greatest variation in quality from the average.
Disregarding for the moment the last two of these factors and sup-
posing the ore particles to be approximately uniform in size it is evi-
dent that the sample must contain enough particles so that one ad-
ditional particle of the richest mineral would practically cause no
variation in the value. This means that the sample of the ordinary
ore must contain a very large number of particles perhaps 500,000.
Having determined how many particles of the ore it is necessary
to include in the sample, and assuming the different minerals to be
entirely detached from one another, it would be fair to take such a
weight of ore after each reduction as would contain this established
number of particles. Or as the weight of a lump is proportional to
the cube of its diameter we may state the rule as follows:—Make the
weight taken for the sample proportional to the cube of the diameter
- of the largest particle of the ore.
In the ordinary ore, however, the different minerals are not entirely
detached from each other, but approach more and more to this con-
dition as the size of the ore is reduced. Hence a fixed number of the
particles of the fine ore is less likely to be a true average of the whole
than the same number of pieces of the lump ore before it was broken.
Therefore as the size of the ore is reduced a larger and larger number
of particles should be taken for the sample. To conform to this con-
dition of affairs the following rule was proposed by Professor R. H.
Richards: ‘‘For any given ore the weight taken for a sample should
be proportional to the square of the diameter of the largest
particle.”’
The accompanying table embodies this rule and is based on figures
taken from the practice of several careful managers. It was arranged
and is now published with the permission of Professor Richards.
The first column shows the safe weight in pounds for a sample of
ore of any of the six grades shown and for sizes as indicated:in the
respective columns. Column 1 applies to iron ores, column 2 to low
grade lead, zinc and copper ores and even to low grade pyritic gold
ores, without native gold, where the pyrite is evenly distributed
through the ore. Columns 3 and 4 apply to ores in which the valuable
minerals are less uniformly distributed. Columns 5 and 6 apply
to ore containing fine particles of native gold or silver, also to telluride
and other “spotty ores.”
30
TABLE IV. WEIGHTS TO BE TAKEN IN SAMPLING ORES,
. a - sak ed
ees
1 2 | 3 4 5 6
etch te Of Diameter us Largest Particles Millimeters:
Sample “i ;
Very Low Low ‘ Very
cons Grade or | Grade or Medi 0 ae Rich and
Very Uni-} Uniform papetarite ae i Spotted
form Ores. Ores. ek Ores.
20,000.000 207.00 | 114.00 76.20 50.80 31.60 5.40
10,000.000 147.00 80.30 53.90 35.90 92.40 3.80
5,000.000 107.00 | 56.80 38.10 25.40 15.80 2.70
2,000.000 65.60 | 35.90 24.10 16.10 | 10.00 170
1,000.000 | 46.40 25.40 17.00 11.40 7.10 1.20
. 500.000 32.80 | 18.00 12.00 8.00 5.00 85
200.000 20.70 | 11.40 7.60 5.10 320 54
100.000 14.70 8.00 5.40 3.60 2.20 38
50.000 10.70 5.70 3.80 2.50 1.60 27
20.000 | 6.60 3.60 2.40 1.60 1.00 17
10.000 4.60 2.50 1.70 1.10 71 12
5.000 | 3.30 1.80 1.20 80 50
2.000 | 2.10 1.10 76 5l a2
1.000 1.50 80 54. 36 22
500 1.00 57 38 25 16
200 | 66 36 24 16 10
100 A6 25 nz 1
050 3G 18 12
.020 21 11 |
010 15
005 | 10 | |
It should be remembered that the above mentioned rules for sam-
pling will not hold for ore containing large pieces of malleable minerals
such as native gold, silver, silver sulphide, chloride etc. These roll
out and do not crush and must be treated by special methods. See
“Sampling Ores Containing Malleable Minerals.”
In using the table, it is not necessary to crush successively to all
of the sizes shown in any of the columns. The ore may be crushed
to any fineness convenient and then a weight of sample corresponding
to that shown in the table may be taken. In sampling mill practice
it is customary to reduce the diameter of the coarsest particles one-
half at each stage or crushing, thus reducing the volume to one-eighth
or 12.5 per cent. It is also customary in practice to take a 20 per
cent sample at each stage, consequently the ratio between the weight
of sample and size of maximum particle is constantly increasing
throughout the sampling process, thereby meeting theoretical condi-
tions previously discussed.
ol
Relation of Size of Sample to Grade of Ore and Effect of
Specific Gravity of Richest Mineral. Although it had long been
appreciated that the size of the sample would have to be greater as
the richness of the ore increased, it remained for Reed’ to develop
a formula by which the proper ratio between these could be scientifi-
eally maintained.
D = diameter of largest pieces in inches.
P = quantity of the lot in Troy ounces.
f = number of parts into which P is to be divided before one part
is taken as a sample.
= percentage of silver or gold in the richest specimens in the lot.
= specific gravity of the richest minerals.
average grade of ore (ounces per ton).
= number of pieces of D size and k value that can be in excess
or deficit in the portion chosen as sample.
largest allowable percentage of error.
So By
II
l
Before taking a sample of ; Troy oz. we must crush the lot so that
D = 05 PL mPl
sk(f-la
- Taking for general purposes s = 7, | = 1 and a value of a of 1.6,
the following table is given for different grades:
Medium m = 50 el
High grade m = 75 Kea
Very rich m = 500 Ko=230
TABLE V.
a ay
Value of D in inches.
Sample reduced from
Medium High Grade | Very Rich
100 to 10 tons
5.28 2.96 2.53
3 10to ) 1 ton 2.46 1.38 Be
2000 to 200 lbs. 1.14 0.6 0.56
200 to 5 lbs. 0.3 0.18 0.16
5 lbs. to 10 assay tons 0.0
This quantity assumed for a appears to be very small and in the
opinion of the writer should be larger, which would have the effect of
reducing the values for D shown in the above table.
Brunton’ derived a formula similar to Reed’s but more convenient
to use.
1School of Mines Quarterly VI. page 351 (1885)
2T. A.I. M.E. XXV. p. 826 (1895)
32
W = weight of sample in pounds.
k = grade of richest mineral in ounces per ton.
c = average grade of ore in ounces per ton.
s = specific gravity of richest mineral.
n = number of maximum sized particles of richest mineral in éx¢éss
or deficit in sample.
f = a factor expressing the ratio of the actual weight of the largest
particle of richest mineral which will pass a screen of a given
size to the weight of the largest cube of the same mineral which
will pass the screen.
p = allowable percentage error in sample.
D = diameter in inches of the holes in the screen, or other normal
diameter to which the ore is crushed.
From these he finds
D = 65
Making p, the allowable percentage error, = 1, the formula be-
comes:
7 = ee
D = .65 Noe
fsn(k—c)
To determine a value to use for n Brunton made a number of assays
on two different lots of high grade ore crushed to pass a certain limit-
ing screen. The average deviation from the mean = p was substi-
tuted in the formula and results of 2.64 and 3.14 respectively were
found for n. Assuming that 3 is a safe value for n and cubing each
side we find.
D? an We
~ 10.8 fs(k-c)
or
ae 10.8fsD*(k—e)
x
from which may be found the safe weight in pounds for 4 sample
of any ore whose largest particle is D inches. Taking four examples
using as the richest minerals pyrite, galena, native silver and native
gold and assuming different values for D, k, ¢ and f the following
table was made after the style of the table first shown in Hofman’s
Metallurgy of Lead. The values for f used for the fine sizes were
those determined by Brunton’s experiments, i.e. 4 for pyrite and —
galena and 6 for native silver and gold. This value of f is reduced
gradually until for 1 inch diameter, it is made equal to one, this
a ee eee ee eee
33
variation therefore tending to compensate for the greater uniformity
of value of the particles as they become larger.
This following table is probably the best and certainly the most
conservative of all. A good deal of intelligent discrimination may
often be used however and mere formula can never be made to cover
all possible contingencies. For instance in sampling an ore in which
the valuable mineral is finely and uniformly disseminated throughout
the gangue, a much smaller sample than that given in the table may
be taken for the coarse sizes, although for the fine sizes the full quanti-
ties shown in the table should be taken. Another ore, with perhaps
the same ratio of value of the richest mineral to average grade, having
the rich mineral in larger crystals or masses will have to be sampled
TABLE VI. WEIGHTS TO BE TAKEN IN SAMPLING ORE.
Ba | Size of Safe Weight in Pounds when Largest Particles are of Size Given
—£&) Particles. in Second Column.
| reer ce ee ee aE eae
eai\S| dea Grade of Richest Mineral Divided by Average Grade.
Sa) oO) 88
pa) ae | an 10 50 200 600 1500 2500
0043} 003 010 025 043
We ‘0055 .0003 0018 .007 021 053 089
50| .0100 0017 0095 039 116 291 485
5.0| 14| .0364 0585 319 1.29 3.90 9.76 16.3
Fecal F450 2.96 16.13 65.5 195. 494. 823.
3| 338 | 30.03 | 163.5 663.9 1998. 5000. Soon
af 75.9 413. 1679. 5054. 12648. 21095.
1.0 486. 2646. 10746. 32346. 80946. | 140346.
120] .0043 005 015 038 064
100) .0055 0005 0027 O11 032 080 134
50! .0100 0026 0143 058 174 437 727
7.5| 14| .0364 0878 479 1.94 5.85 14.64 24.5.
Seeds 1450| 4.44 24.2 98.3 293. 740. 1234.
9) .338 | 45.0 245. 996. 2997. 7500. 12505.
4 114. 620. 2519. 7581. 18972. 31643.
1.0 729 3969. 16119. 48519. |121419. | 210519.
120| .0043 0005 0027 011 .032 081 135
100! .0055 0010 0055 022 .068 170 283
10.5| 50) .0100 0041 0222 .090 272 679 1.133
4| .0364 1476 804 3.26 9.83 24.59 41.00
abe 450). °° 7.78 42.35 172.0 517.7 1295. | 2160.
2) 338 | 78.8 429, 1742. 5245. 13126. 21883.
5 230. 1250. 5077. 15283. 38247. | 63762.
|) milion
Therefore using one assay-ton of ore the weight of the silver or gold
in milligrams gives immediately the assay in ounces pér ton.
Calibration of Weights. The weights supplied by the makers:
cannot always be relied upon and even originally perfect ones are
subject to changes of weight due to wear or accumulation of dirt.
Therefore it behooves the assayer to occasionally check his weights
and to determine the correction to be applied to the marked value.
This requires the use of a standardized weight which should be care-
fully preserved and used for this purpose only.
The method of swings should be used and the weighing is fone by
deflection after the sensibility (deflection of the zero for 0.1 mg.)
has been determined. First determine the position of rest, and the
sensibility with no load, with 100, 250 and 500 milligram loads re-
spectively. The sensibility should not vary much throughout this
range. The method to be followed can be understood from the
following examples.
Calibration of a Set of Assay Weights.
Designate each weight by its marked value in the parenthesis and
when there are several of the same value note some peculiarity by
which they. may be designated. The weights in the set, marked in
milligrams are:
(500) = a, (200) = b, (100) = ¢, (1009 =:d) (50) = eee
(10) = g, (10) = h, (10) = 6) +2) +@)4+@Q) =i
The weight (100) is compared with the standard 100 milligram
49
weight and the weights are then compared among themselves by the
method of swings. The letters represent the true values.
In calibrating the weights from 100 milligrams to 10 milligrams,
observations should be made on the following combinations:
Left Hand Pan Right Hand Pan
(100) 100 mg. standard
(100) pre (20) 410) (10) 4--(5)-- (2), + (2)
ul)
(50) Pee 0) 010!) 5) 2) 29 (1)
(20) GLO) (10")
The recorded observations are as follows:—
100 mg. = c¢ — 0.020 mg.
¢ Peete eth +i 0/190
e afte t+h+i + 0.020
f =g+th + 0.040
g sah + 0.015
i + 0.040
Solving these equations |
I=}
Dei + 0.040
eer + 0.055
iva + 0.135
@ = 5i + 0.450
c= 101 + 0.870
ec = 100.020 mg.
_ From these last two values of ¢
101 = 99.150 mg.
P= 9.915 me.
Substituting this value for i in the above equations we find the
a following values for the other weights of the set:
Designation Actual Weight Correction ‘ to marked value
c = 100.020 + .020 mg.
e= 50.025 020s
f= 19.965 — 035 “
g= 9.970 — .030 «
h= . 9.955 — 045 “
ae 9.915 — 085 “
The smaller weights in 7, may be calibrated in a similar manner.
_ The large weights (100), (200) and (500) may be standardized by a
simple modification of the above.
The process is made much simpler by Baers a complete set of
1 A + correction means that the weight is heavier than the normal value.
50
standard weights which are very carefully handled and kept solely
for standardizing purposes, and these the larger assay offices usually
have.
Testing Riders. Every new rider should be tested before use as
they often vary 0.01 or 0.02 milligrams from their supposed value.
If too heavy, a little bit at a time may be cut off with a pair of scissors
until they come down to the standard.
CHAPTER V.
CUPELLATION.
In every assay of an ore for gold and silver we endeavor to use such
fluxes and to have such conditions as will give us as a resultant two
products :—
Ist. An alloy of lead with practically all of the gold and silver of
the ore and as small amounts of other elements as possible.
2nd. A readily fusible slag containing the balance of the ore and
fluxes. |
The lead button is separated from the slag and then treated by a
process called cupellation to separate the gold and silver from the
lead. This consists of an oxidizing fusion in a porous vessel called
a cupel. If the proper temperature is maintained the lead oxidizes
rapidly to PbO which is partly (98.5 * per cent) absorbed by the cupel
and partly (1.5 per cent) volatilized. When carried to completion
the gold and silver is left in the cupel in the form of a by+* =
The cupel is a shallow, porous dish made of bone-ash, Portland
cement, magnesia or other refractory and non-corrosive material of
spongy texture. The early assayers used cupels of wood ashes from
which the soluble constituents had been leached. Agricola writing
in about the year 1550 mentions ashes from burned bones. Ashes
from deers’ horns alone he pronounces best of all; but the use of these
antedate his time and he states that assayers of his day generally
make the cupels from the ashes of beech wood.
To-day it is thought that the bones of sheep are the best for cupels.
These should be cleaned before burning and as little silica as possible
introduced with the bones. It is important not to burn the bones
at too high a temperature as this makes the ash harderand less absorb-
ent. It is also advisable to boil the bones in water before burning
them as this dissolves a great part of the organic matter which if
burned with the bones yields sulphates and carbonates of the alkalies.
Properly burned sheep bones will yield an ash containing about
90% calcium phosphate, 5.65% calcium oxide, 1.0% magnesium
oxide, and 3.1% calcium fluoride. Ordinary commercial bone-ash
also contains more or less silica and unoxidized carbon. If more
than a fraction of a per cent of silica is found in bone-ash, it is evidence
that sufficient care has not been taken in cleaning the bones, and cupels
1 Liddell, E. & M. J. 89, pp. 1264. June 1910.
52
made from such bone-ash aremore likely to crack during cupellation,
resulting often in the loss of small buttons. If the bone-ash shows
black specks it is an indication of insufficient oxidation and the assayer
should allow the cupels to stand for some time in the hot muffle with
the door open before using. Carbon is an undesirable constituent
of cupels as it reacts with the lead oxide formed giving off CO and
CO: which may cause a loss of the molten alloy due to spitting.
Bone-ash for cupels should be finely ground to pass at least a 40-
mesh screen and the pulverized material should consist of such a
natural mixture of sizes as will give a solid cupel with enough fine
material to fill interstices between coarser particles. Opinions differ
as to the best size of crushing for bone-ash and this will depend no
doubt upon the character of the material. Bone-ash, the screen
analysis of which follows, has, however, yielded particularly good
cupels.
TABLE VIII. SIZE Oe SORE ASH.
oe
Size Mesh Size mm. ne Cent. L Wee
On 40 0.380 9 Pou Cents:
Through 40“ 60 0.244 is as
60 “ 100 0.145 gees ee
+ 100 “ 156 0.098 10: Piva
a 150 50 4c SS
With cupels made from this bone-ash it was possible to obtain
losses not exceeding 1.60 per cent using 100 mgs. of silver and 25
grams of lead, while with some other lots of bone-ash containing smaller
proportions of 150 mesh material it was found impossible to keep the
losses below 2.0 per cent.
Making Cupels. Cupels are made by moistening the bone-ash
with from 8 to 20 per cent of water and compressing ina mold. The
bone-ash and water should be thoroughly mixed by kneading, and
finally it should be sifted through a 10 or 12 mesh sieve to break up
the lumps. Some authorities recommend adding a little potassium
carbonate, molasses or flour to the mixture, but with good bone-ash
nothing but pure water need be added. The mixture should be just
sufficiently moist to cohere when strongly squeezed in the hands, but
not so wet as to adhere to the fingers or to the cupel mold. Twelve
per cent of water by weight is about right; but the amount used will —
depend somewhat on the bone-ash and on the pressure used in forming
the cupels. The greater the pressure the smaller the amount of water
which may be used. It is better to err on the side of making the mix-
ture a little too dry than too wet.
53
The cupels may be molded either by hand or machine. The hand
outfit consists of a ring and die. The ring is placed on the anvil
and filled with the moist bone-ash, the die is inserted and pressed down
firmly. It is then struck one or more blows with a heavy hammer or
mallet, turning the die after each blow: finally the cupel is ejected. The
cupels are placed on a board and dried slowly in a warm place. The
amount of compression is a matter of experience and no exact rule
for it can be given; but it may be approached by making the cupels
so hard that when removed from the mold they are scratched only
with difficulty by the finger nail. One man can make about 100
-cupels an hour using the hand mold and die.
Several types of cupel machines are on the market. One machine
has a compound lever arrangement which gives a pressure on the
cupel equal to twenty times that applied to the hand lever, and by
adjusting, different degrees of compression may be obtained. These
machines have interchangeable dies and rings so that different sizes
of cupels may be made. The rated capacity of this machine is 200
cupels an hour. Another machine made by the same company has an
automatic charging arrangement. This machine is claimed to have a
capacity of 600 cupels an hour. Cupels should be uniform in hardness
and it would seem that with a properly designed machine a more
uniform pressure could be obtained than by the use of hammer and
die. Some assayers however, still prefer hand-made cupels.
Cupels should be air dried for several days at least before use. Most
assayers make them up several months in advance so as to insure
complete drying. They should not be kept where fumes from part-
ing can be absorbed by them as the CaO present will be converted into
Ca(NO3)2. This compound is decomposed at the temperature of
cupellation and may cause spitting of the lead button.
Cupels should not crack when heated in the muffle and should be
sufficiently strong so that they will not break when handled with the
tongs. Good cupels give a slight metallic ring when struck together
after air-drying. It is best to heat cupels slowly in the muffle as this
lessens the chance of their cracking.
A good cupel should be perfectly smooth on the inside and of the
right porosity. If it is too dense, the time of cupellation is prolonged
and the temperature of cupellation has to be higher, thus increasing
the loss of silver. If it is too porous it is said that there is again danger
of a greater loss due to the ease with which small particles of alloy
can pass into the cupel. The bowl of the cupel should be made to
hold a weight of lead equal to the weight of the cupel.
The shape of the cupel seems to influence the loss of precious
54
metals. 22" oe See o
The larger the amount of ore used the more necessary it becomes
to keep down the quantity of fluxes. The following charges more
acid than the bi-silicate are regularly used in this laboratory for the
assay of siliceous tailings. If the tailings were pure silica the slags
would be almost tri-silicates.
91
Ore eA a AST. DA esl
Soda (Na2CQOs3) 30 gms. 60 gms. 150 gems.
Borax ae aeons Lia
Litharge Bp 00) Meet OY es
Argols | : for a 25 gram bead button.
| The results obtained with the last mentioned charges are good, the
_ slags of course are more viscous than the bi-silicate slags but they pour
well even when fusions are made in the muffle furnace. The crucibles
are practically unattacked and can be used for many such fusions
if of good quality and especially if care is taken to cool them slowly.
The following table of bi-silicate factors contains all of the data
— necessary for fluxing siliceous ores with any of the common basic
fluxes, and it is thought that the above explanation will make its
use readily understood.
TABLE XVIII. BI-SILICATE SLAG FACTORS. NO. 1.
Quantity of Bases Required.
SiOz Mie
PbO | NavCOs | KeCOs NaHCOs
_ 1 assay-ton 108.4 gms. | 51.5 gms. | 67.1 gms. | 81.7 gms.
_ 10 grams ele TR eh 23 Oe paiy ©
One gram of FeO neutralizes 8/10 gms. SiO. or requires 1.4 gms.
_ borax-glass. One gram of CaCO; neutralizes 0.5 gms. SiO or re-
- quires 1.0 gms. borax-glass.
To make a bi-silicate slag and at the same time to keep the quantity
3 of fluxes at a minimum the procedure would be as follows:—First,
_ find the amount of silica which will be converted into bi-silicates by
q the bases of the ore. Second deduce this quantity from the amount
_ of silica in the ore and add basic fluxes for the remainder.
For example take an ore of the following composition :—Si0O2 80%,
— FeO 10%, CaCO; 10%, ore charge 1 assay-ton.
Silica equivalent FeO 2010 Pia pan ta tater
o = macs 2.916 x 5 = 1.46
Total 3.78
Silica in ore moth 1292166 = 23.33" ema.
less silica equivalent of FeO and CaCQ; . 3.78
Silica for which fluxes are to be added 19.55 gms.
92 . ; ath
Starting with 30 grams of soda (N 240) ad nog ir grams of
borax this gives the following charge :—
Ore LCA
Sodium-carbonate -. 30- gis:
Borax 10235
Litharge for slag 23 BQ
Litharge for lead button 27
Argols for 25 gram button.
Following the more usual custom of adding a given amount of q
soda-litharge flux and then adding extra borax for the bases, we have
simply to compute the borax to add for the 10 per cent of ferrous
oxide and limestone.
Borax glass equivalent FeO 2.916 X 14 = 4.08
Borax glass equivalent CaCO; 2.916 xX 1.0 = 2.92
Total 7.00
Whence the charge becomes
Ore DAG.
Soda (NaeCQs;) 30 gms.
Borax-glass Oita 1G. 25
Litharge 602
Argols for a 25 gram lead button.
Slags for Glass 1 Basic Ores. In the assay of basic ores we have
to add acid fluxes, silica and borax to obtain a fusible slag. Also
on account of the fact that the silicates of iron, manganese, calcium,
magnesium and aluminum are by themselves infusible, or nearly so,
at the temperature of the assay furnace, it is always customary to
add some soda and excess litharge to the charge. These latter, com-
bining with some of the silica and borax, form readily fusible com-
pounds which help to take into solution the silicates of the basic
oxides and by diluting them give more fusible and fluid slags. A
quantity of soda equal at least to that of the ore is generally taken
as a starting point, and very often a quantity of litharge equal to
that of the ore is also allowed for the slag.
The silicate degree of the slag will depend on the character of the
bases. For Class 1 ores consisting principally of iron, manganese,
calcium, magnesium and aluminum it has been found best to approxi-
mate a sesqui or a bi-silicate slag.
The following table of bi-silicate slag factors will ee the
calculation of charges for basic ores.
a
by he. 93
‘ m r ary’ Z :
oe ap ee “ ;
- 2 ' TABLE XIX. BI-SILICATE SLAG FACTORS NO. 2.
“ Quantity of Bases. Quantity of Acid Required.
1 A. T. FeO 24.3 gms. SiOe
1 A. T. CaCOs ig at AE .
1 A. T. MgCOs 21:0 pas “
oF 1 A. T. MgO 4574." ss
10 gms. PbO Din ae ene
30 “ NaHCOs 10.8 a 4
30 “ NaeCOs 16.4 - *
Peet KOO: 4.4 i a
For sesqui-silicates use three-quarters of the above quantities of
silica.
When borax-glass is substituted for silica consider that one gram
of borax-glass is equivalent to 6/10 gram of silica. The amount of
silica which should be replaced by borax has not been determined,
but on account of the greater fusibility and fluidity of boro-silicates
it is well to replace at least } to $ of the silica with its equivalent
of borax or borax-glass.
The following example will illustrate the use of the table. Take
an ore of the following composition CaCO; 90%, SiOz 10%, ore charge
1 assay-ton. Starting with 30 grams of sodium carbonate and 50
grams of litharge, 20 for the slag and 30 for the lead button an jlan-
ning for a bi-silicate slag the silica requirements of the different bases
are as follows:—
The CaCO; of the ore requires 0.9 X 17.6
15.80 gms. SiO,
30 grams soda requires 7 Nek: Pee oo
20 he PbO 2 — 5.4 6 éé
Total 37.6
Deducing the silica of the ore 2.9
Silica or equivalent borax necessary 34.7
Putting in say 20 grams of silica we have to provide borax equivalent
to 34.7 — 20 = 14.7 grams of silica or 24 grams borax-glass.
_ The final charge stands
Ore LAS TD: Litharge 50 gms.
Sodium carbonate 30 gms. Argols for 25 gram button
Borax-glass 24 gms. Silica 20 gms.
Reducing and Oxidizing. Reducing and oxidizing reactions,
are common in fire assaying as in other chemical work, and practically
all fusions are either reducing or oxidizing in nature. For instance,
the scorification assay is an oxidizing fusion in which atmospheric
94
air is the oxidizing agent, while the crucible fusion of a siliceous ore
is a reducing fusion in which argols, flour or charcoal act as the re-
ducing agents.
By reducing power as used in assaying is meant the amount of lead
that one gram of the ore of substance will reduce when fused with
an excess of litharge. For instance, if we use 5.00 grams of ore and
obtain a lead button weighing 16.50 grams the reducing pov of the
ore is
16.50
5.00
By oxidizing power is meant the amount of lead which one gram
= 3.30
of the ore or substance will oxidize in a fusion, or more exactly it is —
the lead equivalent of a certain amount of reducing agent or ore which
is capable of being oxidized by one gram of the ore or substance.
Reducing Reactions. The reduction of lead by charcoal is shown
by the following reaction :—
2PbO + C = 2Pb + COs
From which it is seen that one gram of pure carbon should reduce
2 X 207
rahe Bs: 34.5 grams of lead. As however charcoal is never pure
carbon the results actually obtained in the laboratory will be some-
what less usually from 25 to 30. All carbonaceous materials have
more or less reducing power. Those most commonly used as reducing
agents in assaying are charcoal R. P. = 27.5, argols R. P. 8-12,
cream of tartar R. P. 5.5, flour R. P. 9 — 12.
Besides carbonaceous matter many other substances and elements
are capable of reducing lead from its: oxide. The most important
of these are metallic iron, sulphur and the metallic sulphides. The
reduction of lead by iron is shown by the following reaction :—
PbO + Fe = Pb + FeO
Whence the reducing power of iron is oe =aed
The reducing power of sulphur and the metallic sulphides will
vary, dependent on the amount of alkaline carbonate present. For
instance, the reduction of lead by sulphur in the absence of alkaline
carbonates is shown by the following reaction :—
2PbO +S = 2Pb + SO,
The reducing power of sulphur under these conditions would be
95
In the presence of sufficient alkaline carbonates the sulphur is
oxidized to sulphur-trioxide which combines with the alkali to form
a sulphate. The reaction is as follows:—
From which we see the reducing power of sulphur under these con-
ditions should be :
cele 19.4
32
In the same way we find that the reducing power of the metallic
sulphides varies according to the amount of available alkaline car-
bonate present. For instance, in the absence of alkaline carbonates
the following equation expresses the reaction between iron pyrite
and litharge :—
FeS, + 5PbO = FeO + 5Pb + 280,
Whence the reducing power is found to be on = 8.6
- In the presence of an excess of sodium carbonate the sulphur is
oxidized to trioxide as indicated by the following reaction :—
FeS, + 7TPbO + 2Na2COs = FeO + Pale + 2NaeSO, of ZC.
Which gives a reducing power of ae == 2s
In order to get such a high reducing power as this it is necessary to
have a very basic slag.
Any silica present combines with soda and litharge to form a silicate
and if it is present in any considerable amount the litharge and soda
are rendered unavailable for the higher oxidation of the sulphur.
_ The amount of lead reduces from any charge by any reducing agent
is always a function of the temperature and the silicate degree. Other
things being equal the more basic the charge the greater the amount
of lead reduced by a unit quantity of the reducing agent. Thus,
a certain sample of argols showed a reducing power of 11.04 when
silica for a sub-silicate was added, 10.93 for a mono-silicate, 10.62
for a bi-silicate and only 9.26 for a tri-silicate. :
The accompanying table gives the reducing power of some of the
common sulphides. The theoretical figures are computed both for
sulphur oxidized to SO. and SOs. In the last columns are given the
reducing power of the pure minerals using the following charge NagCO;
5 gms., PbO 30 gms., SiO. 2 gms., ore to yield an approximate 25
gram button.
96
TABLE XX. REDUCING POWER OF MINERALS.
Computed ;
Mineral Formula || ———--__—__ eee pees d
S to SO2 | S to SOs
Cralena PbS 2.6 3.46 3.41
Chalcocite CuzS 3.9 5.2
Arsenopyrite FeAsS 5.7 6.96 8.18
Stibnite SbeSs San 7.30 6.75
Chaleopyrite CuF es: 6.2 8.44 7.85
Sphalerite ZnS 6.37 8.5 7.87
Pyrrhotite Fe7Ss 7.35 9.9 10.00
Pyrite FeS: 8.6 bo be Or 11.05
Oxidizing Reactions. Certain metals notably iron, manganese,
copper, cobalt, arsenic and antimony are capable of existing in two
states of oxidation. When fused with a reducing agent the higher
oxides of these metals are reduced to the lower state of oxidation at
the expense of the reducing agent. Ores containing these higher
oxides are said to have an oxidizing power on account of this property
of using up reducing agent. For convenience this oxidizing power
is measured in terms of lead although the bulk of the oxidizing re-
action in any assay fusion is probably accomplished against the re- —
ducing agent of the charge. :
For instance if in an assay fusion containing silica we have ferric
oxide, sufficient for a bi-silicate, and carbon the following reaction
takes place :—
2Fe.0; + C + 4810, = 4FeSiO; + CO, .
From which we find that one gram of Fe,O3; requires 0.037 gram of
carbon to reduce it to FeO. Expressed in terms of lead the relation
would be as follows :—
Fe,0; + Pb = 2FeO + PbO
That is to say the oxidizing power of FegQs is ~ = 13
Similarly , .
MnO, + Pb = MnO + PbO
The oxidizing power of MnO, is = = 2.4, which means that each
gram of MnO; present in a fusion with litharge and a redueing agent
will prevent the reduction of 2.4 grams of lead. It is easily seen
therefore that this oxidizing power of ores must be taken account of
in computing a furnace charge. The method of determining the
oxidizing power of ores etc. will be discussed later.
97
In the crucible assay of high sulphide ores it is frequently necessary
to add some oxidizing agent to the charge to prevent the reduction
of an inconveniently large lead button. A lead button of 25 or 30
grams is usually sufficiently large to act as a collector of the precious
metals and were a larger button obtained, it would entail an extra
loss due to scorification or a prolonged cupellation as well as consum-
ing extra time in this treatment. When therefore the ore charge
would of itself reduce more than 25 or 30 grams of lead we ordinarily
add potassium nitrate (niter) or some other oxidizing agent. Niter
is almost exclusively used in this country for oxidizing. Its action
with carbon is shown by the following equation :—
AKNO; + 5C = 2K20 + 5CO, + 2N2
From which the theoretical oxidizing power of niter expressed in
terms of lead is found to be 5.17. The actual oxidizing effect of niter
is always found to be lower than this due to the acidity of the charge
and the probably escape of some oxygen.
In the soda-litharge-silica fusion such as commonly used in actual
niter assays (sub-silicate), the reaction between niter and iron pyrite
will be about as indicated by the following equation :—
k 10OKNO; -f 4FeS. a S10. = FesSiOg oh 5KoSO, + 350, + 5No
- Assuming the reducing power of pyrite in this type of charge to be
10.0 the oxidizing power of niter is found to be 4.7.
A slight oxidizing effect may be obtained by using red lead (Pb3Q,)
in place of litharge and this is sometimes done especially in England
and the English colonies. The oxidizing effect of red lead is shown
by the following reaction :—
Pb;0, + Pb = 4PbO
The oxidizing power in terms of lead is wan = 0.30
Slags for Class 2 Ores. When atisere contains any considerable
proportion of sulphide minerals and:especially when they are present
in such proportions that it is necessary-to add niter to prevent the
reduction of too much lead it will be:¢#eund that the charges recom-
mended for class 1 ores will not allow..a satisfactory decomposition
of the ore. Instead of obtaining two:products, slag and lead, as the
result of the fusion a third intermediate product (matte) is often -
obtained. This amounts to incomplete decomposition of the ore
and is a sure indication of low results. The rate of melting down or the
temperature of the furnace during the early part of the fusion seems
to have a great. deal to do with the formation of a matte... A rapid
melting in a hot furnace seems to allow a more complete. oxidation
98
of the sulphur, possibly because the litharge is utilized for this pur-
pose before it is locked up by the silica. At any event as rapid melt-
ing as may be without boiling over seems to give the most satisfactory
results with class 2 ores.
A matte is much less likely to be formed, however, with a less acid
charge and it has been found best therefore to make a slag approach-
ing a sub-silicate for all heavy sulphide ores, as by this means more
uniformly satisfactory results are obtained. :
The Cover. In practically all crucible assay work it is customary
to place on top of the mixed charge in the crucible-a cover of some
fusible substance. Different assayers advocate different materials
as salt, borax, borax-glass, soda as well as different flux mixtures.
The idea of the cover is that melting early it makes a thick glaze
on the sides of the crucible above the ore charge and tends to prevent
sticking of particles of ore or lead globules which might be projected
or left there by the boiling of the charge. As the fusion becomes quiet
and the temperature rises, most of this glaze runs down to join the
main charge and carries with it any small particles of ore or lead which
may have stuck to it in the early part of the fusion.
The salt cover is thinly fluid when melted. It does not enter the
slag but floats on top of it thus serving to keep out the air and to
prevent loss by ebullition.
The borax cover fuses before the rest of the charge. It is thick
and viscous when melted and serves to prevent loss of fine ore by
“dusting”, as well as to stop loss by ebullition. It finally enters the
slag and so ceases to be a cover after the fusion is well under way.
Some assayers object to the use of salt on the ground that it is
liable to cause losses of gold and silver by volatilization. It is a well
known fact that gold chloride is volatile at a comparatively low tem-
perature, commencing at 180°C. and that silver chloride is volatile
in connection with the chlerides of arsenic, antimony, copper, iron,
lead etc. When an ore eentains substances such as manganese
oxide, basic iron sulphate ete., capable of generating chlorine upon
heating with salt it would seem wise to omit the use of salt. If it
is not desired to use salt a good cover may be made from a mixture
of borax-glass and sodium carbonate in the proportion of 10 Bai
. of the former to 15 parts of the latter.
Testing Reagents. Each new lot of litharge and test lead should -
be assayed for silver and gold so that when any is found to be present
a proper correction may be made. Different lots of argols, charcoal
etc. are also found to vary in reducing power so that their reducing
powers should also be determined.
99
The following procedure is designed, Ist, to allow the student to
determine the reducing power of argols, charcoal or other reducing
agents and at the same time to determine the silver correction for
litharge, and, 2nd, to familiarize him with the principal operations
connected with the crucible method of assay.
Procedure. Take two E or F pot furnace crucibles, or 12 or 15
gram muffle crucibles.
Weigh into them, in the order given the following :—
INO. 1 No. 2.
Sodium carbonate 5 grams Sodium carbonate 5 grams
Silica Dites Silica ae’
Litharge iP ge Litharge OUe
Argols Qeee eye: Me harcoal Lat
Weigh the argols and charcoal on the pulp balance as exactly as
possible, the others on the flux balance. Mix thoroughly with the
spatula by turning the crucible slowly with one hand while using
the spatula with the other. When finished tap the crucible several
times with the handle of the spatula to settle the charge and to shake
down any material which had lodged along the inside of the crucible
above the charge. Finally put on a 4 inch cover of salt.
Pot Furnace Fusion. Have a good bright fire in the pot furnace
which should not however be filled with coke more than half wuy to the
bottom of the flue. Figure to so place the crucibles that their tops
shall not be much above the bottom of the flue. Place a piece of
cold coke directly under each crucible as it is put into the furnace.
Cover the crucibles and pack coke around them being careful to pre-
vent the introduction of any coke or dust. Close the top of the fur-
nace, open the draft if necessary and urge the fire until the charges
begin to fuse. Then close the draft and continue the melting slowly
enough to prevent the charges from boiling over. When the charges
have finished boiling, note the time and open the draft if necessary
to get a yellow heat and continue heating for 10 or 15 minutes.
Pour the fusions into the crucible mold, which has been previously —
coated with ruddle, thoroughly dried and warmed. When cold, a
matter of 5 or 10 minutes for a small fusion, break the cone of lead
from the slag and hammer it into a cube to thoroughly remove the
slag. Weigh the buttons on the pulp balance to the nearest tenth
of a gram and record the weights and reducing powers in the note
book. —
Save the lead buttons and cupel them, using. eee cupels. They
should contain all of the gold and silver in the 60 grams of litharge
used. Weigh the beads and part to see if gold is present. Record
_—
VS
100
the weights of the beads and compute the correction for silver in 30
erams of litharge.
Muffle Fusion. If the fusions are to be made in the muffle have
the muffle red and the fire under such control that the muffle can be
brought to a full yellow in the course of a half hour. Melt at suffi-
ciently low temperature to avoid violent boiling and then raise the
temperature and pour as in the case of the pot furnace fusion.
Notes. 1. So-called silver free litharge can now be purchased but even this
often carries traces of gold and silver.
2. In assaying samples of litharge low in silver 120 to 240 grams may be required
to give a button of sufficient size to handle and weigh.
3. It is convenient to use litharge in multiplies of 30 grams and therefore the
silver correction is based on 30 grams of litharge.
4. The temperature which the muffle should have before the crucibles are in-’
troduced depends upon the number of charges which are to be put in at one time.
If only one or two the temperature should be low to avoid danger of boiling over.
If, however the muffle is to be filled with crucibles the initial temperature may be
higher as the crucibles can be depended upon to decidedly lower the temperature.
5. Pour the fusions carefully into the center of the molds and do not disturb
until the lead has had time to solidify.
The following are the reducing powers of some of the common re-
ducing agents.
Charcoal 23-30 Corn Starch 11.5-13
Argols 7-12.5 White Sugar 14.5
Flour 12-15 Cream of Tartar 4.5-6.5
Assay of Class 1 Ores. Gold or Silver.
This is the most common class of ores and as it is also the one which
presents the fewest difficulties for the assayer, it is considered first.
Actually, ores with no traces of sulphides are somewhat of a rarity
but the methods given below may be adapted to ores containing mod-
erate amounts of sulphides by simply decreasing the amount of re-
ducing agent used.
Procedure. Carefully van some of the ore, estimate and record in
the note-book the amount and character of each of the slag forming
constituents and also of any sulphides present. If the ore is mainly
siliceous weigh out one of each of the following charges :—
Charge (a). Charge (b).
_) Ore 0.5 A. T. Ore 0:5 AT
Sodium carbonate 30 grams Sodium carbonate 15 grams
Borax eo) ne Borax 3-5
Litharge SOR Litharge BQ: as
Argols ‘i ~ Argols -
* Argols enough combined with the reducing material of the ore to
give a 25 gram button.
101
_ Weigh out the fluxes and place in the crucible in the order given and
finally add the ore and argols last of all. Weigh the argols and ore on
the pulp balance, the others on the flux balance. Mix thoroughly
and place a 3 inch cover of salt or soda-borax mixture on top.
Use F pot furnace crucibles, 15 or 20 gram muffle crucibles, if
work is to be done in the muffle. :
Fuse at a moderate red heat to avoid danger of the charge boiling over
and when quiet raise the heat to a bright yellow. Allow 15 minutes
of quiet fusion. Pour as usual, tapping the crucible gently against
the mold if necessary to insure getting out the last globules of lead.
When cold, separate the lead buttons from the slag keeping them
in order (a) (b). Record in the note-book the character and appear-
ance of the slags, the ease or difficulty of the separation of each
from the lead buttons, the appearance of the lead buttons and their
greater or less malleability.
Weigh the lead buttons on the flux balance and cupel carefully
to obtain feather litharge. Weigh the silver beads, correct for silver
in the lithage used, part and weigh any gold found and finally report
the value of the ore in oz. per ton.
Both of these charges should give good results on a siliceous ore.
Charge (a) is a little less expensive, but charge (b) is more commonly
used, as the slag contains two bases and the excess litharge will hold
a moderate amount of impurities in solution. Charge (b) also gives
a better separation of lead button and slag and has the further ad-
vantage that if the ore contains slightly more sulphides than was
estimated the litharge will take care of them, giving a lead button
free from matte. In charge (a), if we have more carbonaceous re-
ducing agent plus sulphide mineral than the 30 grams of litharge can
oxidize some of the sulphur will combine with various metals of the
charge, principally lead, and form a matte which will appear immedi-
ately above the lead button.
_ Approximately 28 grams of litharge from each charge will be re-
duced to give the 25 gram lead button and is therefore not available
to combine with the silica. The active! fluxes are then in charge
(a), 30 grams of soda, 3-5 of borax, 2 grams of litharge and a little
-K.O from the argols, totaling approximately 23 times the ore. In
charge (b), the active fluxes are 15 grams of soda, 3-5 of borax, 22
grams of litharge and a small amount of K.O, totaling approximately
3 times the ore. A very good rule to follow in making crucible charges
is always to use at least 25 times as much active flux as ore.
! By active fluxes is meant a flux which is to appear in the slag and therefore
does not include the litharge which goes to form the lead button.
102
Borax in the charge should be increased as the bases increase. For
an ore with 10 or 20 per cent of iron, manganese oxide or limestone
add up to 10 or 15 grams of borax or 5 to 8 grams of borax-glass.
Notes. 1. Some assayers prefer to omit the borax from the charge and use a
cover of crude borax or borax-glass in place of the salt. A borax cover may be used
to advantage with ores which “dust” in the crucible, as the borax swells and melts
early tending to catch and hold down the fine particles of ore which are projected
upward from the charge.
2. The crucible should never be more than two-thirds full when the charge is
all in.
3. If a silver assay alone is asked for it is customary to omit parting and report
the combined precious metals as silver.
4. In assaying for gold alone if sufficient silver for parting is not known to be
present, a piece of C. P. silver should always be added to the crucible or to the lead
button before cupelling. If the approximate amount of gold is known allow about
eight times its weight of silver.
5. The slag should be fluid on pouring and should be free from lead shot. If
it strings out in long threads on pouring it is too acid. If it is pasty or lumpy,
either the fusion has not been long enough to thoroughly decompose the ore, or the
charge is too basic and more borax and silica should be added. ‘The crucible should
have a thin glaze of slag and should be but little corroded. It should show no
particles of undecomposed ore or ‘‘shots’ of lead. These latter can best be seen
immediately after pouring and the student should make it a point to examine his
crucible immediately after every pour. Neither the cover nor the outside of the
crucible should show any glazing, as this indicates that the fusion has boiled over.
The cold slag should be homogeneous, as otherwise it indicates incomplete decom-
position of the ore. Glassy slags are usually preferred by assayers but are not
essential to all ores.
6. If the button is hard or brittle or weighs more than 30 grams it should be
scorified before cupelling. Hard buttons indicate the presence of copper, antimony,
or nickel. Brittle buttons may be due to antimony, arsenic, zinc, sulphur, litharge
or may be rich alloys of lead and the precious metals.
7. Examine carefully the line of separation of the slag and lead. The separation
should be clean with no films of lead adhering to the slag. There should be no third
substance between the slag and lead, nor should the surface of the lead show any
disposition to crumble when hammered. Any lead gray, brittle substance between
the lead and slag or attached to the lead button is probably matte. This indicates
incomplete decomposition of the ore due to too short a time of fusion or to incorrect
fluxing. If from the latter cause, decreasing the silica and increasing the soda and
litharge will usually prevent its formation in a subsequent fusion.
8. For low grade gold ores or tailings a 1 A. T. charge is commonly used and for
very low grade ores and tailings from 2 to 10 A. T. are run usually in the pot furnace.
In using large charges of siliceous tailings the following charges more acid than the
bi-silicate have been used with excellent results.
Gold Ore | Low Grade Tailings
Ore LAE 2A. T. 5 Age
Sodium carbonate 30 grams 60 grams 150 grams
Borax 3-5. CS 6-10. “ 15-25 “
Litharge HO ties 9005 180 55
Argols for a 25 or 30 gram button in each case.
Crucible G, 20 or 30 gm_ “4H, 30 or 35 gm K.
103
Assay of Class 2 Ores
Ores of this class containing only small amounts of sulphides are
assayed exactly the same as class 1 ores using lesser amounts of argols.
When however, sulphides are present in such amounts as to reduce
a lead button too large to cupel (i.e. over 25 or 30 grams) a different
method of procedure must be followed. The most important methods
for the assay of these ores follow:—
1. Scorification. This method has. already been considered.
It is not well suited for gold ores and fails for many silver ores.
2. Niter Method. The reducing power of the ore is first de-
_termined by means of a preliminary assay. Using the figure thus
obtained a certain amount of niter is added to the regular fusion to
oxidize a part of the sulphur of the ore, thus preventing the reduction
of too large a lead button. This is perhaps the most common method
for the assay of sulphide ‘ores. The sulphides are decomposed partly
by litharge and partly by the niter.
3. Iron Method. The litharge added to the charge is kept low
so that the lead from it plus that in the ore will yield a button of
suitable size for cupelling. The sulphide minerals of the ore are
decomposed by the means of the metallic iron. This is a very good
method for many ores and is very commonly used.
4. Roasting Method. A carefully weighed portion of the ore
is roasted to eliminate sulphur, arsenic, antimony etc. and the roasted
ore is then assayed as a class 1 ore.
5. Combination Wet and Fire Method. The sulphides ete.
of the ore are oxidized with nitric acid, the silver is precipitated as
chloride and combined with the insoluble residue containing the gold,
is assayed by either scorification or srucible.
Niter Assay. Preliminary Fusion. Procedure: Take from
21 to 10 grams of ore depending on ‘the amount of sulphide present,
2} grams for pure pyrite, and correspondingly greater amounts for
ores containing less sulphides. If the ore is mostly galena as much
as 7 grams may be taken, the idea being always to get a button of
25 or 30 grams. (See reducing power of minerals.) Take the same
amount of sodium carbonate as ore, 60 grams of litharge and up to
5 grams of silica. If the ore contains silica a proportionately less
amount should be added. Use an E crucible for the pot furnace or
a 12 or 15 gram crucible for the muffle. Weigh out the fluxes first,
104
in the order given and place the ore on top mixing thoroughly with a
spatula. Place a 4 inch cover of salt on top.
Fuse for 10 or 15 minutes finishing at a good yellow heat. Pour —
into crucible mold, allow to cool, separate the lead from the slag and
weigh on the pulp balance to tenths of grams. Divide the weight of
the lead by the weight of the ore taken to obtain the reducing power.
It should be noted that this reducing power is not an absolute
thing but depends upon many factors such as the ratio of sodium
carbonate to ore, the amount of borax, litharge and silica added as
well as the temperature at which the fusions are conducted. The
size of the lead button reduced in any fusion is decreased by any in-
crease of the borax and silica and is increased by any increase of the
litharge, soda or temperature.
The charge suggested for determining the reducing power of an
ore gives as a rule slightly higher results than are obtained in the regu-
lar fusion, due to the large amount of litharge used in the preliminary
fusion. ‘This seems to be necessary. however to insure the presence
of a sufficient excess of litharge for all ores. The reducing power
obtained in the regular assay is called the working reducing power,
to distinguish it from that obtained in the preliminary fusion.
Estimating the Reducing Power of Ores. In many instances
it is possible to estimate the reducing power of an ore within close
limits. This requires a knowledge of the reducing powers of the com-
mon sulphide minerals (see Table XX), as well as the knack of
vanning. The ore is vanned and the per cent of the various sulphides
estimated, from which data the reducing power is found. For in-
stance, if the ore is 50% pyrite and the rest gangue, the reducing power
will be about 5.5 (50% of R.P.wf pure pyrite). If it is 40% galena
and 10% sphalerite, the redueimggpower will be 40% of 3.4 + 10% of
7.9 = 2.15 approximately. "The reducing power of the ore being
equal to the sum of the products of the reducing powers of the differ-
ent constituents, multiplied’ by? tHe percentage of each | in the ore and
the whole divided by 100. “#* # .
In general if the amount of Stlphides in the ore is comparatively
small and especially if only 0.5 ‘assay-ton of ore is used, it is a very
simple matter to’ obtain a lead’ button of suitable size for cupelling
by thismeans. If for example we have a mixture of galena and gangue
mineral containing 50% of pews the reducing power of the ore will be
ely = 4. 70. aside + assay ae of this ore we heala ebtnin & lead
9 : ; ; : i
make Peep: tt as) 3; ‘WES Waa Fog
a oe
Se a
|
|
105
button weighing 24.8 grams without either argols or niter. If we
had estimated the galena at 40%, we would have added 4 gram of
argols (R. P. 10) and would have obtained a 29.8 gram button which
could still be cupelled. In a similar manner if we had estimated the
galena at 60%, we would have added about | gram of niter and would
obtain a button of about 20.8 grams, which is also all right for cupella-
tion.
When we have practically pure sulphides, as in the case of pyrite .
or galena concentrates it is again easy to estimate the reducing power
and properly control the size of the lead button.
Determining the Oxidizing Power of Niter. The oxidizing
power of niter is found by fusing & weighed amount with an ore whose
reducing power is known. To obtain comparative results the slags
must be exactly like those used for the reducing power fusion and
moreover to obtain the proper size of lead buttons in the final assay
the slag that is made there must be similar as regards acidity, litharge
excess etc. to that made in the preliminary fusion.
The following example illustrates the method of finding the oxidiz-
ing power of niter:—
Ore 5 grams 5 grams
Sodium carbonate ee Dears
Litharge 60“ hee
Niter ess
Silica Spe tas See
Lead obtained 24.31 grams _— 6.61 grams
2 2A-3 1 bt / a
Reducing power of ore = 4.86
Lead oxidized by 4 grams of niter 24.31 — 6.61 = 17.70
he ' 1s
Oxidizing power of niter = ieee 4.42
if} ng
09
Niter Assay. Regular Fusion. Knowing the reducing power
of the ore and the oxidizing power of niter we are ready to make up
the charge for the regular assay. As in the case of class 1 ores it
seems best always to use at least as much normal sodium carbonate
as ore and we may start off on this basis. More litharge is used in
this assay than in those previously discussed and assayers usually
increase the litharge in proportion to the sulphur. The rule for the
use of litharge proposed by Lodge is a good one and calls for 20 per
106
cent in excess of the amount required to satisfy the reducing power
of the ore. On account of the large amount of sulphur present a
matte is often obtained if the acidity of the charge is not carefully
controlled. It is therefore best in adding silica and borax to avoid
using more than required for a sub-silicate.
The following examples may serve to show the method of computing
charges. They are all based on 0.5 assay ton of ore as that is usually
the maximum amount used for the niter assay. The litharge is
computed according to Lodge’s rule, then the silica required for the
ore, soda, niter and active litharge is found and the rest of the com-
putation is exactly like that discussed under class 1 basic ores.
No. 1. No. 2. No. 3.
90% Galena 50% FeAsS 90% Fes:
10% CaCO; 50% SiO» 10% SiOz
Rebs 15 R. P. 4.10 R. P. 9.50
Ore C.5eA TL: Ob Ass 0.5°A, Te
Sodium carbonate 15 grams 15 grams 15 grams
Borax-glass Dat an ae 10s ee i
Litharge 1 Ue ae $0 es 180-23
Niter (O. P. 4.2) Sarees: Siar 26.5: *
Silica = = 5.025
Procedure: Make up charges for your ores according to the rules
outlined above. Conduct the fusions as for class 1 ores taking par-
ticular care when much niter is used until the boiling period is passed.
As soon as all danger of boiling over has passed, heat rapidly to a full —
yellow and pour after 20 minutes of quiet fusion. a
The following table of sub-silicate slag factors will aid in ae de- 7
termination of the quantity of acids to add. .
TABLE XXL. SUB-SILICATE SLAG FACTORS.
Quantity of Acids Required
Quantity of. Bases
Silica - Borax-glass
LANE. | FeO. . oy: 6.08 gms. 7.4 gms.
LAST. CaCOs 440 oe aac
1A.T. MgCOs OALO (AOS
10 gms. PbO 0.67 S -0:82.
30 . NaeCOs 41 ret Oar
40 “ NaHCOs ee ree 4a
10. “2 7 KsGOs 1 125s Ll.
107
Notes. 1. It has been found possible to use somewhat less litharge than that
necessary to satisfy the reducing power of the ore when a greater proportion of alkal-
ine carbonate flux is used. This is probably due both to ‘the oxidizing effect of COs
at a high temperature and to the solvent power of basic alkaline slags for matte.
2. As sulphide ores usually contain more or less copper, nickel, arsenic, antimony,
zinc, tellurium and other so-called impurities, the large amount of litharge used
serves the double purpose of helping to decompose the ore by oxidizing the sulphur
and associated metals, and also tends to prevent these metal impurities from enter-
ing the lead button.
3. Aside from the inconvenience of the preliminary fusion the principal objection
to the niter assay is due to the low results yielded. This is undoubtedly caused by
the oxidation and slagging of the precious metals, particularly silver. To avoid
this source of error only small amounts of niter should be used. When silver alone
is being sought the niter may be entirely done away with by reducing the ore charge
to a quantity sufficient to give a lead button weighing between 20 and 30 grams.
In gold assays, however, a charge less than 0.5 assay ton is to be avoided as it fails
to give a sufficiently close valuation of the ore.
4. Part of the oxidized precious metals may be recovered from the slag after the
fusion is quiet by the addition of some reducing agent. For instance, if the fusion
has been made in the muffle and without salt covers some crucibles of soft coal may
be placed ni the mouth of the muffle after the fusions have become quiet. The
smoke filling the muffle will enter the crucibles and reduce some lead from the slag
which will in turn take with it part of the silver and gold.
The Iron Assay. The iron nail thethod of assaying sulphide ores
is radically different from any of the other methods described. The
principal difference being that metallic iron, usually in the form of
nails, is used as the reducing and desulphurizing agent. As iron
reduces lead from litharge this latter reagent is limited to 30 grams
or less and to make up for this the quantity of alkali carbonate is
increasing to two or three times that of the ore. Just before pouring
the excess iron is removed. |
~The chemical reactions taking place in the crucible are entirely
different from those of the other crucible methods. In the case of
-the argols, niter and roasting methods of assaying, the sulphides
of the ore are oxidized by litharge, niter or the oxygen of the air and
the sulphur either passes off as SO2 and SQ; or in the presence of sodium
carbonate is converted into sodium sulphate which floats on top of
the slag. In the iron assay, part of the sulphur is oxidized by the small
amount of litharge used and the rest stays as sulphide, appearing
either as an iron matte on top of the lead button or dissolved in the
excess of basic slag. |
The following reactions are illustrative of the chemical changes
which take place. They are arranged i in order of their occurrence.—
feo | PHS'-+-i 2PbO'= 3Pb + SO.
Bee CuS + 2PbO°= 2CuPb + SO. 3
FeS, + 5PbO = 5Pb + FeO + 2580,
Fe + PbO = Pb + FeO
108
when the litharge is all reduced the following occur
PbS + Fe = Pb + FeS (matte)
FeS. + Fe = 2FeS (matte)
SboS3 + 3Fe = 28b + 3FeS (matte)
AsoS3 + 13Fe = 2Fe;As (speiss) + 3FeS (matte)
Cus + Fe = Cue + FeS (Partial)
Finally, if there is a sufficient excess of alkali flux used, the iron
matte is dissolved by this basic slag, probably as a double sulphide
of iron and sodium or potassium.
From the equations it will be seen that copper, arsenic and antimony
are reduced, at least in part, and go into the lead button, orin the case ~
of arsenic form a speiss which ordinarily carries some of the precious
metals. In general it may be said that the process is not suited for
ores carrying much nickel, copper, cobalt, arsenic, antimony or
tellurium. One or two per cent of copper in an ore does not interfere
seriously with the assay, but when much more than this is present
some other method should be chosen. Ores containing nickel are
least of all suited to the process.
The slag made should not be more acid than a mono-silicate and
probably a sub-silicate is better for most ores. Occasionally a matte
is found-on top of the lead button and this generally contains more or
less gold and silver. It indicates a too acid slag, an insufficient
amount of alkaline flux or too short a fusion. The slag although
basic does not attack the crucible to any extent and crucibles may
ordinarily be used a number of times.
The method is a most excellent one on suitable ores and the author’s
experience has been that in nine cases out of ten, students will obtain
considerably higher results using this method than the niter method.
It has the advantage that no preliminary fusion to determine the
reducing power is necessary and that if the lead of the ore is allowed
for, a button of the proper size for cupellation may always be obtained.
As before mentioned the method is limited to pure ores and occasion-
ally a hard button, or speiss may be obtained when no copper, antimony
or arsenic was suspected in the ore. . Occasionally also, but only when
the slag is not properly constituted, or when the temperature of fusion
is too high there may be difficulty in separating the lead from the slag
and sometimes a thin film of lead may adhere to the slag when the
two are broken apart. The only other objection is the difficulty
of removing the nails free from shots of lead, but in general when the
fusion has gone far enough this will not cause serious inconvenience.
Procedure: Pan the ore, estimate and record its mineral composi-
109
tion. Note especially the per cent of lead minerals. Use a G pot
furnace or a 20 or 30 gram muffle crucible and weigh out one of the
following charges.
+ Galena
Pure Galena 3 Pyrite Pyrite
Ore Go.AtrL. HSS ple ei Ge O.55A, ol
Sodium carbonate 30 grams 40 grams 50 grams
Borax A eid in. 20 Oe a
Litharge 79 he aa ets
Silica Aaa Aa Sas
Nails;from 3 to 5 (twenty-penny) cut nails or preferably one 3”
to 4’’ track spike inserted point downward.
Cover Salt or borax-soda mixture.
Heat gradually to fusion, fuse from 40 to 60 minutes. Examine
the nails occasionally and if badly eaten add several fresh ones, leav-
ing the old ones in the crucible if they cannot be removed free from
lead. Fuse until the nails may be freed from lead by tapping them
gently and washing them around in the slag. Remove all nails and
pour as usual. The slag will be black and should separate easily
from the lead button.
Notes. 1. If the ore contains two or more grams of silica none need be added.
2. If bicarbonate of soda is substituted for the normal carbonate use a currespond-
a ingly greater weight.
3. This fusion requires a somewhat longer time than the niter fusion owing to
the fact that time must be allowed for all of the charge to come in contact with the
surface of the iron nails.
4. The lead may not start to drive in cupelling quite as rapidly as other buttons
owing to a small amount of iron which is often present.
The Roasting Method. This method of assaying sulphide ores
is rarely used, but might be used to advantage on very low grade
pyritic ores, so will be briefly described.
| Procedure: Take from 0.5 to 5.0 assay tons of ore and spread out
‘in a well chalked roasting dish of sufficient size to allow of stirring
without loss. Have the muffle at a dull red only and the fire so low
that the temperature of the muffle may be held stationary or raised
but slowly. Place the dish in the muffle and cover it if the ore con-
- tains minerals which decrepitate and keep it covered until danger
from this source is passed. The ore should soon begin to roast.
When fumes are noticed coming from the ore, check the fire and hold
it at this temperature for some time stirring frequently. After all
danger of fusing is over gradually raise the temperature stirring at
intervals of 20 minutes or one-half hour. Finally heat to about 700°
GC. for one-half hour, when if the ore contains only sulphides of iron
and copper, practically all of the sulphur will be removed. If there
110
is any doubt about the roast being complete, remove from the muffle,
add a small amount of charcoal and see if any odor of sulphur dioxide
is noticed. If the ore contains zinc, a much higher temperature will
be required to break up the zine sulphate. It is not best, however,
to carry the roasting temperature above 700° C.
If the ore is principally galena or stibnite, add an equal weight of
fine sand or assay silica before commencing the roasting, which should
be done at a very low temperature to prevent the fusion of the sul-
phides.
If the ore contains arsenic or antimony, the roasting operation is
more difficult. The best conditions for the elimination of these ele-
ments are alternate oxidation and reduction at a low temperature.
The presence of sulphur aids in the elimination of these elements due
to the fact that their sulphides are volatile. To obtain the reducing
action necessary for the elimination of arsenic and antimony take the
partially roasted from the muffle, allow it to cool for a few moments,
and then mix powdered charcoal or coal dust with it and roast at a
dull-red heat until the coal is burned off. Then add more coal and
re-roast. Repeat this until no more fumes of arsenic or antimony
are noticed, then heat with frequent stirring to about 700° C.
After the ore is roasted, the dish is carefully cleaned out and the |
ore is charged into a crucible with fluxes and treated exactly as a
class 1 ore. If the sulphide mineral was mostly iron, the ore will
probably be found to have a slight oxidizing power due to the forma-
tion of FeO; and Fe;Q, in the roasting.
The roasting method of assaying is slow and takes up much muffle
space. It is open to the liability of serious mechanical and vola-
tilization losses. Its most useful field would seem to be the assay
of low-grade pyritic gold ores where a very accurate determination
of gold is desired. The method usually gives low results in silver. 3
The combination wet and fire assay is used principally for the de-
termination of gold and silver in copper and nickel matte, copper
bullion, ete. A description of the method will be found in the chapter
on bullion assay.
Assay of Class 3 Ores.
The principal ores belonging to this class are those containing some ©
of the higher oxides of iron or manganese, i. e. Fe,O3, Fe3O1, MnQOsz.
These are reduced by carbon and tend to enter the slag as ferrous and
manganous silicates respectively. If a charge was made up for these
ores using only the ordinary amount of argols this might be all used
111
up in reducing the oxides of the ore and no lead button would result.
To remedy this the oxidizing power of the ore should be known be-
fore making up the charge.
To determine the oxidizing power of an ore, fuse a known weight
of it, say 10 or 20 grams with a regular crucible charge for that amount
of ore and a carefully weighed amount of argols of known reducing
power sufficient to more than oxidize the ore. The weight of lead
then, that the argols could reduce from an excess of litharge, minus
the weight of lead obtained is evidently the amount oxidized by the
ore. This weight divided by the weight of ore taken gives the oxidiz-
ing power.
Having determined the oxidizing power of the ore the assay is
made in the same manner as for class 1 ores with the addition of the
extra argols required.
The following table shows the proper size of crucibles for different
charges.
TABLE XXII. SIZE OF CRUCIBLES FOR VARIOUS CHARGES.
SSS a
ae ; ' Furnace
ES aa a Character of Ore
Ore Taken Pot Muffle
Age asl Siliceous ia 15 or 20 gms.
ma Ad 4 G 20 or 30 “
SAT iM H 30 or 35“
eG A. TL i Ix
ee AST Basic. Iron or Niter Fusions | G 20 or 30 “
2 1 A. T (4 res ‘ (<3 ce | H 30 66
CHAPTER IX.
SPECIAL METHODS OF ASSAY.
The Assay of Telluride Ores. The determination of the precious
metals in ores containing tellurium has always been considered more
than ordinarily difficult. Results obtained by different assayers
and even duplicate assays by the same man were often widely diverg-
ent. The literature of telluride ore assaying is extensive and none
too satisfactory; however, it is safe to say that most of the reported
differences between duplicates and different assayers have been due
more to difficulties in sampling than to the chemical interference of
the element tellurium. When it is considered that most of the tellur-
ide ores which are mined contain less than 0.1 per cent telluride
mineral, it is apparent that more than ordinary care must be taken -
to insure obtaining a fair proportion of this in the final assay portion.
The telluride mineral itself may contain 40 per cent of gold, so that
one 100 mesh particle more or less in the assay portion may make a
difference of several hundredths ounces of gold to the ton. To ob-
viate as far as possible this lack of homogeneity, all telluride ores
should be pulverized to at least 150 and preferably 200 mesh and
then very thoroughly mixed before the assay portions are weighed
out. |
Effect of Tellurium. Tellurium is a close associate of both gold and
silver and is difficult to separate from these metals either in the
crucible, scorification or cupellation processes. It is not however
often found in abundance, and even in high grade ores tellurium it-
self is found in comparatively small amounts. For instance, in two
high grade ores used by Hillebrand and Allen in their experiments on
the assay of telluride ores, containing respectively 15 and 19 oz. of
gold per ton, there was tellurium amounting to 0.074 and 0.092 per
cent respectively. It seems unreasonable to expect such small quan-
tities of any element to influence seriously the results of a fire assay.
In order to study the effects of tellurium in the gold and silver assay
it is necessary to experiment with ores or alloys containing much more
tellurium than those above mentioned. The following facts regarding
the behavior of tellurium in cupellation and fusion are mostly due to
113
the work of Holloway,' Pease’ and Smith, whom we have to thank
for co-ordinating and elucidating much information which was hitherto
much scattered and of doubtful value.
Hffect of Tellurium on Cupellation. The presence of tellurium in
a lead button causes a weakening of the surface tension of the molten
metal. The result is the metal tends to “‘wet’’ the surface of the cupel
and this allows particles of alloy to pass into the cupel and others to
be left behind to cupel by themselves on its surface forming minute
beads. In the case of a button containing 10 or more per cent of
tellurium with an equal weight of gold or silver, complete absorption
may take place. As the proportion of lead in the alloy is increased,
the amount of absorption becomes less, until when the lead: amounts
to 80 times the tellurium very little loss of precious metal occurs in
a properly conducted cupellation. (Smith).
Tellurium is removed comparatively slowly during cupellation
particularly in the early stages, as might be expected on comparing
the heat of formation of its oxide with that of lead oxide. Roses
gives the following figures for the heat of combination of these metals
with 16 grams of oxygen,—Pb to PbO 503 Cal., Te to TeOz 386 Cal.
_ To avoid danger of undue loss in cupellation of buttons from the assay
of such ores, as much as possible of the tellurium should be removed
prior to cupellation. It is also evident that large lead buttons (30
or more grams) should be allowed for in order that the ratio of lead
to tellurium be high.
Silver in the alloy protects gold from losses due to the presence of
tellurium. It appears to act as a dilutant for the gold and should
always be added to every gold assay for this reason if no other.
In the case of imperfect cupellation, tellurium is retained by the bead
and gives it a frosted appearance. In perfect cupellation the final
eondition of the tellurium is that of complete oxidation to TeOs.
Owing to its effect in reducing surface tension, resulting often in minute
beads being left behind, it would be well to use a cupel having a finer
surface when cupelling buttons containing tellurium. Smith states
that the loss due to sub-division and absorption in this case is much
less when a “patent” (magnesia) cupel is used. Losses of gold and
silver by volatilization during properly conducted cupellation of lead
buttons from ordinary telluride ores is extremely small.
Effect of Tellurium in Fusions. Tellurium was formerly thought to
1 The assay of Telluride Ores. G. T. ESTEE and L. E. B. Pease, Trans.
I. M. M., 17 p. 175.
2 The Behavior of Tellurium in Assaying, Bynes W. Smith, Trans. I. M. M.,
17 p. 463.
3 Trans. Inst. Min. & Met. alt p. 384.
114
be oxidized to the di-oxide during fusion and to go into the slag as
a sodium or lead tellurate. Smith disagrees with this and argues
that tellurates are decomposed at a red heat, and that lead tellurate
is white, while he found the litharge slags obtained in the fusion of
telluride compounds to be black. He believes that tellurium exists
in the slag as the black monoxide (TeO).
The slag best suited to the oxidation and retention of tellurium
in crucible assaying is a basic one containing a considerable excess
of litharge. The temperature of fusion should be moderately low
as a high temperature prevents the satisfactory oxidation and slagging
of the tellurium, probably owing to the formation of lead silicates
before the litharge has had time to oxidize the tellurium. Smith
gives the following reaction for the oxidation of tellurtum :—
2 PbO + Te = PhO + TeO 7
In support of this he claims to have found the black sub-oxide of
lead in the slag.
Practically all authorities agree that the scorification process is
not reliable for telluride ores. When a button from a crucible assay
contains too much tellurium for direct cupellation Smith recommends
fusing or ‘soaking’ the button under an ample amount of litharge
at a moderate temperature (700—900° C.).
Hillebrand and Allen used the following charges for ores containing
from 15 to 19 oz. gold and 0.074 to 0.092 per cent tellurium.
Ore TPAC Litharge 1380 grams
Sodium carbonate 30 grams Reducing agent for 25 gram buttons
Borax-glass The Silver 24 to 3 times gold.
They find the slag losses no higher than with ordinary gold ores
and no serious cupellation losses. With ores containing much more
tellurium than the above, the quantity taken should be reduced and
the rest of the charge maintained as before.
The Assay of Ores and Products High in Copper. Crucible
methods for the assay of matte and ores high in copper have largely
supplanted the older scorification method. This is due to the fact
that a larger amount of pulp may be used for each individual assay,
thus increasing the accuracy of the results. The copper is eliminated — 4
as it is in the scorification assay by the solution of its oxide in the basic
lead oxide slag. The assay thus combines the advantages of the scori-
fication with those of the crucible assay.
Perkins ' has made a careful study of this process, and calls atten-
tion to the fact that the litharge used must be in proportion to the _
1 The Litharge method of Assaying Copper Bearing Ores and Products, and the a
Method of Calculating Charges. W. G. Perkins, T. A. I. M. E., 31 p. 913.
is 3
amount of copper and other impurities in the ore. The amounts he
uses are very large (from 137 to 300 parts PbO to | part Cu), and make
the method an expensive one. Others have reduced this amount
considerably, and still manage to get buttons which will cupel.
The Slag. The slag should be decidedly basic, for if we combine
the litharge with large amounts of silica-and borax, it will no longer
retain its power of holding the copper in solution. A small amount
of silica is necessary to prevent to some extent the action of the
litharge upon the crucible. One part of silica to from 15 to 20 parts
of litharge is generally allowed in the charge. Borax should be en-
tirely omitted as it acts to decrease the copper holding capacity of
the slag, and also causes boiling of the charge. Perkins states that
the best results are obtained with a slag which exhibits when cooled
and broken a somewhat glassy exterior gradually passing to litharge
like crystals towards the center. The amount of crystallization which
takes place is, of course, a function of the rate of cooling and will
depend among other things upon the size of the charge, the tempera-
ture of the charge when poured, and of the mold, so that too much
weight should not be given to the above. ‘The slag should however
be crystalline resembling litharge, and if dull or glassy throughout,
indicates the presence of too much acid for a good elimination of
— copper.
Conduct of the Assay. On account of the very corrosive action of
the litharge slag it is especially necessary that the fusion be made
rapidly. The muffle should be hot to start (1000° to 1100° C.),
the hotter the better, and the fusion should be finished in from 20
to 30 minutes. This not only preserves the crucibles, but also as a
necessary sequel prevents the slag from becoming charged with silica
and thus forcing the copper into the button. The slag melts at a low
- temperature and a very high finishing temperature is not necessary.
With a quick fusion there is less chance for oxidation of lead with the
consequent reduction of too small a lead button.
For the best work the hole in the back of the muffle should be
stopped up, and a reducing atmosphere maintained in the muffle.
This may be accomplished by filling the mouth of the muffle with
charcoal or coke, or by distributing a few crucibles part full of soft
coal throughout the charge and using a tight-fitting door. If this
precaution is not observed part of the silver will be oxidized and lost
in the slag.
The following charges. kindly furnished by the Boston and Mon-
tana Reduction Department of the Anaconda Copper Mining Com-
pany, Great Falls, Montana are recommiended for these ores.
116
TABLE XXIII. CHARGES FOR COPPER BEARING MATERIAL.
Approximate Charge for Silver Charge for Gold
Material Analysis (In 20 gram crucible) | (In 30 gram crucible)
Cu 9%-15% Sample 1 A.T. | Sample Lage
SiO 15% 23% Soda 20 grams | Soda 30 grams
FeO 33 %—-40% Litharge 100 “ Litharge 150 ‘“
Concts. S 33 %—40% Silica yee Silica olen
Ag 3 02.—5 OZ. NitsE) Lo 20s Niter 40-60 “
Au 0.0150z.-—0.0250z.| Cover mixture Cover mixture
Cu 30%—-45% Sample SAE: Sample 2 A. T.
Fe 40 %-30% Soda 18 grams | Soda 25 grams
Matte S 30%-27% Litharge 100 “ Litharge 200 “
Ag 10 0z.-18 oz. Silica Te Silica IDivke
Au 0.07 oz.-0.11 oz. | Niter ae Niter 1S
Cover mixture Cover mixture
Cu 45%-60% Sample 1A. T. Sample A
Fe 30Z-15% Soda 18 grams -| Soda 25 grams
Matte 8 27%-24% Litharge 125 “ Litharge 240 “
Ag 15 0z.-25 oz. Silica (jee Silica 1a
Au 0.10 oz.—0.14 oz. | Niter 7 hee Niter 14505
Cover mixture Cover mixture
—
Assay of Antimonial Gold Ores. The niter method-is universally
recognized as being the best method for the sulphide ores of antimony. -
Considerable litharge is necessary to keep the antimony out of the
lead button. The following charge is recommended by two English
authorities: —
Ore 5A. T. Litharge 100-120 grams
Na,CO; 10-20 grams Niter LO ee
Borax-glass 5-10 “ Silica 10 Say
A preliminary assay to determine the reducing power is of course
necessary. The above charge will be found to correspond almost
exactly with our standard for sulphide ores, with litharge according
to Lodge’s rule.
George T. Holloway in discussing this method recommended using
a much larger proportion of soda in the charge, 1.e., three times as
much as stibnite, in order to aid in the retention of the antimony in
the slag as a sodium antimoniate.
Assay of Auriferous Tinstone. C. O. Bannister’ finds a crucible
assay with the following charge to be the most satisfactory
method :—
' William Kitto, Tr. Inst. of Min. & Met., 16 p. 89.
' William Smith, Tr. Inst. of Min. & Met., 9 p. 332.
2 Trans. Inst. of Min. and Met. (London) 15 p. 513.
Shoe ee ee tess 25 grams
Memmmacarpondte..;...2..2i7..2 > 40° “
eh). ho Sok ee, LOR?
eM are So 4 ober ee
mmr at Gg a eS Lo yee
In this method the tin is converted into a fusible sodium stannate.
The author found no tin reduced during the fusion as shown by the
button cupelling without difficulty. In all ores carrying over 1 oz.
of gold per ton, the slags were cleaned by a second fusion with 10
grams of soda, 30 grams of red lead and 1.5 grams of charcoal.
Various other methods of assay were tested but none were as satis-
factory as this.
Corrected Assays. In the assay of high-grade ores and bullion
it is often desirable to make a correction for the inevitable slag and
cupel losses. This is done in one of two ways: either by the use of
a “check” or synthetic assay or by assaying the slags and cupels re-
sulting from the original or commercial assays.
In correcting by a ‘“‘check’”’ assay a preliminary assay is first. made
and then an amount of proof silver or gold, or both, approximately
equivalent to the amount present in the sample, is weighed out and
made up to approximately the composition of the sample by the ad-
dition of base metal, etc. The check thus made is assayed in the
same furnace parallel with the real assay. Whatever loss the known
amounts of precious metal in the check sustain is added to the weight
of metal obtained from the sample as a correction, the sum being sup-
posed to represent the actual metal present in the sample. This
method of correction is always used in the assay of gold and other
precious metal bullions, and is sometimes used in the assay of high-
grade ores. A more detailed description of the method will be found
in the chapter on the assay of bullion. This method when properly
applied is the better and gives a very close approximation to the actual
precious metal contents of a sample.
In the case of rich ores and furnace products other than bullion,
a correction is usually made by assaying the slags and cupels result-
ing from the original assay. The metals thus recovered are added
as a correction to the weight first obtained. This method, while
approximating the actual contents of an ore, may occasionally give
results a little too high, for although gold and silver lost by volatiliza-
tion is not recovered and the corrections themselves must invariably |
suffer a second slag and cupel loss, yet on the other hand, the cupelled
metal from both the first and second operations is not pure and may
retain enough lead and occasionally other impurities from the ore
118
to more than offset the above small losses. The results of assays cor-
rected by this method are evidently somewhat uncertain, but are
nevertheless much nearer to the real silver content than are the re-
sults of the uncorrected or ordinary commercial assay.
Smelter contracts are almost invariably still written on the basis
of the ordinary or uncorrected assay and when the corrected assay
is made the basis of settlement, a deduction is made amounting to
the average correction. This amounted to 1.1 per cent in the case
of certain Cobalt ores.
When a corrected assay is to be made it is well to use a Portland
cement or magnesia cupel for the first cupellation as these materials
-are easier to flux than bone-ash.
To assay a Portland cement cupel the following charge, a sesqui-
silicate, is found to give satisfactory results :—
Cupel (45 grams of cement).
Na2zCO;—45 grams,
Borax-glass—21 grams.
Litharge—Dependent on size of original button,
to make a total of 75 grams.
Argols (R. P. 10) 3.2 grams.
Silica—32 grams.
To assay a magnesia cupel, good results are obtained by adding
soda equal to the original weight of the cupel and litharge to make a
total of 30 grams more than the original weight of the cupel (allow-
ing for litharge in the cupel). Compute the silica necessary to make
a sesqui-silicate with magnesia, soda and active litharge, and add
two-thirds of this weight of silica and substitute for the other third
twice its weight of borax-glass. )
To assay a bone-ash cupel, first remove and reject the unsaturated
part of the cupel in order to have as little of this refractory material
as possible to deal with. The saturated part will be about 50 per
cent bone-ash and 50 per cent litharge. Grind to 80 mesh and clean —
the bucking board or machine by grinding 10 or 15 grams of 10 mesh q
silica. This should be reserved and added to the charge. To assay,
add a weight of soda equal to the weight of saturated cupel material, —
two-thirds as much borax-glass, 25 grams of litharge plus enough ©
more to make a total equal to the weight of saturated cupel material, —
silica one-third as much cupei material, reduane agent for a es gram
lead button. For example:— :
Cupel material}. : 45 grams Litharge 47% grams
od askOave tok aie ee Argols (R. P.10) 2.2 “
. Borax-glass . : pOaia. Silica (from clean-
ing board) tbo
CHAPTER X.
THE ASSAY OF BULLION.
Bullion from an assayer’s point of view is an alloy containing enough
of the precious metals to pay for parting.
The different bullions are usually named to correspond with their
major components, for instance, copper bullion an alloy of copper
with small amounts of other impurities, as well as some gold and silver.
In the same way we have lead, silver and gold bullions. Doré bullion
is silver bullion containing gold. The term base bullion is used in
two different senses. According to the lead smelters definition base
bullion is argentiferous lead, usually the product of the lead blast
furnace; according to the mints and refiners definition it is bullion
containing from 10 to 60 per cent of silver, usually some gold, and a
large percentage of base metals particularly copper, lead, zine and
antimony. Fine gold bars are those which are free from silver and
sufficiently free from other impurities to make them fit for coinage
and use in the arts usually 990 to 999 fine.
The results of lead and copper bullion assays are reported | In ounces
per ton as in the case of ore assays, but in the assay of silver, gold and
doré bullions the results are reported in “‘fineness,” i.e., so many
parts of silver or gold in one thousand parts of bullion. Thus sterling
silver is 925 parts fine, that is to say, it is 92.5 per cent silver,
Weights. In assaying gold, silver and dore bullion, a special
set of weights called gold assay weights are used. This is termed
the ‘“‘millime”’ system, and the unit one millime weighs 0.5 milligram,
and therefore the 1000 millime weight equals 0.5 grams. Ordinary
weights in the gram system may be used but as 0.5 gram is the quan-
tity of bullion commonly taken for assay the use of the millime system
saves computation in obtaining the fineness.
Sampling Bullion.
Bullion may be sampled either in the molten or in the solid condi-
tion. When it may be melted and kept free from dross the dip or
ladle sample is usually the more accurate method. As the weight,
as well as the assay of the bullion must be known in order to value
it, the sampling of large lots of bullion by the dip sample method
120
often presents difficulties owing to changes in weight or purity in
the considerable length of time necessary for pouring. Again it is
not always convenient to melt a lot of bullion to obtain a sample,
and other means must be found. Sampling solid bullion by punch-
ing, boring, sawing or chipping, under certain conditions, may be
made to yield good results. Lead bullion is usually sampled by punch-
ing one or more holes in each bar, and combining and melting the
punchings. Copper bullion is now generally cast in the form of slabs
or anodes, and these are drilled.
Sampling Molten Bullion. The most satisfactory method of
sampling bullion is to melt the whole in a suitable vessel, stir
thoroughly with a graphite rod or iron bar to mix and then immediately
before pouring, ladle out a small amount and granulate it by pouring
into a pailof water. If these operations are correctly performed there is
no chance for segregation, and each particle of the granulated metal
should be a true representative of the whole. If a granulated sample
is not desired, a ladleful of the mixed molten metal may be poured
into a thick-walled flat mold so that it chills almost instantly, and a
drill or saw sample may be taken from this. When a ladle sample is
taken, the ladle must be so hot as not to allow the forming of any
solidified metal or “‘sculls’’ as this would interfere with the homogeneity
of the sample. This method of sampling is most satisfactory on
bullons which do not oxidize or form dross on melting, as this of
course, adds a complication hard to allow for.
Sampling Solid Bullion. The principal difficulty in the sampling
of bullion in the form of bars or ingots is caused by the segregation
of the various metals in cooling. If it were, possible to cool a bar
instantly, segregation would be prevented, and a chip or boring taken
from any part would be representative. As instant cooling is im-
possible, the sampling of bars of the ordinary dimensions becomes a
difficult problem. As the result of a careful study of this problem
Keller' has concluded that it is almost impossible to obtain samples
of satisfactory accuracy from bars or pigs of the usual dimensions.
To eliminate the difficulties of sampling from a bar he recommends
casting the metal in the form of a thin plate. Of course some con-
centration would take place here also, but as the plate would solidify
so much faster than the same metal cast in a bar or ingot this factor
would have less weight. Owing to the fact that concentration takes
place from or toward every surface, we will have all around the plate
a zone not wider than the thickness of the plate where concentration
has taken place both horizontally and vertically, but which should
1 'T. A. I. M. E,, 27, p. 106.
121
of itself be a sample of the whole. In the part of the plate enclosed
by this zone we have concentration in the vertical direction only.
If we drill or punch through this part of the plate we should obtain
a correct sample of the whole. Keller cites experiments to prove the
above theory.
Some typical methods of sampling lead and copper bullion follow.
Sampling Lead Bullion. Lead bullion is sampled both in
the liquid and in the solid state. In either case it is now customary
to transfer the lead from the blast furnace either into a reverberatory
furnace or into large kettles holding 20 to 30 tons. Here it is purified
either by liquation, or by cooling to a little above the melting point of
pure lead. By doing this, a large part of the impurities which are
held in solution by the superheated lead are separated out as a dross
which is carefully removed by skimming. The remaining lead, which
is now in a better condition to sample, is drawn off by means of a
syphon and cast into bars of about 100 pounds.
In taking a dip sample a small ladleful is taken at regular intervals
from the stream coming from the syphon. These individual samples
are carefully remelted at a dark red heat in a graphite crucible, the
melt is well stirred and cast in a heavy-walled shallow mold, making
a cake about 10’’ long, 5’’ wide and 4’’ thick. This cools so quickly
that there is little or no chance for segregation. The final assay
samples are taken from this cake either by sawing and taking the
sawdust, or by boring entirely through the slab in a number of places,
and taking the borings, or by cutting out four or more 3 A. T. pieces
from different parts of the bar and using these directly.
In sampling solid lead bullion the bars are sampled by means of a
heavy punch which takes a cylindrical sample about 2’’ long and
1/8’’ in diameter. There are naturally a number of different systems
but the most common method is to place five bars side by side and
face up, and punch a hole in each extending half-way through. Each
bar is punched in a different place and in such a way that the holes
make a diagonal across the five bars. The bars are then turned over
and another sample is taken from each along the opposite diagonal.
Usually one carload of about 20 or 30 tons is sampled as one lot. The
-punchings from such a lot, weighing from 8 to 15 pounds are melted
in a graphite crucible and cast into a flat bar, from which the final
assay samples are taken by sawing, drilling or cutting.
Sampling Copper Bullion. The sampling of copper bullion may
be classified into smelter methods, and refinery methods. The bullion
is quite universally cast in the form of anodes at the smelter, and
]
122
shipped to the refinery in this form. This renders remelting at the
refinery unnecessary, and the result is that the refiners sample the
solid bullion by drilling. The smelters, having the bullion in the
molten condition, generally sample it in this condition on account of
the greater ease and less expense.
Probably the most satisfactory smelter method of sampling is the
“splash shot method”’, which consists in shotting into water a small
portion of the molten stream of copper as it flows from the refining
furnace by “batting” the stream with a wet stick. This operation
is repeated at uniform intervals during the pouring, the amount taken
each time being kept about the same. The samples are dried and
dirt and pieces of burned wood are removed. All material over four
mesh and under 10 mesh is rejected, and the remainder taken as the ©
sample. This method when properly carried out gives results which
check within practical limits with the drill sample of the anodes taken
at the refinery.
Another method which is used to some extent for sampling molten
copper bullion is known as the ‘“‘ladle-shot method.” This consists
in taking a ladleful from the furnace or from the stream of the casting
machine and shotting it by pouring over a wooden paddle into water.
In this method at least three ladlefuls are taken, one near the begin-
ning, one at the middle, and one near the end of the pour. The shots
are treated in the same manner as before. This method is not thought
so well of as the previous one on account of segregation toward or
from the “ sculls ”’ which are left in the ladles.
Instead of shotting and taking the shot for the final sample, W. H.
Howard of Garfield, Utah, recommends ladling into a flat dise.
This ‘‘pie sample” is sawed radially a number of times, and the saw-
dust used for the final sample.
The following description of the method of sampling anodes at
Perth Amboy, N.J. is typical of refinery methods of sampling and is
the method developed by Dr. Edward Keller. The copper is re-
ceived in the form of anodes 36”’ long, 28’’ wide and 2”’ thick. These
are carefully swept to remove foreign matter, and then drilled with
a 0.5’’ drill completely through the anode, all of the drillings being
carefully saved. A 99-hole template is used to locate the holes which
are spaced 3 1/16’’ center to center, and the outside row is approxi-
mately 23’’ from the edge of the anode. The holes of the template
are used in continuous order, one hole to the anode.
For very rich anodes some refiners use a template having as many |
as 240 holes, but it seems doubtful if this arrangement of spacing a
single hole in each anode will yield any better sample.
——— tan
123
With low-grade, uniform bullion every fourth anode only is drilled.
A 30 ton lot of anodes in which each one is drilled will yield 6 or 8
pounds of drillings, which are ground in a drug-mill fitted with man-
ganese steel plates and reduced by quartering to about 2 pounds.
This sample is reground until it will all pass a 16-mesh screen and is
then divided into the sample packages.
The Assay of Lead Bullion.
A description of the cupellation assay of lead bullion has already
been given in the chapter on cupellation. In smelter control work
the assay is usually made in quadruplicate. If the bullion contains
sufficient copper, arsenic, antimony, tin or other base metals to
influence the results of the cupellation assay, three or four portions
of 0.5 or 1.0 A. T. are scorified with the addition of lead until the
impurities are eliminated, when the resultant buttons are cupelled.
Correction for Cupel Loss. In some instances the slags and cupels
are re-assayed and the weight of the gold and silver found is added
to that obtained from the first cupellation. There is no fixed custom
as yet regarding the use of corrected assays. In most of the custom
smelters, the uncorrected assay is used as the basic of settlement;
but some of the large concerns who have their own refineries are using
the corrected assay in their inter-plant business.
The Assay of Copper Bullion.
Copper bullion may be assayed by the scorification, crucible or
by a combination of wet and fire methods. In the combination method
the bullion is treated with sulphuric or nitric acid which dissolved
the copper and the silver and leaves the gold. The silver is pre-
_cipitated by suitable reagents and filtered off together with the gold.
The filter paper and contents are put into a scorifier or crucible with
reagents and the assay finished by fire methods.
The scorification method is generally accepted as standard for gold,
and many smelter contracts state that ‘gold shall be determined by
the all fire method or its equivalent.’’ The mercury-sulphuric acid
combination method on many bullions gives gold results equal to
the scorification. The silver results obtained by the scorification
method are open to suspicion owing to the considerable slag and
~ cupellation losses, and the doubt concerning the purity of the buttons,
which often contain noticeable amounts of lead and copper.
~The crucible method has not as yet come into common use for the
determination of gold and silver in copper bullion, but according to
124
Perkins! it gives gold results equal to the “all-scorification‘’ method.
In smelter practice, silver in copper bullion is determined usually
by the nitric acid combination method, sometimes by the mercury-
sulphuric acid combination method. This later method tends to
give high silver results, owing to the incomplete solution of the copper
in the acid, and the possibility of some copper being retained in the
silver bead.
The nitric acid combination method is recognized as giving low
results in gold. Van Liew’ attributes this to the solution of the gold
in the mixture of nitrous and nitric acids present. He found a decided
loss (33.7%) of gold on treating gold leaf with a mixture of nitrous
and nitric acids for two and a half hours. He gives a method of slow
solution in cold dilute acid which reduces this loss to a minimum.
The Scorification Method. The following method commonly
referred to as the “‘all fire’’ method is a modification kindly supplied
by Mr. H. D. Greenwood, Chief Chemist for the United States Metals
Refining Co., Chrome, N. J.
Sample down the finely ground bullion on a split sampler in such a
way as to obtain a sample of about 1 A. T., which will include the
proper proportion of finer and coarser parts of the borings. This
sampling must be conducted carefully, as the precious metal contents
of the finer parts differs somewhat from that of the coarser portion
of the sample. Portions “dipped” from the sample bottle or from the
sample spread out on paper are likely to contain undue amounts of
coarse or fine.
Weigh out 4 portions of copper borings of } assay ton each, mix with
50 grams test lead, put in 38-inch Bartlett scorifiers, cover with 40
grams test lead and add about 1 gram SiOs. Scorify hot, heating
at finish so as to pour properly. Add test lead to make weight of
button plus test lead equal to 70 grams, add 1 gram SiO, and scorify
rather cool. Pour, make up to 60 grams with test lead, adding 1
gram SiO: and scorify.
Combine the buttons two and two, and make up each lot to 85
grams with test lead, adding | gram SiO» and scorify very cool. Make
up button to 70 grams by adding test lead, add 1 gram SiO. and
scorify for the fifth time. The buttons should be free from slag and
weigh 14 grams.
Cupel at a temperature to feather nicely, and raise heat at finish.
Cupels to be made of 60 mesh bone-ash, and to be of medium hard-
ness.
PPA. LANA Boo ppuonl:
2E. & M. J. 69, pp. 496, et seq.
a my cafe eee
Pee | wee oe
125
Weigh the bead and part as usual. Dry, anneal and weigh. The
two results should check within .02 oz. per ton, and the average
figure is to be reported. If the silver contents of the bullion is low,
add enough fine silver to the copper borings before the first scorifica-
tion to make the total silver in the mixture equal to about 8 times the
amount of gold.
The Crucible Method. The crucible method for gold and silver
in copper bull.on was first described by Perkins! and as described by
him showed no great advantage over the scorification method as to
saving in time or cost of materials, or increased furnace capacity.
The following modified procedure requires about one-third of the
materials, time and furnace capacity as that described by Perkins,
and yet gives buttons sufficiently free from copper so that they may
be cupelled direct.
Sample down the finely ground bullion to about + A. T. and adjust
the weight of the sampled portion to exactly + A.T. Place in a
20 gram crucible and mix with it 1.2 grams of powdered sulphur.
Cover this with a mixture of 15 grams of sodium carbonate, 240
grams of litharge, and 8 grams of silica; but do not mix with the sul-
phur and copper which should be allowed to remain in the bottom of
the crucible. Cover with salt or flux mixture and place in a hut muflie
so that the charge will begin to melt in 6 or 8 minutes. ‘lhe fusions
should be quiet and ready to pour in 25 or 30 minutes.
If a salt cover is used the lead buttons should weigh abeut 32 grams,
if a flux cover is used they may be somewhat smaller. With a properly
conducted assay the buttons are soft enough for direct cupellation;
but the cupels are quite green. If the assayer prefers, the buttons
may be made up to 50 or 60 grams with test lead and scorified in a
3 inch scorifier to further eliminate the copper. After cupellation
_ the buttons are weighed and parted as usual. It is well to do four
fusions, and to combine the buttons two and two for parting.
Remarks. As soon as the sulphur melts (115° C.) it combines with
the copper to form a matte which is decomposed and most of its copper
oxidized and slagged by the litharge of the charge. The fusions melt
down very quietly almost without boiling, and with a short period
of fusion the crucibles are not badly attacked. The final temperature
need not be higher than a good bright red or full yellow. The slag
is heavy but very fluid and should not contain any lead shots.
The method gives results in gold equal to the scorification method;
1 An “All-Fire’’ Method for the Assay of Gold and Silver in Blister Copper.
W. G. Perkins, T. A. I. M. E., 33 pp. 670.
126
but like any method using high litharge, the silver is apue to be some-
what low. :
Mercury—Sulphuric Acid Method. Sample down the’ finely
ground bullion on a split sampler in such a way as to obtain a sample
of about 1 A. T., which will include a proper proportion of the finer
and coarser parts of the borings. This sampling must be conducted
carefully as the precious metal contents of the finer parts differs some-
what from that of the coarser portion of the sample. Portions
“dipped”’ from the sample bottle or from the sample spread out on
paper are likely to contain undue amounts of the coarse or fine.
Adjust the weight of the sampled portion to exactly 1 A. T. and
transfer it to a No. 5 beaker (capacity about 750 ¢c.c.). The beaker
should have a watch glass cover.
Treat the sample with 10 c.c. of mercury solution and shake the
beaker until the copper is thoroughly amalgamated; then add 80
c.c. of strong sulphuric acid, place the beaker on a hot plate and
boil until the copper is dissolved. This requires about twenty
- minutes.
Remove the beaker and allow it to cool. The contents will be a
semi-liquid sludge. When cool, add about 100 ¢.c. of cold water and
mix; then add about 450 c.c. of boiling water and stir until the copper
Bilphite dissolves.
Bring to boiling and add 4 to 6 c.c. of salt soins 1 c.c. equivalent
to 50 mg. Ag. Remove from the hot plate and add 10 c.c. of a 10
per cent solution of lead acetate. Stir well, allow to settle and filter
at once through double filter papers (124 or 15 c.m.) washing the
beaker with hot water. Finally wipe the inside with filter paper and
add it to the filter. Thorough washing of the filter is not necessary.
Transfer the wet filter and its contents to a 24’’ scorifier which
has been glazed on the inside by melting litharge in it and pouring
away the excess.
Burn off the filter paper at a low temperature—best in a closed oven
which may be heated to, say 175°C. This chars the paper slowly
without danger of loss of silver.
When the paper is consumed, add 30 grams of test lead and scorify;
pour so as to obtain a 12 gram button, cupel as usual to produce feather
litharge, weigh the gold and silver bead and part with dilute nitric
acid.
The mercury solution mentioned above is made by dissolving 100
grams of pure mercury in nitric acid, and diluting to 500 c.c. From
this stock solution take 50 c.c. and dilute it to 1 liter, for the working
|
:
:
%
:
:
:
127
solution—10 c.c. of the latter will serve for each assay-ton of copper
bullion.
The object of adding mercury is to secure an easy solution of the
copper in sulphuric acid. If the copper is treated directly without
previous amalgamation, it is very difficult to dissolve it in sulphuric
acid. In fact a considerable portion of it will remain insoluble,
partly in the form of sulphide of copper. If the copper be amal-
gamated on the other hand, solution proceeds smoothly until prac-
tically all of the copper is dissolved. When the bullion is low in
precious metals, say less than 50 ounces per ton, no silver dissolves
in the sulphuric acid. No gold dissolves whatever the grade. If the
bullion is very rich in silver a little of it may dissolve in the acid.
The assays should be made in duplicate or triplicate, and the aver-
age results reported. Differences on silver seldom exceed 0.2’ ozs.;
on gold the results are usually exactly the same. The sulphuric
acid used should be chemically pure and full strength (1.84 sp. gr.).
Nitric Acid Combination Method.’ Sample down the finely
ground bullion on a split sampler in such a way as to obtain a sample
about 1 A. T., which will include the proper proportion of the finer
and coarser parts of the borings. This sampling must be conducted
carefully as the precious metal contents of the finer parts differ
somewhat from that of the coarser portion of the sample. Portions
“dipped” from the sample bottle or from the sample spread out on
paper are likely to contain undue amounts of coarse or fine.
Weigh out two portions of copper borings of 1 A. T. each, and carry
the assay through on each portion as follows:— _
Place in a No. 5 beaker, add 100 c.c. of distilled water and 90 c.c.
HNOs, sp. gr. 1.42, the latter being added in portions of 30 c.c. each
at intervals of about 1 hour. When all is in solution precipitate a
small amount of silver chloride with salt solution in order to collect
the gold, filter through double filter papers and wash the filter papers
free from copper. To the filtrate add the calculated amount of salt
solution to precipitate all the silver and a slight excess, measuring the
solution with a burette and varying the amount added with the rich-
ness of the bullion. Allow to stand over night after stirring well.
Filter the silver chloride through double papers, wash papers free
from copper, then sprinkle 5 grams of test lead in the filter paper and
fold into a 234 inch Bartlett shape scorifier, the bottom of which is
lined with sheet lead. To this add also the filter papers containing,
“1 Procedure kindly supplied by Mr. D. H. Greenwood, Chief Chemist, for the
United States Metals Refining Company, Chrome, N. J.
128
the gold. Dry and ignite the filter papers carefully, cover with
35 grams of test lead, a little borax-glass, and scorify at a low heat
so that the resultant button will weigh about 12 grams. Cupels —
should be feathered nicely. Cupel to be made of 60 mesh bone-ash
and to be of medium hardness. Weigh the bead and part. Anneal
and weigh the gold. The two results on gold should check within
0.02 oz. per ton, and the silver within 1 per cent.
The Assay of Dore Bullion.
This method is the one generally adopted by assayers in this coun-
try, and may also be used for the assay of silver bullion. A better
method for the accurate determination of silver in doré or silver bullion
is probably the Gay-Lussac or salt titration, also known as the mint —
method. This later method requires considerable equipment and
preparation and for this reason the occasional assay is more easily
done by fire methods.
The Check. In order to correct for the inevitable losses in cupelling
as well as for any other errors in the assay, silver, doré, and gold
bullions are always run with a check. This check or “proof center”
is a synthetic sample made up of known weights of pure silver, gold
and copper to approximate as closely as possible the composition of
the bullion to be assayed. It is cupelled at the same time and under
the same conditions as the regular assays, and whatever gain or loss
it suffers is added as a correction to the regular assay. To obtain
data to make up the check a preliminary assay is made to determine —
the approximate composition of the bullion.
Preliminary Assay. A sample of 500 mgs. of bullion or as near this
amount as possible is weighed out on the assay balance, and the exact
weight recorded. This is compactly wrapped in 6 or 8 grams of lead
foil and cupelled in a small cupel with feather crystals of litharge.
The cupel should be pushed back in the muffle for the last two or
three minutes to insure the removal of the last of the lead. After
the play of colors has ceased it should be drawn toward the front of
the muffle and then covered with a very hot cupel to prevent sprout-
ing. It is then removed gradually from the muffle and when cool
is cleaned, weighed and parted in the ordinary manner. The gold
will require more than the ordinary amount of washing on account
of the large quantity of silver present.
If the cupelling has been properly conducted it will be fair to assume
a loss of one per cent of silver in determining the approximate silver.
The weight of gold may be taken as approximately correct. The
ORO OT ee a | a ee ae ee oe
129
~ sum of the weights of approximate gold and silver is subtracted
from the weight of bullion taken to obtain the amount of base metal.
This will usually be copper, but whatever it is the assayer should be
able to determine by the appearance of the bullion and the cupel.
Final Assay. Two portions of approximately 500 mgs. are weighed
accurately and wrapped in the proper amount of lead foil as shown
by the following table which assumes the impurity to be copper.
TABLE XXIV. LEAD RATIO IN CUPELLATION.
Fineness of Au. + Ag. | Wt. of Lead Fineness Au. + Ag. Wt. of Lead
950 5 gms. 750 llgms
900 7 Os ' 700 12 66
850 8 650 ist:
800 PO eae s 600 15a
A check is made up with C. P. silver and proof gold equal to the
approximate silver and gold found by the preliminary assay and the
necessary amount of copper or other base metal. These are wrapped
’ up in the same amount of sheet lead as was used for the bullion. The
lead for these assays is best cut into equal sized rectangles with pro-
portions approximately 13’ X 23’’, and twisted intu (the whape of
little cornucopias with the bottoms folded up. The bullion aid metals
going to make up the check are transferred to these directly from the
seale pans after which they are folded over and made into compact
bundles.
The cupels are placed in a row across the muffle and when they are
hot the buttons are dropped quickly into them with the check in the
middle. They should be cupelled at a low temperature so that plenti-
ful crystals of litharge are obtained all around the buttons, but toward
the end the temperature should be increased to insure driving off.
the last of the lead.
The buttons are cleaned, weighed, parted and the gold weighed.
The per cent loss of gold and silver is determined and a corresponding
correction made to the weights of gold and silver found. From these
figures the fineness in both gold and silver is determined. The gold
should check within 0.1 part and the silver within 0.5 parts.
Notes. 1. When the doré contains antimony weigh the samples into 2.5’
scorifiers with 30 grams of test lead. The proofs are made up according to the
preliminary assay. All are scorified in the same muffle at the same time. Pour
and hammer the lead buttons into a cube. Should the weight of these lead buttons
vary over a gram, make up to the same weight with sheet-test-lead, cupel and
part as usual. / ; |
2. When the doré contains bismuth, selenium or tellurium, three 3 gram
130
portions are weighed out into 24’’ scorifiers with forty grams of test lead, scorified
and the lead buttons flattened out into a sheet about 3 inches square. This sheet
of lead is dissolved in about 200 c.c., of dil. HNOs (1-3) and boiled to expel all red
fumes. Dilute to 400 c.c., filter through triple folded 15 em. filter, washing the pre-_
cipitate only once. To the filtrate is added sufficient NaCl solution to precipitate
all the silver. Heat to boiling and allow to stand over night. Filter through 15 em.
filter washing the precipitate only once. Place the two filter papers in a 23’" lead
lined scorifier, dry in an oven, burn, then cover with 30 grams of test lead and
scorify. Open the scorification at a rather high heat, continuing with a gradually
falling temperature. When the scorifiers have entirely closed over, close the muffle
door, raise the heat and pour; then treat exactly as in No. 1.
If the silver fineness of the doré is not three or more times greater than the pole
fineness, another set of assays must be run with the addition of proof silver at the
weighing out of the doré. ;
U. S. Mint Assay of Gold Bullion.
Melting. Every lot of bullion or dust received at any U. S. Assay
Office or Mint is immediately weighed and given a number. It is
then sent to the melting room. Here it is melted in a graphite crucible
with borax and soda and cast into a bar. Usually no attempt is
made to refine it unless it is very impure. Occasionally, in the case
of very impure bullions, a small dip sample is taken and granulated,
but in general the whole melt is cast and sampled as noted below.
The slag is poured with the bar and when solid is ground, panned and -
the recovered prills are dried, weighed and allowed for in computing
the value of the bar. |
Sampling. After the bar is cleaned of slag it is dried, weighed and
numbered and from diagonally opposite corners two samples of 3 or
4 grams each are chipped. These are flattened with a heavy hammer,
annealed and rolled into sheets thin enough to be easily cut with
shears. The use of the shears can only be learned by practice, but
assayers become very skillful after a time so that it is no unusual thing
to see a bullion assayer weigh out five samples in almost as many
minutes.
Preliminary Assay. Assay for Bases. To determine the ap-
proximate composition of the bullion a preliminary assay is made.
A sample of 1000 gold weights (500 mgs.) is weighed out, wrapped in
five grams of lead foil and cupelled. The weight of the bullion taken,
less the weight of the button obtained gives the base metals.
The button now consists of gold and silver, the approximate relative
proportions of which must be determined. This may be done by
adding silver, cupelling and parting or by touchstone. This later
method is used at the Government Assay Offices and Mints. The —
touchstone consists of a piece of black jasper on which the sample is
rubbed and the mark compared with marks made with alloy slips
(needles) of known compositoin. The needles range from 500. to
131
1000 fine and are 20 points apart. This gives the fineness within
2 per cent which is close enough to show how much silver to add to
inquart the main assay and to make up the check or proof center.
Final Assay. The final assay is usually made by two assayers each
working on one of the chipped samples. In the case of a small bar
each makes one assay, while in the case of a large bar each assayer
makes two or more assays. The balances used for the assay are
usually adjusted so that a deviation of the needle of one division on
the ivory scale amounts to some simple fraction of the weights used.
Thus at one assay office a deviation of the swing of one division on
the ivory scale amounts to 0.1 mg. = 0.2 gold weights. With this
adjustment it is not necessary to make so many trials with the rider
to get the final weight, nor is it necessary to weigh out exactly an even
half gram of bullion for the assay. Instead we weigh out 1000 = 3
divisions on the ivory scale, record the difference, and make a cor-
responding correction when the gold cornet is weighed.
As stated above the weight of bullion taken for each assay is
1000 (500 mg.). To this is added sufficient silver to make the ratio
of silver to gold 2 to 1, and the whole is wrapped up in 5 or 6 grams
of lead foil. The lead foil pieces are all cut to exact size, about 14’’
23’’, and rolled up into cornucopia shape with the bottom pinched in.
The bullion is poured directly into these from the scale pan. ‘I'he
silver is added in the form of discs made for convenience into 4 or 5
different sizes. These discs are punched out of sheets carefully rolled
to gauge so that the punchings will weigh exactly even tens and hun-
dreds in the gold weight system. If the bullion contains no copper
it is advisable to add about 30 parts gold weight (15 mg.). This cop-
per may be alloyed with the silver used for parting.
One or more proofs of pure gold weighing usually 900 (0.450 grams)
are also weighed and made up to the 2 to | ratio and copper added to
approximate that in the bullion. These are wrapped in the same
quantity of lead foil as the bullion, and one or more are run in each row
of cupels in the muffle. The lead packets are pressed into spherical
shape by pliers specially designed for the purpose.
The lead packets are put in order as prepared in the numbered
compartments of a wooden tray and taken to the furnace room where
they are cupelled in a rather hot muffle. The cupels are surrounded
by a row of extra cupels so that the temperature may be kept as
uniform as possible for all the assays. The cupels are withdrawn
while the buttons are still fluid. With a two to one ratio of silver to
gold, and with copper present, there is no danger of sprouting.
The buttons are removed from the cupels by means of pliers and
132
carefully cleaned from all adhering bone-ash. They are then placed
on a special anvil and flattened by a middle and two end blows with
a heavy polished hammer. They are then annealed at a redheat and
passed twice through the rolls which are adjusted each time so that
after the second passage they are about 23’’ long by 4’’ wide, and
about as thick as an ordinary visiting card. It is important that the
fillets be all of the same size and thickness with smooth edges. They
are then re-annealed and each one is numbered on one end with small
steel dies to correspond with the number of the assay, and rolled up
into ‘“cornets” or spirals between the finger and thumb, with the
number outside. It is important that an even space be left between
all turns of the spiral in order that the acid shall have easy access
to all parts of the gold.
The cornets are parted in platinum thimbles which are supported —
in a platinum basket, and the whole thing is placed in a platinum
vessel containing boiling nitric acid of 32° B. (Sp. Gr. 1.28). They
are boiled for 10 minutes and then transferred to another vessel
containing acid of the same strength and boiled 10 minutes longer.
The basket and contents is then washed by dipping vertically in and
out in three changes of distilled water, drained, dried, and annealed
usually in the muffle.
When cold the cornets are ready to weigh. The gold should be
entirely in one piece, and the original numbers easily discernable on
the parted cornets. The proofs are weighed first and the corrections
applied to other cornets. The proofs always show a slight gain in
weight. The correction thus determined is termed the be
and is really the algebraic sum of all the gains and losses.
When more than 14 cornets are parted at one time the lot is given
a preliminary 3 minute treatment in an extra lot of acid followed by
the two regular 10 minute boilings.
The purpose of the copper which is added to the assays is to render
the button tough and permit of its being rolled out into a smooth
edged fillet. Without the copper, the fillet is apt to crack in rolling,
or to come through with a ragged edge which might give rise to a loss
in parting. The action of copper in this case is probably due to its
effect in aiding in the removal of the last of the lead in cupelling.1
The time required for cupellation is approximately 12 minutes.
1 Rose Trans. Inst. Min. & Met. 14 pp. 545.
ES
CHAPTER XI.
THE ASSAY OF SOLUTIONS.
A large variety of methods for the assay of gold and silver bearing
solutions have been published in the technical press, and quite a
number of these have been adopted by assayers. These methods
may be classified as follows:—
1. Methods involving evaporation in lead trays with subsequent
cupellation, or scorification and cupellation of the tray and contents.
2. Methods involving evaporation with litharge and other fluxes
followed by a crucible fusion and cupellation.
3. Methods in which the precious metals are precipitated and either
cupelled directly or first fused or scorified and cupelled.
4. Electrolytic methods in which the precious metals are deposited
directly on cathodes of lead foil, which are later wrapped up with
the deposit and cupelled.
5. Colorimetric methods (for gold only) all of which depend upon
obtaining the “‘purple of Cassius’’ color which may be compared with
proper standards.
Evaporation in Lead Tray. This method is a good one on rich,
neutral solutions containing only salts of the precious metals. A
tray of suitable size is made by turning up the edges of a piece of lead
foil. If many of these assays are to be made it is well to have a wooden
block as a form on which the trays may be shaped. A tray 2’’ X
2’’ x 2#’’ deep is about right to hold 1 assay-ton of solution.
Having made a tray which will not leak, the solution is added and
carefully evaporated to prevent spattering. The tray is then folded
into a compact mass and dropped into a hot cupel.
Among the disadvantages of the method are the following: It
does not permit of the use of a large quantity of solution, and there-
fore is suited only to rich solutions. If the solutions are acid they
will corrode the tray, and if they contain salts other than those of
gold and silver these will interfere with cupellation. As both AuCl,
and KAu(CN), are volatile at moderate temperatures, many assayers
- do not consider the method a reliable one for solutions of these salts.
Evaporation with Litharge. (First Method.) A measured
quantity of the solution is placed in a porcelain evaporating dish and
134
from 30 to 60 grams of litharge is sprinkled over the surface. The
mixture is allowed to evaporate at a gentle heat to prevent both
spitting and baking of the residue. When dry the residue is scraped
out, mixed with suitable fluxes, transferred to a crucible and fused in
the ordinary manner. The last portions remaining on the dish may
be removed by means of a small piece of filter paper slightly moistened
which is afterwards added to the charge.
Some assayers add a little fine silica and charcoal with ihe litharge.
The soluble constituents of a crucible charge, soda and borax, should
not be added to the solution as they form a hard cake which is difficult
to remove from the dish. The most important point in the process
_ is the proper manipulation of the temperature. If this is right there
will be no spattering and the dry residue will come away practically
clean from the dish after prying It up with the point of a spatula.
Evaporation with Litharge. (Second Method.) A measured
amount of solution is evaporated in a porcelain or enameled iron dish
to a small volume, so that when the litharge and silica of a crucible
charge are added, they will absorb practically all of the liquid forming
a thick paste. The heating is continued and the material is stirred
constantly to keep the residue granular, and to prevent it from stick-
ing to the dish. When dry the residue is cleaned out with the aid of
a spatula and a mixture of soda, borax and are argols placed with it
in a crucible which is heated to quiet fusion, poured and treated in
the usual way. .
This method requires more manipulation than the first one, and
the only advantage is in a possible hastening of the process.
If any residue sticks to the dish it may be removed by rubbaiee it
with a little fine silica on a piece of filter paper, the whole being after-
wards added to the charge.
A modification of the foregoing evaporation methods consists in
evaporating to a small volume without the addition of any reagents,
and then transferring the concentrated solution to a small dish of
very thin glass (Hoffmeister’s dish). The solution is evaporated
to dryness either with or without litharge, and the dish and contents ~
broken up directly into a crucible containing the usual fluxes. The
assay is finished in the usual manner. The advantage of this method
lies in the fact that there is no chance for loss of residue by not properly _
cleaning the dish, as the dish and all are fused.
The evaporation method while somewhat long, is the most reliable
and accurate one known, and is the standard with which all other
methods are compared. By arranging to allow the evaporation to
135
run over night, the samples taken one night may be assayed and re-
ported early next morning. The method is adapted to the treatment
of any quantity of solution and of almost any character. If the solu-
tion contains much sulphuric acid, the litharge may be converted
into lead sulphate which is unsuited either to act as a flux or to pro-
vide lead for a collecting agent. A fusion made on such a substance
using a carbonaceous reducing agent, will give either no button at
all, or a button of matte. The reaction between lead sulphate and
carbon is as follows: )
PbSO; + 2C = PbS + 2CO,
If the solution is one of AuCl; a little charcoal should be added dur-
ing the evaporation to insure the reduction and precipitation of the
gold, asin this way we avoid the danger of loss gold by volatilization
as the chloride. The gold being precipitated on the charcoal is in
the best possible position to be alloyed with the lead which will be
reduced by the carbon.
Precipitation by Zinc and Lead Acetate. The Chiddey
Method. (For Cyanide Solutions.) This method which was first
described by Alfred Chiddey ' is suitable for both gold and silver and
is used almost exclusively in this country for the assay of cyanide
solutions. It works equally well on strong or weak, foul or pure
solutions, and almost any quantity may be taken. Many changes
of detail have been suggested and innumerable modifications of the
original process are found described in the technical press. The
following method has been found satisfactory:
Take from one to twenty assay-tons of solution in a beaker or
evaporating dish and heat. Add 10 or 20 c.c. of a 10 per cent solution
of lead acetate containing 40 c.c. of acetic acid per liter. Then add
one or two grams of fine zine shavings rolled lightly into a ball. The
gold, silver and lead will immediately commence to precipitate on
the zinc. At first the solution may become cloudy but will soon clear
as more of the lead is precipitated. Heat, but not to boiling until
the lead is well precipitated. This usually takes about twenty or
twenty-five minutes. Then add slowly (about 5 c.c. at a time),
20 c.c. hydrochloric acid (1.12 sp. gr.), to dissolve the excess zinc.
Continue heating until effervescence stops. It is often found that
action ceases while there is still some undissolved zine remaining.
This is entirely covered and thus protected from the acid by the spongy
lead. To be sure that all the zinc is dissolved feel of the sponge with
a stirring rod and drop a little hydrochloric acid from a pipette di-
rectly on it.
1H. & M. J. 75, p. 478, March 1903.
136
As soon as the zine is dissolved decant off the solution and wash
the sponge two or three times with tap water. Next, moisten the |
fingers and press the sponge, which should be all in one piece, into a
compact mass. Dry by squeezing between pieces of soft filter paper
or by placing on a piece of lead foil and rolling with a piece of large
glass tubing. Finally roll into a ball with lead foil, puncture to allow
for steam escape, add silver for parting, and place in a hot cupel.
As soon as the zine is dissolved the assay should be removed from
the heat, and the sponge removed. If this is not done the lead will
start to dissolve and the sponge will soon break up. Washing by
decantation and manipulation with the fingers may appear crude, -
but after a little practice the operator becomes so proficient that
there is practically no chance of loosing any of the lead.
If any considerable amount of water is left the assay will split in
the cupel. To avoid this danger some assayers dry the assays on the
steam table before cupelling. Any zinc left will also probably cause
spitting. Chiddey recommends placing a piece of dry pine wood in
the mouth of the muffle immediately after changing the cupels,
probably with the idea that this aids to prevent spitting when some
zinc has been left undissolved. When working with small quantities
of solutions it is best to add water occasionally to maintain a volume
of at least 100-150 ¢.c. The secret of keeping the lead from breaking
up is not to allow the solution to come to a boil at any stage of the
procedure. ?
Zine dust is used by many chemists in place of zine shavings, a
small amount being added on the end of a spatula. Many chemists
agree that 4 gram is sufficient.
William H. Barton‘ suggests the addition of a small piece of alum-
inum foil dropped into the solution after the hydrochloric acid is
added, to prevent the dissolving of the lead and the consequent
breaking up of the sponge by the hydrochloric acid after the zinc is
all dissolved.
T. P. Holt * recommends the substitution of a square of aluminum
foil for the zinc. The lead sponge is removed from the aluminum
with a rubber-tipped stirring rod. Care must be taken to use a
sufficiently thick sheet of aluminum (1/16’’ does well), to prevent
small pieces becoming detached. These would remain with the lead
sponge and might cause the cupels to spit.
Precipitation as Sulphide.’ Acidify five or ten assay tons
1! Western Chemist and Metallurgist, 4 p. 67, Feb. 1908.
* Mining & Scientific Press, 100 p. 863, June 1910.
3 Henry Watson, E. & M. J. 66 p. 753, Dec. 1898.
— ee re ee a ee ee Se
137
of solution with HCl and heat to boiling. While boiling add a solu-
tion containing two grams of lead acetate and pass in a current of
hydrogen sulphide until all the lead is precipitated. Allow to cool
somewhat still passing in H2S, then filter and dry. Collect the gold
and silver with lead either by a crucible fusion or scorification assay.
The method is said to be quick, accurate and economical.
Precipitation by Cement Copper.’ To 8 assay tons of the solu-
tion add a few cubic centimeters of sulphuric acid, and one gram of
finely divided cement copper. Hear to boiling and boil 10 minutes.
Filter through a strong 7 inch paper and place on the drained filter
one-third of a crucible charge of mixed flux. Place the filter in a
crucible containing another third of a charge of flux, and cover with
the final third. Fuse and cupel as usual. The filter itself furnishes
the reducing agent for the assay. If cement copper is not available,
a solution of copper sulphate may be added together with a small
piece of aluminum foil. Boil until all the copper is precipitated and
add the remaining aluminum foil to the fusion. This modification
takes more time than the first.
Precipitation by Silver Nitrate (For Gold in Cyanide Solu-
tions). Add an excess of silver nitrate solution which will cause the
gold and silver to precipitate as an auric-argentic-cyanide. Allow
the precipitate to settle, filter through a thin paper, and wash several
times. Dry the filter and either scorify with test lead or fuse in a
crucible with litharge and the regular fluxes. The method gives fairly
good results in solutions not too low in gold. With solutions very
low in gold the precipitation of the gold is not perfect.
Precipitation by a Copper Salt’ (For Cyanide Solutions Only).
Add to one liter of solution in a two liter flask 25 c.c. of a 10 per cent
solution of copper sulphate, then add 5 to 7 ¢c.c. of concentrated hydro-
chloric acid and lastly 10 to 20 ¢c.c. of a 10 per cent solution of sodium
sulphite. Shake vigorously for at least two minutes then filter,
dry, and fuse the filter and precipitate in the usual way. With weak
solutions it is best to bring up the strength by the addition of cyanide
_ before adding the copper salt. The gold and silver are carried down
by the precipitate of cuprous cyanide formed. Assays may be
completed in three hours and the results are said to be good on both
low and high grade solutions.
Albert Arents, T. A. I. M. E. 34 p. 184.
1
2 Andrew F. Cross, Jl. Chem. Met. and Min. Soe. of So. Af. 1 p. 28, and 3 p. lL.
3 A. Whitby, Proc. Chem. Met. and Min. Soe. of So. Af. 3 p. 6.
138
The Electrolytic Assay of Cyanide Solutions. The following
method is abstracted from the Journal of the Chemical, Metallurgical
and Mining Society of South Africa ' which describes the method and
installation used at the Kleinfontein Group Central Administration
Assay Offices.
Ten-assay-ton samples of the solution to be assayed are placed in
No. 3 beakers, which are held in‘a frame, and electrolyzed using a
current of 0.1 ampere. The anodés used consist of ordinary 5/16
inch arc lamp carbons which are held in position in the center of each
beaker by suitable clamps. They are arranged so that they may be
lifted out of the solution when no current is passing. The cathodes
are made from strips of ordinary assay lead foil 23’’ x 9’’ with the
lower edge coarsely serrated to allow for circulation of the solution.
To connect with the battery a }’’
of the foil, and turned upward to make a terminal. The two ends of
the lead are brought together and connected by folding the edges,
making a cylinder about 3 inches in diameter.
The time required for the complete deposition of the gold is four
hours, after which the carbons are removed, the lead cathodes dis-
connected and dried on a hot plate. When dry, they are folded into
a compact mass and cupelled.
With weak solutions a small quantity of cyanide should be added
in order to decrease the resistance and thus accelerate the deposition
of the precious metals. The author reports no difficulty in obtaining
a complete and adherent deposit of the gold, which separates as a
bright yellow deposit.
This of course was the only metal worked for on the Rand, but
there seems to be no reason why silver as well as gold can not be de-
termined by this method.
The principal advantage of the method lies in the small amount of
actual personal attention required. The method works as well
for a 20 A. T. sample as for one of 10 A. T. The time required for
the deposition of the gold is somewhat longer than for some of the
precipitation methods and this appears to be the principal disad-
vantage of the process.
Colorimetric Methods. (For Gold only.) Several attempts
have been made to adapt the ‘“‘Purple of Cassius”’ test to the estimation
of gold in chlorine and cyanide solutions. So far as the author is
aware, none of the methods have been adopted as practical assay
laboratory methods in this country. They were used for a time in
1 Vol. 12 p. 90. C. Crichton.
strip is all but severed from one end |
a ee
Pe ae ee, a ey
=< =
a
Peni
er aa ae ee hee vi)
aaa 139
one for two South African plants, but have never come into great
favo r. The two most promising methods were described by Henry
R. Cassel (E. & M. J. 76 p. 661) and James Moir (Proc. Chem.
. and Min. Soc. of So. Af. 4 p. 298) and to those original articles
interested reader is referred.
an. ha ay pa A >
¥ sa Ne A ba > ‘
we 7
CHAPTER XII.
THE LEAD ASSAY.
The fire assay for lead consists of a reducing fusion with iron,
fluxes, and some carbonaceous reducing agent, and is conducted much
as is the iron nail assay for gold and silver ores except of course no lith-
arge or other lead bearing fluxis added. The object of the fusion is to
reduce and collect all of the lead in a button free from other 2lements.
Lead Ores. Lead ores are classified by metallurgists as oxidized
or sulphide ores, also as pure or impure ores. The oxidized ores
contain the lead principally in the form of carbonate, occasionally
as sulphate and rarely as oxide or in combination with phosphorous,
molybdenum, vanadium, chromium, etc. The corresponding lead
minerals are cerussite, PbCO; (77.6% Pb), anglesite PbSO, (68.3%
Pb), minium Pb3O, (90.6% Pb), pyromorphite PbsCl (POx)3 (75.6%
Pb), vanadinite 3Pb3(VOz)2 PbCl, (72.4% Pb) and wulfenite PhMoO,
(56.5% Pb). The most important sulphide lead minerals are galena
PbS (86.6% Pb) jamesonite PbeSb.S, (50.8% Pb) and bournonite
PbCuSbS; (42.5% Pb). The principal associated minerals are
argentite, pyrite, chalcopyrite, sphalerite, stibnite, quartz, calcite
and dolomite, as well as the oxidation compounds of the above sul-
phides. Impure ores, from the assayers point of view are those con-
taining more or less arsenic, antimony, bismuth, copper, zine, and
other rarer metals which interfere with the lead assay.
Besides ores, the assayer may have brought to him various furnace.
products such as litharge, slag, matte, flue dust and cupel bottoms.
The fire assay for lead is not always as accurate as a carefully made
wet determination but it is so simple, inexpensive and rapid that for
a long time it served to govern the purchase and sale of all lead ores.
Today it is still largely used by the smelters and others for the assay
of pure ores, although for ores containing such base metal impuritie:
as antimony, copper, zinc, etc. the wet method is usually preferred.
The results of the fire assay may be either lower or higher than the
actual lead content, depending on the nature and quantity of the
other minerals present in the ore. Pure ores give low results owing
to losses of lead by volatilization and slagging. Both the sulphide
and oxide of lead are volatile at moderate temperatures and for this
reason great care must be taken to keep the temperature as low as
141
possible consistent with a proper decomposition of the ore, and of the
lead compounds which are formed during the fusion. On the other
hand lead compounds, particularly the oxide, tend to pass into the
slag which tendency is increased by the presence of zinc, and to some
extent by arsenic and antimony. Impure ores containing arsenic,
antimony, bismuth and copper, usually give high results as these
metals are partly or wholly reduced and pass into the lead button.
Quantity of Ore and Reagents Used. The amount of ore
used is generally 10 grams, occasionally 5 grams. With low grade
ores 20, 25, or more grams may be used. ‘The reagents used are the
alakli carbonates, borax-glass, some reducing -agent usually argols
or flour, and occasionally sulphur. Iron in some form is always used.
It may be in the form of nails or spikes, coiled wire, or the crucible
itself may be of iron, and in this case will be used over and over again
until worn out. A very satisfactory way of introducing iron is to
use a rail or boat spike 25 or 3 inches long, and about 0.5 inches
through. In this assay it is customary to use a mixture of sodium
and potassium carbonates as the mixture fuses at a lower temperature
than either one alone. The alkali carbonates act as fluxes for the
silica, and serve to give a basic slag which is necessary in this assay.
Usually from two to three times as much alkaline carbonate as ore
is taken. Borax-glass acts as a flux for the metallic oxides for lime-
stone and the other alkaline earths. From one-half to twice as much
borax-glass as ore is used. An excess of reducing agent 1s always used
to insure keeping the necessary strongly reducing condition. Sul-
phur is only occasionally used and then when assaying an oxidized
ore containing copper.
In the lead assay it is customary to use a mixed flux called a “‘lead
flux.’ This may be bought already prepared or may be made up in
the laboratory. Many different formulas are given among which are
the following:
L 2 o.
Sodium-bicarbonate 12 parts 4 parts 6.5 parts.
Potassium carbonate 1Dtae ria Aine
Borax-glass ny ime — 26D ies
Borax powdered = Chae _
Flour Zsa Liven A te
No. 1 and 2 are found in use in the Coeur d’ Alene lead district
where the fire assay for lead has been brought to the highest degree
of perfection. No. 1 is better for ores having a basic gangue, No. 2
for siliceous ores. No. 3 is perhaps the best of all for general use.
Fa
142
About 380 grams of flux are intimately mixed with 10 grams of ore,
a nail is inserted and a cover of 8 or 10 grams more of flux is added.
Very few assayers use a cover of salt in the lead assay on account of
the danger of the loss of lead as chloride.
The fusion should always be made in a muffle furnace owing to the
better control of temperature available. In fact the secret of the
successful fire assay for lead is largely in the proper manipulation
and control of the temperature throughout the process:
At the start the temperature should be low, sufficient only to barely _
melt the charge. This is necessary owing to the fact that in the early
part of the assay the charge is in active motion and particles of the
various lead compounds are continually being brought to the surface,
where if the temperature were high they would suffer an appreciable
loss by volatilization. When the charge has finished boiling and most
of the lead is reduced and collected in the bottom of the crucible
there is less danger of a loss by volatilization, owing first to the fact
that lead itself is not so readily volatile as are some of its compounds
and second to the difficulty of migration of the molecules through the
heavy layer of reducing slag which covers the lead.
After the boiling has entirely ceased the temperature is raised gradu-
ally to decompose the lead compounds which still remain in the slag.
These are principally the silicate and the double sulphide of lead and
sodium or potassium and require a bright yellow heat for their com-
plete decomposition. The fusion period is finished when the nails
can be removed free from shots of lead. Sulphide ores require a much
longer fusion than oxides owing to the fact that their decomposition
is effected principally by iron, and therefore time must be allowed for
every particle of the charge to come into contact with the iron.
Oxide ores, on the other hand, are decomposed by the carbon of the
charge and as this is uniformly distributed a much shorter time will
suffice. Sulphide ores will require from one to one and one half
hours of fusion, oxide ores from three quarters of an hour to an hour.
Influence of Other Metals on Lead Assay. Silver. Practically
all of the silver in an ore is reduced and passes into the lead button.
If present in sufficiently large quantities a correction for it may be
made, i.e., 291.66 oz. per ton equals one per cent.
Gold. This metal is also reduced and passes into the lead button,
but it is usually present in such small quantities that it may be dis-
regarded.
Arsenic. Arsenic is occasionally found in lead ores usually in the
form of arsenical iron pyrite. During the assay part of the arsenic
is volatilized as metal or as arsenic sulphide but the larger part re-
*
eee ee _
143
mains in the crucible. Here it usually enters into combination with
the iron forming a speiss. After pouring it will be found as a hard
white button on top of the lead from which it may be removed by
hammering. Little if any arsenic enters the lead button. Under
certain conditions, 1.e., a long fusion at a low temperature with high
soda excess, the formation of a speiss may be prevented.
Antimony. This metal is frequently found associated with lead,
usually however only in small amounts. In the assay with iron,
antimony is reduced and passes into the lead button. Buttons con-
taining antimony are harder and whiter than those from pure lead
ores and when they contain much antimony are brittle, breaking with
a bright crystalline fracture.
If much antimony is present (over one-half as much as the lead)
an antimony speiss will be found lying on top of the button.
Bismuth. This metal is rarely found associated with lead ores, but
if present will be reduced and pass into the lead buttons.
Copper. Copper is often found in lead ores in the form of chalco-
pyrite, chalcocite, and oxidized copper compounds. If the ore is
fully oxidized and a high temperature is employed most of the copper
will pass into the lead button. If the ore contains much pyrite or
sulphur in other forms most of the copper will remain as a sulphide
and be dissolved in the alkaline slag. A button containing copper
will be hard and tough and may show a reddish tinge.
Iron. This metal is often present in lead ores usually in the form
of iron pyrite. It goes into the slag forming either a silicate or a
double sulphide of iron with sodium or potassium. The lead button
is practically free from iron.
Zinc. Zine is often found associated with lead in ores usually
in the form of the sulphide. During the assay part of the zinc is
volatilized and part remains in the slag. Zine sulphide is only de-
composed by iron at a very high temperature so that only a very
small amount of zinc passes into the lead button. Zine sulphide is
practically infusible, so that if present in too great an amount, may
make the slag thick and pasty, and thus interfere with the separation
of the lead.
Procedure. Assay ores in duplicate using 10 grams of ore and 40
grams of prepared lead flux. Use a 12 or 15 gram muffle crucible.
Weigh out first 30 grams of lead flux, place the ore on top of this and
mix thoroughly with the spatula. Insert a spike or nails point down-
ward and finally cover with 10 grams more of lead flux. Have the
144
muffle only moderately red so that it will take at least 30 minutes
from the time the charges are put in until they are boiled down.
Close the door to the muffle as soon as the crucibles are in and after
the charges are melted place two crucibles part full of soft coal in —
the mouth of the muffle just inside of the door, which should be kept
as tightly closed as possible. Raise the temperature gradually to a
bright yellow and continue at this temperature until the nails can be
removed free from lead. |
Finally take the crucibles from the muffle using a pair of muffle
crucible tongs and without setting them down quickly remove the
nails with a large pair of steel forceps, tapping against the side of the
crucible and washing the nails in the slag to remove all adhering lead
globules. Pour the fusion into a mold and when cool separate the
lead from the slag and hammer clean. Weigh to centigrams and
report the results in percentage. Duplicates should be checked
within 0.5 per cent.
The slag should be black and glassy. If dull, more borax-glass
should be added. It should pour well from the crucible and immedi- ~
ately after pouring, the crucible should be examined for shots of lead.
If these are found it is usually an indication of too low a temperature
at pouring.
Notes. 1. If the ore is an oxide and contains copper add a gram or two of finely
pulverized sulphur to the charge to prevent the copper from entering the button.
2. The soft coal is added to insure reducing conditions in the muffle and it may
be renewed if necessary. When a muffle is used solely for fusion purposes the hole
in the back is stopped up, thus preventing the entrance of so much air.
3. The removal of nails and pouring must be done without a moments delay as
the charges are small and cool rapidly.
4. If the ore contains much silver the button should be cupelled and the weight of
silver found deducted.
5. The lead should be soft and malleable and a fresh cut surface should have the
bluish gray color of pure lead. The button should be capable of being hammered
out into a thin sheet without breaking or cracking. A button that is bright, brittle
and brilliantly white in the fracture indicates the presence of arsenic or antimony.
6. If there is doubt regarding the purity of the lead button it may be tested by
cupellation. The only metals except lead likely to be present are gold, silver, anti-
ee copper and possibly zinc; each of which gives characteristic indications in
cupelling.
7. Crucibles may be used a number of times as they are but little corroded but
those used previously for gold and silver assays must nol be used for this assay as the
slag left in them contains lead. It is well to use a special size of crucible for the lead
assay in order to prevent errors from mixing crucibles.
Assay of Slags, Furnace Products and Low Grade Ores or
Tailings. In the assay of low grade materials such as slags and
tailings a larger quantity of ore should be used and a different mixture
of fluxes. The slag should be between a singulo and a sub-silicate
and part of the iron may be added in the form of filings. On account
of the size of the charge it is well to add a number of nails, as this
will lessen the time necessary for complete reduction.
The following charges have been found satisfactory :
145
Limestone (3-2% Pb) Slag Slag
Ore 25 gm. Slag 25 gm. Slag 100 gm.
NasCO; 25 ay NaeCO3 20 y NaeCO; 50
K.CO; 20 os K.COs; 20 a K.CO; mr
Borax-glass 20 ‘“ Borax-glass 10 ‘“ Borax-glass 10 “
Flour Pose Flour Mg Se Nelielnte LOP os
Nails ree Nails Det pee Nets 5 els
(20 penny) (20 penny) (20 penny)
20 gram crucible
20 gram crucible -
30 gram crucib!e
Allow some time at a high temperature to allow opportunity for
all of the slag to come in contact with the iron.
Corrected Lead Assay. To recover any lead which may have
been left in the slag the following procedure is recommended. Save
all the slag and remelt in the original crucible with the spikes or nails
formerly used. If the first slag was quite glassy and viscous in
pouring, add from 5 to 15 grams more sodium carbonate. Heat to
redness and drop into each crucible a lump of about 5 grams of
potassium cyanide. Close the door to the muffle and heat to a bright
yellow and pour as soon as quiet. Add the weight of any small
- button found to the lead from the original fusion.
Chemical Reactions of the Lead Assay.
With an ore containing PbCOs;, PbSO., PbS, SiO. and CaCO; the
following reactions may occur :—
PbCO; = PbO + CO,
2PbO=- C = 2Pb + CO,
PbO + S810, = Pb SiO; *(Begins at 625° C.)
PbSO, + 2C = PbS + 2CO, (Begins at a dark red heat.)
7 PbS + 4K.CO; = 4Pb + 3 (K2PbS,) + K,SO, + 4CO2
(Begins at a red heat.)
If carbon were not present some oxide and sulphate would probably
-remain to react as follows :—
PbS + 2PbO = 3Pb + SO, (Begins at 720° C.)
PbS + PbSO, = 2 Pb + 250, (Begins at 670° C.)
2PbSO, + SiO2 = Phe SiO, + 2802 + O2 (High heat.)
Toward the end as the heat is raised to a bright red and above, the
reactions with iron become of importance, particularly the following :—
PbS + Fe = Pb + FeS
PbSi0; + Fe = Pb + FeSiO;
(Begins at 200° C.)
, (Begins at 550° C.)
(Requires a bright yellow: heat for
completion. )
(Requires a bright eo heat for
completion. )
K»PbS + Fe = Pb + KeFeS:
INDEX
Acid slags, 88.
Annealing parted gold, 69, 70.
Antimony, effect in iron nail assay, 108.
effect in lead assay, 143.
effect in scorification, 77, 78.
Antimonial ores, crucible assay of, 116.
Argols, 4, 5.
Arsenic, effect in iron nail assay, 108.
effect in lead assay, 142.
effect in scorification, 77, 78.
Assay—ton weights, 48.
Balance, alignment of knife edges, test-
ing, 46.
arms, equality, testing, 46.
button, 38.
construction of, 38.
directions for use of, 40, 41.
equilibrium, testing, 44.
flux, 37.
multiple rider attachment, 47.
pulp, 37.
sensibility of, 45, 46.
stability of, 45.
testing an assay balance, 44-46.
theory of, 38-40.
time of oscillation, 45.
Basic ores, assay of, 92.
calculation of charge for, 92.
slags for, 92, 93.
Basic slags, 88.
Bone-ash, 51, 52.
best size for cupels, 52.
temperature of burning, influence of,
51.
Bone-ash cupel, assay of, 118.
Borax, 2.
effect of in slags, 86-88, 98.
use of in assaying, 86-88, 98, 102.
Borax-glass, 2.
Bismuth, effect in lead assay, 143.
Bullion, copper, assay of, 123-128.
doré, 128-130.
gold, 130-133.
lead, 123.
nomenclature, 119.
sampling of, 119-123.
silver, 128-130.
Capsules, parting, 69.
Character of sample, determination of,
85.
Charcoal, 5.
Chiddey method for assay of cyanide
solution, 135, 136.
Class 1 ores, assay of 89-101.
Class 2 ores, assay of, 97-110.
Class 3 ores, assay of, 110.
Classification of ores, 84.
Classification of silicates, 86.
“Cleaning the slag” in scorification, 82.
Coal furnaces, 12.
firing of, 12.
Cobalt, effect in scorification, 78.
Coke furnaces, 13.
Colorimetric method of assay, 138.
Combination assay of copper bullion,
123, 124, 126, 127.
Copper bullion, assay of, 123-128.
sampling, 121-123.
Copper, crucible method for ores high
in, 114, 116.
effect of in cupellation, 62, 63.
effect of in iron nail assay, 108.
effect in lead assay, 143.
effect in scorification, 77, 78.
matte, assay of, 80.
Corrected assays, gold and silver, 117,
118.
lead, 145.
Cover, the, 98.
Cryolite, 7.
148
Crucible assay, 83.
copper bullion, 123, 125, 126.
procedure for, 100, 101, 103-107.
theory of, 83, 108-111.
Crucible furnaces, 9.
Crucible slags, ‘properties of, 85.
Crucibles, 18, 19.
capacity of different sizes, 19.
desirable properties of, 18.
size of for various charges, i11.
Cupels, 51, 52.
assay of, 118, 121.
cracking of, 52, 53.
effect of shape of, 54.
instructions for making, 52, 53.
machines for making, 53.
magnesia, 66.
Portland cement, 66.
size of, 53, 54.
testing, 53, 65, 66.
Cupellation, 51, 54-57.
appearance of buttons containing
platinum, 58, 65.
correct temperature for, 55.
‘freezing’ of button in, 56, 58.
indications of metals present, 64.
instructions for, 57-59.
loss of gold in, 61.
loss of silver in, 59-61.
regulation of temperature during, 57,
58.
retention of base metals in beads
from, 66.
“spitting” during, 53.
“sprouting”’ of silver after, 56.
Cupellation losses, 59-64.
influence of copper on, 62, 63.
influence of impurities, 62.
influence of quantity of lead, 60.
influence of tellurium, 113.
influence of temperature, 59-61.
Cyanide solutions, assay of, 133-139.
Doré bullion, assay of, 128-130.
Electrolytic assay of cyanide solutions,
138.
Fire brick, directions for laying, 16.
Fire brick lining vs. tiles for lining, 11.
Flasks, parting, 71.
Flour, 5.
Fluidity of slags, 88.
Fluorspar, 7.
Fluxes and reagents, 1-7.
Fluxing, principles of, 84.
Fuel, advantages of gas and oil over
solid, LW
for assay furnaces, 10, 11.
Furnace repairs, 15, 16.'
Furnaces, 9-15.
directions for firing, 12.
Fusion products, 7, 8.
Gasolene furnaces, 13, 14.
Gold, ores containing coarse particles,
assay of, 35, 36.
Gold bullion, assay of, 130-133.
Gold, weighing, 69, 70.
Granulated lead, assay of, 79.
Heat of formation of metallic oxides,
74, Pia:
Ignition temperature of metallic sul-
phides, 74.
Inquartation, 70, 71.
Iron, 5.
effect in lead assay, 143.
effect in scorification, 77, 78.
reducing action of, 5.
Iron assay, 103, 107-109.
Jones sampler, 26.
Lead, 4.
fire assay for, 140-145.
granulated, assay of, 79.
granulated to make, 4.
ores, 140.
Lead bullion, assays of, 58, 59, 123.
sampling, 121.
Litharge, 3, 4.
assay of, 99, 100.
corrosive action of, 75.
Lodge’s rule for, 106.
solubility of metallic oxides in, 73,
74.
use in the scorification assay, 75.
Lodge’s rule for the use of litharge, 106.
Magnesia cupel, assay of, 118.
Manganese, effect in scorification, 78.
Matte, 8.
copper, assay of, 80.
Mercury-sulphuric acid combination
method for copper bullion, 123,
124, 126, 127.
Metallic assay, 35, 36.
Metallic oxides, heat of formation of.
74, 113.
solubility in litharge, 73, 74.
Metallic sulphides, ignition tempera-
ture, of 74.
Minerals, oxidizing power of, 96.
reducing power of, 96.
Moisture sample, 22.
Muffles, 17.
care of, 67.
directions for replacing, 16.
method of supporting, 12.
spilling in, 67.
Muffle furnaces, 9, 11, 12.
Multiple rider attachment for balances,
47.
Neutral ores, 84.
Nickel, difficulty of eliminating in
scorification, 74.
effect of in iron nail assay, 108.
effect in scorification, 74, 79.
Niter, 5, 6.
oxidizing power of, 105.
oxidizing reactions, 5, 6.
Niter assay, 102-107.
objections to, 107.
preliminary fusion, 105-107.
regular fusion, 105-107.
Nitric acid combination method for
copper bullion, 124, 127.
Ores, classification of, 84.
determining reducing power of, 103,
104.
estimating reducing power of, 104,
105.
Oxides metallic, heat of formation of,
7a) 118. |
solubility in litharge, 73, 74.
149
Oxidizing power, definition, 94.
of minerals, 96.
of niter, 97, 105.
of niter, determining, 105.
power of ore, to find, 111.
ores having, 84.
of red lead, 97.
Oxidizing reactions, 96, 97.
Parting, 68.
acid for, 68, 69, 133.
capsules, 68.
flasks, 71.
preparing beads for, 71.
procedure, 69-72, 133.
ratio of silver to gold necessary, 68,
70.
Portland cement cupel, assay of, 118.
Potassium carbonate, 3. :
Potassium cyanide, 6.
Reactions, during iron assay, 107, 108.
lead assay, 145.
oxidizing, 96, 97.
reducing, 94, 95.
Reagents, 1-7.
testing, 98, 99:
Reducing agents, 4, 5.
Reducing power, definition, 94.
of minerals, 97.
of ores, determination of, 103, 104.
of ores, estimation of, 104, 105.
ores having, 84.
of reagents, 94.
of reagents, determination of, 99.
Reducing reactions, 94—96.
Rescorifying buttons, 78, 80, 124.
Riders, 47, 50.
Riffle cutter, 26.
Roasting method, 103, 109, 110.
Salt, action of, 98.
objections to, 98.
Sample, definitions, 21.
finishing, 34, 35.
moisture, 34, 35.
Samples, labelling, 22.
Sampling, copper bullion, 121-123.
duplicate, 34.
150
Sampling, gold bullion, 130.
hand and machine compared, 27.
lead bullion, 121.
Sampling, machine, 26, 27.
operations, 22~27.
ores containing malleable minerals,
35, 36.
Brunton’s formula for, 32.
Reed’s formula for, 31.
Richard’s rule for, 29,
tables showing weights to be taken,
30-33.
theory of, 28-32,
Scorification assay, 74.
charges for different ores, 81.
chemical reactions in, 77,
of copper bullion, 123-125,
of copper matte, 80.
effect of various metals, 78.
for gold, 79.
indications of metals present. 78,
losses in, 81.
of matte, procedure for, 80.
ores suited, 80.
procedure, 75-77, 80, 124.
Scorifiers, 19, 20.
sizes, 74,
spitting of, 79, 81.
Segregation of metals in cooling, samp!-
ing effeeted by, 120, 121.
miliea 142.
Silicates, classification of, 86.
mixed, 89.
Siliceous ores, assay of, 89-92.
calculation of charge for, 89-92.
Silver, effect in lead assay, 142.
Slags, 85, 86.
acid and basic distinguished, 88, 89.
action of borax in, 86-88.
for Class 1 basic ores, 92, 93.
for Class 1 siliceous ores, 89-92.
Slags, for Class 2 ores, 97, 98, 105, 106.
fluidity of, 88.
formation temperature, 88.
Slag factors, bi-silicate, 91, 93.
sub-silicate, 106.
Slag forming constituents of ores, 84.
Sodium carbonate, 2, 3.
Solubility of metallic oxides in litharge,
73, 74.
Solutions, assay of, 133-139.
Speiss, 8.
Stack, height of, 12.
size of, 12.
support of, 12.
Stanniferous: ores, crucible assay of,
Ligeia
Sulphuric acid combination method,
for copper bullion, 123, 124, 126,
127,
Telluride ores, assay of, 112-114.
Tellurium, effect of in cupellation, 113.
in fusion, 113, 114.
Thompson rider, 47.
Tin ores, assay of, 116, 117.
Vanning, 85.
shovel, 85.
Weighing, 41-43.
by equal swings, 42.
by method of swing, 42, 43.
by ‘‘no deflection,” 43, 44.
by substitution, 44.
checking of, 44.
gold, 69, 70.
out ore, 76.
Weights, 47, 48.
calibration of, 48, 49.
Zine effect in lead assay, 148.
ii 00 N10
Eta
te vligh wae e
x Ae
fang mia ase
‘~! .
taint ale tg sty
re
‘st
4
I
we pine
site oe eT
Rtas
Serratia 7
vias
Ce eee
ee bee's
tah
ae
ey