Chapter IX (4)

If, however, such a button is yellow, then, from its weight and depth of colour, a rough estimate can be made of how much gold is contained in it. Silver must be added to make the total weight 3-1/2 times as much as that of the gold supposed to be present. Thus, if the button weighs 10 milligrams and is supposed to contain 8 milligrams of gold, then 8 multiplied by 3-1/2 is 28; the button must, in such case, be made up to 28 milligrams by adding 18 milligrams of silver. In judging of the quality of the gold button, no ordinary error will very seriously affect the result. If, in the example just given, the quantity of gold present was really 7 or even 9 milligrams of gold, the resulting alloy would still have been suitable for such partings. In fact, in routine assays, where the quantity as well as the quality of the gold is known within fair limits, it is often the custom to add the silver for inquartation to the lead during the first cupellation.

But in the assay of rich gold alloys such approximate work will not do. If the composition is not already known with a fair degree of accuracy _preliminary assays_ must be made. Weigh up two lots of 100 milligrams of the alloy and wrap each in 3 grams of lead. To one add 300 milligrams of silver. Cupel both. The button containing the added silver must be flattened and boiled with 15 c.c. of nitric acid; and the resulting gold must be washed, dried, ignited and weighed. This, in milligrams, gives directly the percentage of gold. The weight of the other button gives the percentage of gold and silver; the difference between the two gives the percentage of silver. The rest will, perhaps, be copper.

The composition of the alloy being known, or having been determined as just described, the calculation of how much silver must be added is fairly simple. The following is an example. Suppose the bullion contains 92 per cent. of gold, 1 per cent. of silver and 7 per cent. of copper, and that .5 gram of it is to be taken for an assay. The .5 gram, then, will contain

Gold .460 gram
Silver .005 "
Copper .035 "

But the total silver required is .46 gram × 2.5. This equals 1.15. Allowing for the .005 gram of silver already present, 1.145 gram of silver must be added.

The silver is incorporated with the gold, and at the same time the copper is eliminated, by cupelling with sheet lead. How much sheet lead must be used will depend partly on how much bullion is taken, partly on how much copper it contains. Four grams of lead will do for a .5 gram charge; and for a .3 gram charge, 3 grams may be used. But with 20 per cent. of copper these amounts should be doubled; with 40 per cent. of copper they should be trebled; and with over 60 per cent. of copper four times as much lead should be used. For small buttons of gold as little lead as may be relied on to start cupelling may be taken; the lead may conveniently be in the form of little cups made by folding lead foil on a piece of glass rod. With a large number of bullion assays systematically worked and checked a simple plan would be to always use the quantity of lead required by the alloy containing most copper which turns up for assay. This weight, cut out of lead foil, would be kept in stock folded into little bags ready to receive the bullion and silver.

The silver used for inquartation must, of course, be free from gold and is best prepared by the assayer who is to use it (see p. 66). It should not be in long strips or angular pieces likely to perforate the lead in which it is folded. When wrapped in the lead it should be in the middle and should make as compact a parcel as possible.

Each little parcel, as completed, should be placed on a tray in its properly numbered compartment. Its position here should correspond to that it will occupy in the muffle and eventually in the cupel tray. The cupellation must be made with all the requisite precautions. A good smooth malleable button is needed for the next operation, which is known as flatting.

~Flatting.~--Small buttons, such as are got in assaying most gold ores, are placed on a polished steel anvil and flattened by one or two blows with a hammer. The flattened discs are heated to dull redness on a clean cupel and are then ready for parting. Somewhat larger buttons may be similarly treated, but they should be annealed (_i.e._ heated to redness and allowed to cool) during the flattening. The silver-gold alloy left from the cupellation is soft and bends like lead; but after hammering or rolling it becomes harder, gets a spring in it like a piece of mainspring and cracks or splits somewhat easily. There should be no cracks or stripping or even roughness on the flattened metal, since such defects may cause the loss of small particles either during the flattening or in the subsequent treatment with acid. The softness of the metal is restored by heating. In bullion assays the flatting of the buttons requires care and practice for its skilful working. The strips of alloy for parting should be of uniform thickness and condition so that the action of the acid shall be equal in all cases. The button is taken from the cupel, cleaned and placed on the anvil: it is then struck a heavy blow which widens it to about 3/4 inch in diameter; this blow is followed by two others, one a little in front, the other behind, which lengthen the disc and give a very blunt roof-like slope to its upper face. It should then be annealed. This may be done by putting it in a just red-hot scorifier heated in a muffle: it very soon attains the right heat and may then be transferred to a cold scorifier; the hot scorifier should be put back into the muffle. The softened disc is then taken to the rolls (Fig. 45). The rolls are loosened until the disc can be pressed between them. Looking through the interval between them the rolls should appear exactly parallel; if they are not, one adjusting screw should be loosened and the other tightened until parallelism is obtained. The rolls are now turned and the disc should be drawn through without any great effort. Beginners are apt to err by trying to do too much with one turn of the handle. It is easy to stop whilst the rolls are only just gripping the metal and then to bring the disc back by reversing the action. If the disc was originally level and the rolls are parallel, the metal will appear as a strip which has been merely lengthened. If the rolls are tighter on one side the strip will be bowed; the tighter side will correspond with the outer curve of the crescent. A mistake of this kind may be amended by passing the strip through the rolls the other way, so as to reverse the irregularity and so straighten the strip. The screw on the looser side should then be tightened until parallelism is obtained; after which more care should be taken to tighten the two screws equally. The rolling should be stopped when the strip is 3 or 4 inches long and of the thickness of an ordinary visiting card. The strip should be annealed during the rolling and again at the finish.

~Parting.~--The thin sheet of metal is dropped into hot dilute nitric acid and boiled for five or six minutes after the brisk action of the acid on the metal has ceased. At this stage nearly all the silver has gone into solution as nitrate of silver and the acid is charged with this salt. This acid is poured off and the residual metal is again boiled for from 20 to 30 minutes with a second lot of stronger acid. This leaves the gold almost pure, though it may still retain from .05 to .1 per cent. of silver. Treatment with the first acid only would probably leave three or four times as much.

The _nitric acid_ used should be free from hydrochloric, sulphuric, iodic and telluric acids. In testing it for the first of these add nitrate of silver and dilute with distilled water; there should be no turbidity. In testing for the others evaporate three lots in dishes over a water-bath. Test one for sulphates by adding water and barium chloride. Test another for iodates by taking up with a little water, adding a few drops of starch paste and then dilute sulphurous acid solution a little at a time; there should be no blue colour. Test the third for tellurium by heating with 1 c.c. of strong sulphuric acid until dense fumes come off; allow to cool considerably; a piece of tin foil added to the warm acid develops a fine purple colour if only a trace of tellurium is present.

The presence of lower oxides of nitrogen, which impart a brown colour to the acid, is objectionable; they, however, are removed by boiling the diluted acid before using it for parting. It is usual to keep a stock of the acid suitably diluted to the two strengths required for the parting. These are known as the parting acids. The _first parting acid_ is the weaker and is used in the first attack on the metal. The specific gravity generally recommended for it is about 1.2. It may be prepared either by diluting the strong acid with about its own volume of distilled water, or by suitably diluting the second parting acid which has been already used in an assay; the small proportion of silver this contains is not harmful for this purpose. The _second parting acid_ has a specific gravity of about 1.3, and may be made by diluting the strong acid with half its volume of distilled water.

_Parting in Flasks._--Flasks are most convenient for the larger partings, as in bullion assays; and should always be used for this purpose unless some of the special parting apparatus, like that used in Mints, is available. Many assayers use flasks, though of a smaller size, for the ordinary partings in assaying gold ores. The flasks are either bulbs with long necks (Fig. 46) which ought to be heated on rose burners of special construction; or they are small flat-bottomed conical flasks which may be conveniently heated on a hot-plate and are, in this respect, much easier to deal with in general work. The following instructions apply to the parting of an alloy containing a few decigrams of gold together with the proper proportion of silver.

The strip from the rolls, after being softened by annealing, is folded on itself on a glass rod into a roll or cornet. It should be so plastic that it will retain the shape thus given it and not spring open on removing the pressure of the fingers. About 50 c.c. of the first parting acid are placed in a 6-ounce conical flask and heated to boiling; the flask is then withdrawn, and tilted a little to one side, whilst the cornet is cautiously dropped into it; there will be a sudden issue of hot vapours and a prompt withdrawal of the hand is advisable. The flask is replaced on the hot plate and the acid is kept boiling for 10 or 15 minutes. The flask is then withdrawn and the acid diluted with about an equal volume of distilled water. If the flask has a thick glass band around its neck, a little way down,[28] care must be taken to use hot water, for any sudden chill will certainly crack the flask where it is thus thickened. The liquor is carefully decanted into a clean beaker and is then thrown into a jar marked "waste silver." About 40 c.c. of the second parting acid, heated to boiling, is then poured into the flask, which is then replaced on the hot plate. The boiling is continued for 15 or 20 minutes or even longer. At this stage bumping has to be specially guarded against; after a little experience it is easy to see when this is imminent and the flask should be withdrawn to a cooler part of the plate; it is better to prolong the heating at a temperature below boiling than to run the risk of disaster. Some of the older writers, however, are rather insistent on vigorous boiling with large bubbles. The addition of a small ball of well-burnt clay of about the size of a pea has been recommended, as it lessens the tendency to irregular and dangerous boiling. At the end of the treatment with the second acid the flask is withdrawn from the plate and the acid is diluted with an equal volume of distilled water. The liquor is carefully decanted into a beaker, and then poured into a jar or Winchester marked "acid waste"; it serves for making the first parting acid. The flask is then washed twice with hot distilled water; the washings must be carefully decanted from the gold. The flask is then filled with water. A parting cup (size B) is then placed over its mouth, like a thimble on the tip of a finger. This cup is of unglazed porous earthenware of such texture that it absorbs the last few drops of water left on drying; and with a surface to which the gold does not adhere even on ignition. The gold should fall out cleanly and completely on merely inverting the cup over the pan of the balance. The flask and cup are then inverted so that the flask stands mouth down in the cup; a little of the water from the flask flows into the cup, but only a little. The gold falls steadily through the water into the cup. When time has been allowed for even the finest of the gold to have settled into the cup, the flask is removed. This is easiest done under water. The cup, with the flask still resting in it, is dipped under water in a basin; as soon as the neck of the flask is immersed the crucible can safely be drawn away from under it and then lifted out of the water. The flask should not be taken away first, for the rush of water from it may easily sweep the gold out of the cup. The water in the cup is then drained off and the cup is dried at not too high a temperature; for if the last drop or two of water should boil there is danger of spattering the gold out of the crucible. When it is dry, the cup is heated on a pipe-clay triangle over a Bunsen burner, or on a slab of asbestos in a muffle, to a dull-red heat. This brings the gold to "colour"; that is, the loose tender dark coloured gold becomes bright yellow and coherent; and is in a state fit to be transferred to the balance and weighed. All unnecessary transferences must be avoided. As soon as the cup is cool it may be inverted over the pan of the balance, when the gold will fall out cleanly or, at the worst, a gentle tap with the finger will be sufficient to detach it.

_Parting in test-tubes_, or in the smaller conical flasks, is used in the assay of gold ores of ordinary richness. The work is exactly like that just described in all its main features. Generally speaking much less acid will be used; for example, in test-tubes and for small buttons, 3 or 4 c.c. of each acid is quite enough. Again, the action need not be so prolonged; 10 or 15 minutes in each acid is sufficient. So, too, the heating may be less; it is very convenient to support the test-tubes in a water-bath, or merely to rest them in a beaker of boiling water; and there is no serious objection to doing this. A smaller parting cup should be used; the A size is suitable. The button, on the other hand, should be beaten thinner than is needed for the larger partings. If the silver should be in excess and the gold becomes much broken up, ample time should be given for subsidence from the test-tube or flask into the parting cup.

_Parting in glazed crucibles or dishes._--This method of working has the advantage that there is no transference of the gold until it is placed on the pan of the balance. On the other hand, in the boiling more care is required in adjusting the temperature. The following instructions apply to the treatment of very small buttons, to which the method is more particularly applicable; but very little modification is needed for the treatment of larger buttons. The smallest sized Berlin crucibles answer admirably. They should be cleaned by treatment with hot and strong sulphuric acid, followed by washing in distilled water; the comfort and ease of working mainly depends on the thoroughness of this cleaning. The crucible, one-third full with the first parting acid, is heated on the hot plate until the acid is almost boiling. The flattened and annealed button is dropped into it and the heating continued with, at most, gentle boiling for a few minutes. The crucible is then filled with distilled water, which cools it enough for easy handling; and when the gold has settled the liquor is poured off along a glass rod into a clean beaker. Any greasiness of the crucible makes itself felt here and is very objectionable. The crucible is then one-third filled with the second parting acid and the heating resumed, care being taken not to raise the temperature too high; this should be continued much longer than before, say for five or ten minutes or even longer according to the size of the button. Distilled water is again added and, when it is drained off, the washing with distilled water is twice repeated. It will not be possible to drain off the last drop of water; but if the gold is coherent, the crucible can be so inclined that this drop drains away from the gold, in which case the drying can be done rapidly; the boiling of the water will do no harm. But when the gold is much broken up, it will collect in the middle of this drop and the drying must be done gently; best by putting the crucible in a warm place. When dry, the crucible is heated till the gold changes colour, but the heat must be kept well below redness. When cold, the gold is transferred directly to the pan of the balance. With minute specks of gold which will require measuring, it is best to put a small piece of lead foil (say .1 gram) in the crucible over the gold, and then heat the crucible to above redness over a blowpipe. Whilst the lead is oxidising it is easily swept round in a bath of molten litharge by merely tilting the crucible. In this way any separated specks of gold can be taken up with certainty. When the worker is satisfied that the lead has had ample opportunity for taking up the gold, the lead must be kept in one place and the heat slowly lowered. By this means the button becomes supported in comparatively pure litharge and when solid can be picked out quite easily with a pair of pliers and in a very clean condition. The lead button is then cupelled on a very fine cupel, as already described. The method of working last described destroys the crucible. If the gold is not quite so small this may be avoided. A small piece of lead foil should be hammered out until it is perfectly flexible. It is then shaped into a tray and the gold is transferred to it. The lead is then folded over, with the help of two pins; and cupelled.

If the crucible shows a black stain on heating it is because some silver remains through bad washing. It shows poor work and the assay should be repeated.

_The silver retained in the gold after parting_ is, in bullion assays, an important matter; it is roughly equal to the loss of gold due to absorption by the cupel. Mr. Lowe working on .5 oz. of gold, obtained by parting in assaying bullion, found it to contain .123 per cent. of silver. Dr. Rose in some special assay pieces found by a less direct method of assaying, from .06 to .09 per cent. of silver. The proportion of silver retained varies in a marked way with the proportion of gold to silver in the alloy before parting. It is generally stated that the retained silver is least when this proportion is 1 to 2-1/2, and more or less silver than this leads to a less pure gold after parting.

_Platinum_ in an alloy being parted is dissolved along with the silver either altogether or in part. It imparts a straw yellow colour to the parting acid. _Palladium_ gives an orange colour to the acid.

_The loss of gold by solution in the acid during parting_ is small, but easily demonstrable. On a 500-milligram charge of bullion it may amount to from .05 to .15 milligram; _i.e._ from .01 to .03 per cent. It is due to gold actually dissolved and not merely held in suspension.

~Assaying with checks. Surcharge.~--It will be seen from what has been stated that the errors in gold parting are of two kinds: viz. (1) a loss of gold on the cupel and to a less extent by solution in the acid, and (2) an apparent gain of gold due to the retention of silver in the parted material. Both errors are small, and as they are of an opposite character they tend to neutralise each other. Hence they are altogether without effect on the accuracy of the assays of ores when the total gold is reckoned in milligrams. And even with the larger amounts present in bullion assays their influence is so small that an uncorrected result is still fairly accurate; the resultant error would not be more than one part in two or three thousand.

It is customary to report the purity of bullion, or its fineness as it is called, in parts per thousand of bullion. The sum of the errors of an assay, which is called the _surcharge_, is reported in the same way. Thus a surcharge of + .3 means that the gold as weighed was .3 part per 1000 more than the gold actually present. But a surcharge - .3 means that on the whole there was a loss of .3 part per 1000 in the assay.

Speaking roughly the retained silver will vary with the weight of gold present; if one alloy contains twice as much gold as another the retained silver will be about twice as much also. On the other hand, as already explained, the cupellation loss on the poorer alloy is as much as, or even more than, with the richer one, because of the copper, &c. present. With rich gold alloys the silver more than compensates for the loss and the surcharge is positive; but with poorer alloys the loss is greater and the surcharge is negative.

In Mints and places where bullion assays must be made with the highest attainable accuracy, the surcharge is determined by experiment, and the proper correction is made in the reports on the bullion. This is done by making assays of gold of the highest degree of purity alongside of those of the bullion whose quality has to be determined. These "checks" are so made that they do not differ from the actual assays in any material point. Thus, being of the same quality and weight and undergoing exactly the same treatment, they may reasonably be expected to have the same surcharge as the assays they imitate. Suppose the bullion being assayed varies only a little, up or down, from 900 gold and 100 copper in the thousand, and that .5 gram of it is used in each assay. A quantity of gold differing only a little from .450 gram would be very exactly weighed and placed with .050 gram of copper in the same weight of lead as is being used in the other assays. It would be cupelled, parted, &c., as nearly as possible under the same conditions as the actual assays. Suppose the pure gold weighed .45016 gram and the parted gold weighed .45025 gram, the gain in weight, .00009 gram, would be deducted from the actual assays. A surcharge correction is never applied except to bullion of the same quality as that represented by the "check assay" it was calculated from.

It is evident that unless the gold is of the highest degree of purity these check assays will introduce an error almost equal to that which it is designed to remedy. Moreover, to work the checks to the greatest advantage, a very systematic and uniform method of working must be adopted.

~Parting in special apparatus.~--One plan for obtaining greater uniformity is to stamp each cornet with a number for purposes of identification, and to treat several, including one or more check assays in the same acid contained in a beaker; all the assays under these conditions evidently receive precisely the same acid treatment. Such a plan can of course only be adopted where there is no risk of the gold breaking up during the parting. An improvement on this is to have a porcelain basin[29] about 8-1/2 inches in diameter and with a capacity of about 1-1/2 litres. It is provided with a porcelain cover with 30 numbered holes through which tubes dip into the acid. The cover is removable. The tubes are like test-tubes and are supported by the cover; their bottoms are perforated with holes or slits. The acid is placed in the basin and boiled over a flat burner; it enters the tubes through the slits. The cornets are placed each in its proper tube. When the boiling is finished, the cover with the tubes is lifted and at the same time the acid drains back into the basin. A dip into a basin of distilled water washes at one operation all 30 assays. The cover is then put on a basin containing the stronger parting acid which is already boiling. This boiling is continued for half an hour. The cover with the 30 cornets is then lifted out from the acid and dipped two or three times in distilled water to wash off the last traces of acid. To transfer the cornets from the tubes to the porous cups the whole of the tube must be dipped under the water; otherwise the operation is exactly as when working with test-tubes.

A still simpler method of working is to use small platinum cups[30] provided with fine slits which admit the acid but retain the gold. A number of these, say 60, are supported on a platinum tray. The parting acids are boiled in platinum dishes under a hood; and the 60 cornets (each in its proper cup) are placed in the acid all at once: the tray carrying the cups is provided with a handle suitable for this purpose. After a proper boiling the tray is lifted out of the weaker acid into the stronger one, where it undergoes the second boiling. It is next dipped several times in distilled water and lastly, after a gentle drying, it is raised to an annealing temperature which must not be too high for fear of the gold sticking to the platinum. After cooling, the cornets are transferred from the platinum cups directly to the pan of the balance. Here all 60 cornets have exactly the same treatment and the "checks" may be compared with great exactness with the other assays accompanying them. There is, too, a great saving of labour.[31]

~Silver, &c., in gold bullion.~--The base metals are generally determined by cupelling .5 gram of the alloy with 5 grams of lead. The loss in cupellation having been allowed for by any of the usual methods (see p. 104) the gold and silver contents are given. By deducting the gold the proportion of silver is obtained. The silver is generally determined by difference in this way. If it is desired to dissolve out the copper, silver, &c., and to determine them in the wet way, the gold must first be alloyed with a sufficiency of some other metal to render it amenable to the attack by acid. Cadmium is the metal generally recommended, and the alloy is made by melting together a weighed portion of the gold with five or six times its weight of cadmium in a Berlin crucible and under a thin layer of potassium cyanide.

~Lead with gold or silver.~--Large quantities of lead carrying gold and silver are sold to refiners in bars weighing about 100 lbs. each. The assay of these alloys presents no special difficulties, but the sampling of them is a question which may be profitably discussed.[32]

A molten metal may be conceived to have all the physical states observed in ordinary liquids, although these cannot be actually seen owing to its opaqueness. There is no doubt that _pure_ lead at a temperature only a little above its melting-point can contain a large proportion of gold in such a manner that it may in a figurative way be spoken of as a clear solution. Any small portion withdrawn from the molten metal would afford a perfect sample. The same would be true of any pure alloy of lead and silver in which the silver does not exceed the proportion of 2-1/2 per cent.[33] On the other hand, if the molten metal contains much more than .5 per cent. of zinc, more than .1 per cent. of copper, or a larger quantity of silver, it may be likened to a turbid liquor. The resemblance holds good so far that if the molten lead be further heated, whereby its solvent power on the added metal is increased, the turbidity will disappear, or at least be considerably diminished. A portion taken at random from such a molten metal may, or may not, give a good sample. The suspended insoluble matter will tend to concentrate itself in the upper or lower parts of the liquid according to whether it is heavier or lighter than it; and this separation may occur with extreme slowness or with fair rapidity. However, it is generally agreed that in the case of such alloys as occur in practice, samples taken in this way are quite satisfactory and are the best obtainable. The precautions insisted on are that the lead shall be made as hot as practicable; that it shall be stirred up at the time of taking the sample; and that the portion withdrawn shall be taken out with a ladle at least as hot as the molten metal. The further precaution that if any dross be on the surface of the metal it shall be skimmed off and separately sampled and assayed is almost too obvious to require mention. An alternative and, perhaps, better way of taking the sample is to withdraw portions at equal intervals from the stream of metal whilst the pot is being emptied; equal weights taken from these portions and mixed (by melting or in some other way) give a fair sample of the whole. In addition, separate assays of each portion will show to what extent the metal lacks uniformity in composition For example, samples taken at the beginning, middle, and end of a run gave the following results in ozs. of silver per ton: 475, 472, 466, showing an average result of 471 ozs. Fifteen fractions taken at regular intervals during the same pouring ranged from 475 ozs. to 464 ozs.: the average result was 469.8 ozs. The same lead cast into bars and sampled by sawing gave an average of 470 ozs.[34] In another case[35] samples drawn at the beginning, middle, and end of a run gave 1345 ozs., 1335 ozs. and 1331 ozs. The mean result in such cases is always a reasonably safe one, but evidently where the metal varies a good deal it is safer to take more than three dips.

Imagine such lead run into moulds and allowed to become solid as bars; the difference between bar and bar would not be greater than that between corresponding dip samples. But in each bar the distribution of the silver and gold is very seriously affected during solidification. Chips taken from the same bar of auriferous lead may show in one place 23 ozs. of gold to the ton, in another 39 ozs.; similarly with silver they may vary as much as from 900 ozs. to 1500 ozs. to the ton.

This rearrangement of the constituents of a bar takes place whilst the lead is partly solid, partly liquid. The most useful conception of such half-solidified metal is that of a felted spongy mass of skeleton crystals of comparatively pure lead saturated with a still fluid enriched alloy. If the solidification of an ingot of impure tin be watched it will be evident that the frosted appearance of the surface is due to the withdrawal of the fluid portion from a mat of crystals of purer tin which have been for some time solid and a contraction of the mass. The shrinking of the last part to become solid is further shown by the collapse of the surface of the ingot where weakest; that is, a furrow is formed on the flat surface. In other cases of fused metal there is expansion instead of contraction in this final stage of the solidification, and the enriched alloy then causes the upper face of the ingot to bulge outwards. There are other causes effecting the redistribution of the metals through the ingot. There can be no general rule of wide application showing which part of a bar is richest and which poorest in the precious metals. This will depend on the quantities of gold or silver, on the quantities and kinds of other metals present and on the manner of casting. The student is advised to consult Mr. Claudet's paper which has been already referred to.

The best method of sampling such bars is to melt them all down and to take a dip sample of the molten metal in one or other of the methods already described. According to Mr. Claudet this should be done in all cases where the gold exceeds one or two ounces or where the silver exceeds 200 ozs. to the ton. If during the melting down some dross has formed this must be skimmed off, weighed and separately sampled and assayed. The clean lead also must be weighed, sampled and assayed. The mean result must be calculated. Thus 14 tons 5 cwts. of clean lead assaying 32 ozs. to the ton will contain 456 ozs. of silver; 15 cwt. dross assaying 20 ozs. to the ton will contain 15 ozs. of silver. The 15 tons of lead and dross will contain 471 ozs. of silver or 31.4 ozs. per ton.

Of the methods of sampling which avoid melting the bars, that known as sawing is the only one which is thoroughly satisfactory. In it the bars are brought to a circular saw having fine teeth and are sawn across either completely or halfway through; in this way a quantity of lead sawdust is obtained (say 1 lb. or so from a bar) which represents exactly the average of the bar along the particular cross section taken and approximately that of the whole bar. A bar of lead, which by dip assay gave 334 ozs. to the ton, gave on three transverse sections 333 ozs., 335 ozs. and 331 ozs. The variation may be greater than this, but with a large number of bars, where each bar is cut across in as far as possible a different place, these variations tend to neutralise each other and a good sample is obtained. Two or three cwt. of sawdust may be obtained in this way; this is thoroughly mixed and reduced by quartering in the usual way or by a mechanical sampler. A sample of 2 or 3 lbs. is sent to the assayer. This being contaminated with the oil used in lubricating the saw is freed from it by washing with carbon bisulphide, ether or benzene and dried. Then, after mixing, 100 to 200 grams of it are carefully weighed and placed in a hot crucible, the heat of which should be sufficient to melt all the lead. The molten lead should not be overheated and should show no loss due to the melting. The removal of the oil may have decreased the weight by perhaps one half per cent. If the lead gives dross on heating it may be melted under 10 or 20 grams of potassium cyanide, which prevents the formation of dross. Samples are sometimes taken with a drill, gouge or chisel, though no method of this kind is quite satisfactory. One plan adopted is to use a punch which, when driven into the bar, gives a core or rod of metal about half as long as the bar is thick and about one-eighth of an inch across. With five bars side by side it is customary to drive in the punch at one end on the first bar, and at the opposite end on the last one, and on the others in intermediate positions in such a manner that all the holes will be along a diagonal of the rectangle enclosing the bars. The bars are then turned over and similar portions punched out through the bottoms of the bars and along the other diagonal. Or one set of five may be sampled along the top and the next set along the bottom of the bars.

Silver and gold present in bars of copper are subject to the same irregularity of distribution as in lead. The sampling of such bars is guided by the same principles.[36]

CYANIDES.

The cyanides ought perhaps to be considered along with chlorides, bromides and iodides in Chapter XV. But they are treated here because they owe their importance to their use in the extraction of gold and because their determination has become a part of the ordinary work of an assayer of gold ores.

Formerly, the cyanide most easily obtained in commerce was potassium cyanide; and it was generally sold in cakes which might contain as little as 40 per cent. or as much as 95 per cent. of the pure salt. It became customary to express the quality of a sample of commercial cyanide by saying it contained so much per cent. of potassium cyanide. The commercial product now made by improved methods of manufacture is actually sodium cyanide, but is called "potassium cyanide" (probably with the words "double salt" on the label); it contains cyanide equivalent to something over 100 per cent. of potassium cyanide in addition to a large proportion of sodium carbonate and other impurities. What is wanted in most cases is merely a soluble cyanide, and it is a matter of indifference whether the base be sodium or potassium. But since 49 parts of sodium cyanide (NaCN = 49) are equivalent to 65 parts of potassium cyanide (KCN = 65) it is evident that a pure sample of sodium cyanide would contain cyanide equivalent to little less than 133 per cent. of potassium cyanide. Therefore a sample of cyanide reported on in this way may be rich in cyanide, and yet have much impurity.

The commonest impurity in commercial cyanide is carbonate of sodium or potassium. This may be tested for by dissolving, say, 2 grams in a little water and adding barium chloride. There may be formed a white precipitate of barium carbonate, which if filtered off, washed and treated with acid, will dissolve with effervescence. Cyanate may be tested for in the solution from which the barium carbonate has been filtered by adding a little soda and boiling; if cyanates are present they decompose, giving off ammonia (which may be tested for in the steam) and yielding a further precipitate of barium carbonate.[37] If the soda alone gave a further precipitate of barium carbonate, this may, perhaps, be due to the presence of bicarbonates. Alkaline sulphides may be present in small quantity in commercial cyanide. Their presence is shown at once when the sample is being tested for its strength in cyanide, inasmuch as the first few drops of silver nitrate solution produce at once a darkening of the liquor. A special test for sulphide may be made by adding a drop or two of solution of acetate of lead to four or five c.c. of soda solution and adding this to a clear solution of the suspected cyanide. This will cause a black precipitate or colour, if any sulphide is present.

The cyanides of the heavier metals combine with the alkaline cyanides to form double cyanides. Some of these, ferrocyanide and ferricyanide of potassium for example, have such characteristic properties that the fact that they are cyanides may be overlooked. Others, such as potassium zinc cyanide (K_{2}ZnCy_{4}), have much less distinctiveness: they behave more or less as a mixture of two cyanides and are, moreover, so easily decomposed that it may be doubted if they can exist in dilute alkaline solutions. In reporting the cyanide strength of a cyanide liquor as equivalent to so much per cent. of potassium cyanide, there is a question as to whether the cyanide present in the form of any of these double cyanides should be taken into account. It must be remembered that the object of the assay is not to learn how much of the cyanide exists in the solution as actual potassium cyanide; reporting the strength in terms of this salt is a mere matter of convenience; what is really desired is to know how much of the cyanide present in the liquor is "free" or "available" for the purposes of dissolving gold. Every one is agreed as to the exclusion of such cyanides as the following: potassium ferrocyanide (K_{4}FeCy_{6}), potassium ferricyanide (K_{3}FeCy_{6}), potassium silver cyanide (KAgCy_{2}), and potassium aurocyanide (KAuCy_{2}); and the double cyanides with copper or nickel. But with cyanide liquors containing zinc the position is less satisfactory. One method of assay gives a lower proportion of cyanide when this metal is present; and the loss of available cyanide thus reported depends, though in a fitful and uncertain way, upon the quantity of zinc present. The other method of assay reports as full a strength in cyanide as if no zinc were present. Unfortunately, using both methods and accepting the difference in the results as a measure of the quantity of zinc present, or at any rate of the zinc present as cyanide, is not satisfactory. It appears best to use the method which ignores the zinc; and to determine the amount of zinc by a special assay of the liquor for this metal.

The cyanide present as hydrogen cyanide or prussic acid (HCy) is practically useless as a gold solvent. Hence any report on the strength of a cyanide liquor which assigned to this the same value as its equivalent of alkaline cyanide would be misleading. On the other hand, it is "available cyanide" inasmuch as a proper addition of sodium hydrate[38] would restore its value. The question of the presence or absence of free prussic acid is involved in the larger one as to whether the cyanide solution has the right degree of alkalinity. The assay for "cyanide" should include the hydrogen cyanide with the rest.

A rough test of the power of a cyanide liquor for dissolving gold may be made by floating a gold leaf on its surface and noting the time required for its solution. This test might, perhaps, be improved by taking, say, 20 c.c. of the liquor and adding three or four gold leaves so that the gold shall always be in considerable excess. The liquor should not be diluted as this will affect the result. It should be allowed to stand for a definite time, say at least two or three hours, or better, that corresponding to the time the liquor is left in contact with the ore in actual practice. The liquor should then be filtered off and, with the washings, be evaporated in a lead dish as in the assay of cyanide liquors for gold (p. 141). The gold obtained on cupelling, less any gold and silver originally present in the liquor, would be the measure of the gold dissolving power.

THE ASSAY FOR CYANIDE BY TITRATION WITH SILVER NITRATE.

The determination of the quantity of a cyanide is made by finding how much silver nitrate is required to convert the whole of the cyanide into potassium silver cyanide[39] or one of the allied compounds. It will be seen from the equation that 170 parts by weight of silver nitrate are required for 130 parts by weight of potassium cyanide. As already explained it is customary to report the cyanide-strength in terms of potassium cyanide, even when only the sodium salt is present. One gram of potassium cyanide will require 1.3076 gram of silver nitrate. _The standard solution of silver nitrate_ is made by dissolving 13.076 grams of silver nitrate in distilled water and diluting to 1 litre; 100 c.c. of such a solution are equivalent to 1 gram of potassium cyanide.[40]

The titration is performed in the usual way, running the standard solution of silver nitrate into a solution containing a known weight or volume of the material containing the cyanide. The _finishing point_ is determined in one of two ways, both of which are largely used. In the first place, as long as there remains any free cyanide in the solution the silver nitrate will combine with it forming the double cyanide and yielding a clear solution; but as soon as all the free cyanide is used up the silver nitrate will react with the double cyanide[41] forming silver cyanide, which separates as a white precipitate and renders the solution turbid. But, in the second place, if potassium iodide is present in the solution the excess of silver nitrate will react with it,[42] rather than with the double cyanide; and silver iodide will separate as a yellowish turbidity which is easily recognised.

In working with pure solutions, the two finishing points give the same results; and this is true even when there is much difference in the degree of dilution. The finishing point with the iodide, however, has an advantage in precision. Moreover, it is but little affected by variations in alkalinity, which render the other finishing point quite useless. The great difference between the two is shown when zinc is present in the solution. In this case, when working without the iodide, the first appearance of a turbidity is less distinct; the turbidity increases on standing and as a finishing point is unsatisfactory. It can be determined with precision only by very systematic working and after some experience. The turbidity is due to the separation of an insoluble zinc compound. A most important point (to which reference has already been made) is that less silver nitrate is required to give this turbidity and, consequently, a lower strength in cyanide is reported. On the other hand, as much silver nitrate is required to give the yellow turbidity due to silver iodide as would be required if no zinc were present.

Unfortunately the difference in the two titrations does not depend merely on the quantity of zinc present; as it is also influenced by the extent of dilution, the degree of alkalinity of the solution, and the quantity of cyanide present. In an experiment with .055 gram of zinc sulphate and .1 gram of potassium cyanide the difference in the two finishing points was only .1 c.c.; whereas with .4 gram of potassium cyanide, the other conditions being the same, the difference was 1.5 c.c. of standard silver nitrate. On the assumption that all the zinc was present as potassium zinc cyanide (K_{2}ZnCy_{4}) the difference should have been 5 c.c. in each case. Again, repeating the experiment with .4 gram of potassium cyanide, but with .11 gram of crystallised zinc sulphate, the difference was 6.5 c.c.: that is, merely doubling the quantity of zinc increased the difference by more than four times. Hence it would appear better to use the method with the iodide and make a separate assay for the zinc. But since the student may be called on to use the other method, he is advised to practice it also.

~The assay without iodide.~--The standard solution of silver nitrate is placed in a small burette divided into tenths of a c.c. Ten c.c. of the cyanide solution to be assayed is transferred to a small flask and diluted with water to about 70 c.c. The silver solution is then run in from the burette (with constant shaking of the flask), a little at a time but somewhat rapidly, until a permanent turbidity appears. Since 1 c.c. of the silver nitrate solution corresponds to .01 gram of potassium cyanide, it also corresponds to .1 per cent. of this salt counted on the 10 c.c. of cyanide solution taken. The titration should be performed in a fairly good uniform light. The learner should practice on a fairly pure solution of potassium cyanide at first, and this may conveniently have a strength of about 1 per cent. For practice with solutions containing zinc make a solution containing 1.1 gram of crystallised zinc sulphate in 100 c.c. and slowly add measured quantities of from 1 to 5 c.c. of this to the 10 c.c. of cyanide liquor before diluting for the titration.

If a cyanide solution blackens on the addition of the silver nitrate it contains sulphide. In this case, shake up a considerable bulk of the liquor with a few grams of lead carbonate, allow to settle and make the assay on 10 c.c. of the clear liquor.

If the cyanide liquor be suspected to contain free prussic acid, take 10 c.c. for the assay as usual; but, before titrating, add .1 or .2 gram of sodium carbonate. On no condition must caustic soda or ammonia be added. The difference between the results, with and without the addition of carbonate of soda, is supposed to measure the quantity of free prussic acid. If this has to be reported it is best done as "prussic acid equivalent to ... per cent. of potassium cyanide." Suppose, for example, the difference in the two titrations equals 1 c.c. of standard silver nitrate; the prussic acid found would be equivalent to .1 per cent. of potassium cyanide.

~The assay with iodide.~--The standard solution of silver nitrate is placed in a burette divided into tenths of a c.c. Take 10 c.c. of the cyanide liquor, which should previously have been treated with white lead for the removal of sulphides if these happened to be present. Transfer to a small flask, add 3 or 4 drops of a solution of potassium iodide and 2 or 3 c.c. of a solution of sodium hydrate; dilute to 60 or 70 c.c. with water. If much zinc is present the soda may be increased to 20 or 30 c.c. with advantage. The standard solution should be run in somewhat rapidly, but a little at a time, so that the precipitate at first formed shall be small and have only a momentary existence. The titration is continued until there is a permanent yellowish turbidity. The most satisfactory and exact finish is got by ignoring any faint suspicion of a turbidity and accepting the unmistakable turbidity which the next drop of silver nitrate is sure to produce. This finishing point gives results which are exactly proportional to the quantity of cyanide present; and it can be recognised with more than ordinary precision even in solutions which are not otherwise perfectly clear.

Each c.c. of the standard silver nitrate solution corresponds to .01 gram of potassium cyanide; and if 10 c.c. of the liquor are taken for assay this corresponds to .1 per cent. or 2 lbs. to the short ton or 2.24 lbs. to the long ton. As already explained the result should be reported as "cyanide equivalent to so much per cent. of potassium cyanide."

The following experimental results were obtained with a solution of potassium cyanide made up to contain about 1.2 per cent. of the salt.

~Effect of varying cyanide.~--The bulk before titration was in each case 60 c.c.; 2 c.c. of soda and 3 drops of potassium iodide were used in each case.

Cyanide added 40 c.c. 30 c.c. 20 c.c. 10 c.c. 5 c.c. 1 c.c.
Silver required 47.0 c.c. 35.25 c.c. 23.5 c.c. 11.7 c.c. 5.8 c.c. 1.15 c.c.

Accepting the result for 40 c.c. as correct, the others are in very satisfactory agreement.

~Effect of varying dilution.~--The conditions were those of the 40 c.c. experiment in the last series; but varying amounts of water were used in diluting.

Water added none 100 c.c. 200 c.c. 400 c.c.
Silver required 47.0 c.c. 47.0 c.c. 47.0 c.c. 47.05 c.c.

Very considerable dilution therefore has no effect.

~Effect of varying soda.~--The conditions were those of the 40 c.c. experiment in the first series, except that varying amounts of soda solution were used.

Soda added none 10 c.c. 30 c.c.
Silver required 46.95 c.c. 47.0 c.c. 47.0 c.c.