UNIVERSITY OF CALIFORNIA. 
 
 aass A1.U 
 
 x 5^.0 
 
The Fire Assay 
 
 of Gold, Silver, and Lead in Ores and 
 Metallurgical Products 
 
 by 
 
 Leonard S.[Austin 
 
 Professor of Metallurgy and Ore-dressing, 
 Michigan College of Mines. 
 
 First Edition 
 
 1907 
 
 Published by the 
 
 Mining and Scientific Press 
 
 San Francisco 
 
GENEBAL 
 
 COPYRIGHT, 1907 
 
 BY 
 MINING AND SCIENTIFIC PRESS 
 
TABLE OF CONTENTS 
 
 PAGE 
 
 Preface 7 
 
 I. The Fire Assay of Ores or Furnace and Mill Products 
 
 Containing Gold, Silver, and Lead 11 
 
 II. 'Sampling and Preparation of the Ore for Assay 12 
 
 III. Care of the Assay Office 15 
 
 IV. Apparatus ; r ;; 17 
 
 V. The Assay Furnace ." 22 
 
 VI. Crucibles and Scorifiers 30 
 
 VII. Assay Balances 32 
 
 VIII. Fluxes Used in Assaying 40 
 
 IX. Ores 48 
 
 X. The Scorification Assay 50 
 
 XL Cupelling 54 
 
 XII. Parting 56 
 
 XIII. The Crucible Assay 60 
 
 XIV. Roasting of Ores 70 
 
 XV. Assay of Matte 72 
 
 XVI. High-Grade Silver-Sulphide Assay 73 
 
 XVII. Assay of Cyanide Solutions 74 
 
 XVIII. Assay of Base r Bullion - 76 
 
 XIX. Assay of Silver Bars or Ingots 78 
 
 XX. Assay of Blister or Pig-Copper Containing Silver and 
 
 Gold 82 
 
 XXI. Assay of Gold Bullion 84 
 
 XXII. Assay of Ores Containing Metallics 85 
 
 XXIII. The Lead Assay 87 
 
 179709 
 
PREFACE 
 
 The following pages present a system of assaying, intended 
 to cover determinations of the precious metals (silver and gold) 
 and of lead, according to methods quite commonly recognized 
 in the Rocky Mountain States as suited to ores and products 
 containing the just-named metals. 
 
 In reading books on the subject, the beginner in assaying is 
 embarrassed in deciding which one of the various prescribed 
 charges he is to employ. It has, therefore, been the aim of the 
 author in this book to carry through a single system, which, while 
 not the only one, would at least be definite and clear to the learner. 
 
AUSTIN'S FIRE ASSAY 
 
I. THE FIRE ASSAY OF ORES OR FURNACE AND 
 
 MILL PRODUCTS CONTAINING GOLD, 
 
 SILVER, AND LEAD. 
 
 The fire assay is a dry or fusion method for the determination 
 of the metal sought with the aid of heat and suitable fluxes, and 
 of weighing it in metallic form. Thus if we desire to determine 
 lead in an ore we fuse a known weight of the ore in a crucible to- 
 gether with suitable fluxes and obtain a lead button which, upon 
 weighing, will give us the percentage of contained lead in the 
 portion of ore so taken, called the sample. This sample must 
 be a true average of the lot of ore whose value we wish thus to 
 determine. 
 
 The steps in assaying are: 
 
 1. Receiving and labelling the ore. 
 
 2. Sampling. 
 
 3. Fluxing, melting, and cupelling. 
 
 4. Weighing the metal obtained. 
 
 5. Recording and reporting results. 
 
 Generally but one metal is determined in an assay, except that 
 in the case of gold and silver, the two metals, recovered as one, 
 are subsequently parted. 
 
 The work should be systematically arranged and carried out. 
 The secret of doing a day's work well, both in quantity and in 
 accuracy, depends largely upon this. 
 
 Never report results where there is reason to believe that they 
 may be incorrect owing to faulty manipulation, to losses, or to 
 the mixing of samples. 
 
 Single assays may be made upon grab samples when an ap- 
 proximation is sufficient, but for results involving the purchase 
 of metals or ores, duplicates should be run which must closely 
 agree, and, if not, then the work must be repeated until they do 
 so. 
 
IL SAMPLING AND PREPARATION OF THE ORE 
 
 FOR ASSAY. 
 
 All metals, ores, furnace and mill-products, which are to be 
 assayed, must first be accurately sampled. This is quite as im- 
 portant as accurate assaying, for, if the sample does not truly 
 represent the lot or quantity of ore whose value is sought, the 
 assay of it will be misleading. 
 
 The following instructions on sampling ores are here given, 
 since the assayer is often called upon to sample the ore which 
 he is about to assay. 
 
 RECEIVING THE SAMPLE. The ore, mineral, or metallurgical 
 product may come as a hand-sample, a 'grab'-sample, or as a 
 regular lot-sample from the sampling mill. Qnly regular or 
 mill-samples are, in general, assayed in duplicate. 
 
 Small tin or graniteware pans are used to receive hand- 
 samples, each being placed in a separate pan, and with a label, 
 upon which is marked the name of the owner, the name of the 
 ore, and the lot-number of the sample. Where the assayer has 
 to number the lots himself, he should also describe the ore so 
 that it may be distinguished again by the sender. 
 
 It is a good practice to reserve one or more of the larger pieces, 
 after having partly broken down a sample, since in this way we 
 may be better able to designate or otherwise recognize the ore. 
 
 The hand-sample may consist of one or of several lumps, of a 
 mixture of fine and of lumpy ore, or of a regularly ground sample 
 of ore ready for assaying. In the latter case, the assayer should 
 be especially on his guard to exercise due care, since another 
 portion of the same 'pulp' can be used as a check on his work 
 at another time or by another assayer. 
 
 PREPARATION OF .THE SAMPLE. A general principle in sam- 
 pling is that the ratio between the larger pieces in the sample 
 and the weight of the sample should not exceed a certain amount. 
 
AUSTIN S FIRE ASSAY. 13 
 
 This is attained by breaking the pieces finer as the bulk of the 
 sample is decreased. 
 
 The breaking up of the lumps of a sample may be accom- 
 plished by means of a laboratory crusher, by pestle and mortar, 
 or by breaking on the grinding-plate with a hammer or muller. 
 In the latter case it is well to provide a wooden frame, four 
 inches high, and but two inches high at the front, which encloses 
 the area of the plate, and which prevents the ore from escaping 
 while being broken up. When the quantity of the sample is 
 considerable, some time and labor may be saved by not breaking 
 so finely at first, but by breaking smaller each time the sample 
 is cut down. 
 
 The work of cutting down or of quartering may be performed 
 as follows : The ore is made up into a coned pile in the centre 
 of the plate, the material being poured or thrown upon the apex 
 of. the pile so as to effect an even distribution of fine and coarse. 
 The cone is now flattened out to a circular cake or disk by 
 scraping the ore outward every way from its apex. Two lines 
 are marked at right angles, dividing the mass into quarters. The 
 two opposite quarters are then removed, rejected, and the re- 
 mainder mixed and quartered as before, taking care, however, 
 first to break the pieces smaller if necessary. 
 
 Finally, as this operation is continued, there remain two to four 
 ounces of the ore. This is ground to pass an 8o-mesh screen. 
 Where the ore is rich, spotty, or contains coarse particles of gold, 
 it is better to use a I2omesh screen. The sample must now be 
 dried, if riot already so. The final mixing of the sample, after 
 it has been ground up, is done by 'rolling' it on a piece of oil-cloth 
 or of rubber cloth. This is performed by laying or folding over 
 the cloth so as to cause the ore to travel from side to side. It 
 is then rolled at right angles to its former direction, and thus 
 mixing proceeds until the operator is sure that the ore is quite 
 uniform in all its parts. A beginner is very apt not to move or 
 roll the ore sufficiently to insure its complete turning over. For 
 practice, it would be well for him to mix some argols and soda 
 in this way, until the absolutely uniform color of the mixture 
 indicates that the mixing is complete. 
 
14 AUSTINS FIRE ASSAY. 
 
 An expeditious and satisfactory way of cutting down a sample 
 consists in the use of the tin sampling-trough or riffle. The ore 
 is scattered into the troughs from a shovel in such wise as not to 
 overtop the troughs. Half of the ore remains in the riffle, and 
 half drops through. When the troughs are full the ore in them is 
 reserved, and more is again added, and so on, until the sample 
 has been all put through. This reserved portion again goes 
 through the riffle, thus again halving it. Of course when finer 
 breaking is needed it must be done. 
 
 Prospectors and others, interested in mines, often deceive them- 
 selves by thinking that by sending in a piece or lump of rock, they 
 can arrive at an idea of the value of the ore they are going to 
 ship. If one will assay separately two such pieces of rock, he 
 will see how much they vary from one another. The only way to 
 know the value of an ore is to sample it in quantity, so that all 
 parts of the pile are equally represented. 
 
HI. CARE OF THE ASSAY OFFICE. 
 
 When the work of sweeping out or cleaning the assay office 
 is done by an attendant, the assayer should take notice that it is 
 thoroughly done. The balances, however, to keep them clean and 
 in the best of order, need his personal attention. As the assayer's 
 work proceeds through the day, he will find that dust, ashes, and 
 scattering particles get on the desk and about the furnace. He 
 has to keep his work going steadily, but there are spare moments 
 when he can brush up and set things in order again. In fact, 
 he should put his tools and appliances in place again as he goes 
 along, and thus, nothing being mislaid, valuable time is saved. 
 
 It assists in the prestige of the assayer to keep his office con- 
 stantly in order. If this is done, his employer, whose good-will 
 he values, will notice these things in a way favorable to him; 
 indeed it is one of the ways of showing that he is ambitious and 
 energetic. 
 
 Clinkers will form upon the interior of the furnace, and these 
 are to be removed with a long-handled cutter-bar or chisel. The 
 assayer also replaces his own muffles, claying up the joint with 
 fire-clay, to which has been added its own weight of sand, or 
 preferably, some coarsely ground fire-brick which he can break 
 up for himself on the grinding-plate. In a reduction works 
 where tailing, slag, or low-grade products are treated, there 
 should be two cast-iron bucking plates, one of which should be 
 reserved for this use. In case a high-grade ore is followed by a 
 low-grade one the assayer must use some sand, coke, or other 
 similar material to be ground on the plate and then brushed off, 
 in order to insure the entire removal of particles of the rich ore. 
 Otherwise, low-grade samples may get 'salted/ This may ac- 
 count for cases where a trace, or even more of metal, has been 
 reported when in fact the material was quite barren. 
 
 When auriferous placer gravel is to be assayed, a considerable 
 sample of a known weight, say 100 assay tons (6.38 lb.), should 
 
16 AUSTIN'S FIRE ASSAY. 
 
 be panned down carefully and the residual concentrate, consisting 
 of black sand, pyrites, and gold, assayed. Even where the gravel 
 has been broken up and mixed, the results are apt to be uncertain, 
 since particles of gold are often lodged in the crevices of the 
 pebbles, and only water treatment will wash them out. 
 
IV. APPARATUS. 
 
 GRINDING PLATE AND MULLER. For breaking, quartering 
 down, and pulverizing samples of ore the grinding or bucking 
 plate is employed, as shown in Fig. I. It is of cast-iron, planed on 
 its upper surface, one inch thick, and has raised edges. The 
 waste part of the sample is swept into a box placed beneath the 
 bench. The muller, which should weigh 25 lb., has a curved 
 rubbing surface, and is rubbed back and forth on the plate to 
 pulverize the ore. Where the pieces are one-fourth inch or larger, 
 they are broken, using the muller as a hammer. Quartering down 
 
 FIG. 1. GRINDING PLATE AND MULLER. 
 
 of the sample is performed on the plate using a piece of Russia 
 sheet iron 3 by 6 in. as a scraper. 
 
 Sieves or screens may be 10 in. diam. with brass screen-cloth 
 and having a removable pan-bottom (See Fig. 2). When the 
 
i8 
 
 AUSTIN S FIRE ASSAY. 
 
 sample has been pulverized it is mixed or rolled on a piece of 
 oil-cloth or a rubber cloth 18 in. square. Neglect of this pre- 
 caution of thorough mixing means varying results in duplicate 
 assays. * 
 
 FIG. 2. TEN-INCH SIEVE AND BOTTOM. 
 
 FIRE TOOLS. For handling crucibles, etc., several kinds of tools 
 are used as shown below: 
 
 Crucible tongs are employed for setting, removing, and pour- 
 
 FIG. 3. CRUCIBLE TONGS. 
 
 ing crucibles. They have bent points for convenience in seizing 
 the crucible and for pouring them. A pair of such tongs is shown 
 in Fig. 3. 
 
 FIG. 4. SHORT CRUCIBLE TONGS. 
 
 Fig. 4 shows the short tongs used for removing nails from 
 the crucible before pouring. They are nine inches long. 
 
AUSTINS FIRE ASSAY. 19 
 
 CRUCIBLE TONGS FOR MUFFLE. (See Fig. 5.) These have 
 curved ends for embracing the crucible and are useful in putting 
 in and removing crucibles (sometimes scorifiers) from the muffle. 
 They have in mid-length a flat tongue for steadying the two legs 
 in relation to one another. The round wire tongue, ordinarily 
 provided, is not satisfactory. 
 
 FIG. 5. CRUCIBLE TONGS FOR MUFFLE USE. 
 
 For handling scorifiers we have scorifier-tongs, of which Fig. 
 6 is a good example. It has a forked leg as well as a flat one, 
 and has a broad spring at the back. The flat end is often used to 
 dip up and transfer borax-glass to the scorifier. 
 
 FIG. 6. SCORIFIER TONGS. 
 
 Fig. 7 represents a pair of cupel-tongs more carefully made 
 and proportioned than those commonly purchased from dealers 
 in assay supplies. It has a broad and flexible spring back, and a 
 
 FIG. 7. CUPEL TONGS. 
 
 flat tongue for steadying the legs. The bent points should be no 
 longer than here given. The assayer can easily cut off and shape 
 the points when, as is usual, they are too long. For large muffles 
 they should be at least 30 in. long. 
 
2O 
 
 AUSTIN S FIRE ASSAY. 
 
 MUFFLE SCRAPER. This tool, intended for promptly scraping 
 out anything spilled upon the floor of the muffle, should hang 
 near the muffle so that it can be used without a moment's delay. 
 In this way the damage to the muffle can be greatly lessened. It 
 is shown in Fig. 8. 
 
 FIG. 8. SCRAPER. 
 
 CUPEL MOLDS. Cupels are made by the assayer (though they 
 can also be bought) from bone-ash, ground fine enough to give 
 a smooth surface and yet have a porous texture. The molds, 
 1^2 in. diam., are in two parts, the ring and the plunger, and are 
 generally made of brass. The assayer must take care to handle 
 them so as not to bruise or deface them in any way. It will be 
 noticed that the ring has an interior taper for the easy removal 
 
 FIG. 9. TWO-HOLE SLAG MOLD. 
 
 of the cupel when molded, and, in making the cupel, the plunger 
 enters the smaller diameter. In making cupels, the bone-ash is 
 
 FIG. 10. SIX-HOLE SLAG MOLD. 
 
 moistened with a little water until it has a grip or adhesion such 
 that, when molded, it will cohere. Before use, cupels must be 
 carefully dried by being put in a warm place, preferably for a 
 month or more, so that quite a stock must be made and kept on 
 hand. There are cupel-making machines which will turn out 
 
AUSTINS FIRE ASSAY. 21 
 
 several hundred cupels in an hour, but for a small assay office, 
 they are generally made in the simpler way. 
 
 SLAG MOLDS. They are made of various shapes, some having 
 a hemispherical, some a conical, cavity. The most satisfactory 
 mold is one having the conical cavity, smoothly turned out, and in 
 a solid cast-iron block, so as quickly to abstract the heat from the 
 molten material poured into it. (See Fig. 9.) Fig. 10 represents 
 a similar mold having six depressions. 
 
 Lighter and cheaper molds, having less iron in them, are not 
 so satisfactory. 
 
V. THE ASSAY FURNACE. 
 
 A variety of assay furnaces have been devised, some using soft 
 coal as a fuel and depending upon the flame from the coal to do 
 the heating, some, in which coke is used, with coals in direct 
 contact with the muffle or the crucible, and some where the flame 
 of gasoline is used, much on the same principle as with soft coal. 
 To these may be added ordinary illuminating gas, where, as in 
 cities, it may be had. 
 
 FlG. 11. COAL-BURNING ASSAY FURNACE. 
 
 Fig. ii is an elevation, and Fig. 12 and 13, two sectional ele- 
 vations of a two-high muffle furnace for burning soft coal, which 
 
AUSTINS FIRE ASSAY. 23 
 
 must have enough volatile constituents to produce a suitable 
 flame. The interior is lined with fire-clay tile of special shapes 
 so that the furnace can be readily and speedily built. The rest 
 of the furnace is of red brick, the whole being firmly clamped 
 or bound with angle-iron and tie rods. The muffles, of the form 
 shown in Fig. 12 and 13, have an inside width of 9 in., and each 
 
 FK;. 12 AND 13. SECTIONS or- COAL-BURNING ASSAY FURNACE. 
 
 one is large enough to take two 2O-gram crucibles (3 in. diam. 
 by 3^4 in- high) abreast, and in at least four rows, making twelve 
 crucibles at a time or five rows of scorifiers, four abreast, or 
 easily four cupels abreast. The lower muffle, which can be 
 brought to the higher temperature, is generally reserved for melt- 
 ing and the upper one for cupelling. By means of a tile-damper 
 the assayer can regulate the heat of the furnace, so that, in 
 cupelling, he can carry several rows of cupels at a time. He then 
 subdues the heat at the back of the muffle by putting in cold 
 crucibles which absorb the heat and cool down the muffle to 
 exactly the. required temperature. The muffles are carried on sup- 
 ports, two on either side. At the front will be noticed a short flue 
 for removing the fumes which may issue from the muffles. The 
 furnace should have a stack with a strong draft, this stack being 
 built independently of the furnace. The fire-door is placed so 
 
24 AUSTINS FIRE ASSAY. 
 
 as to feed the coal at the back, and out of the way of the assayer. 
 It can also be arranged so as to feed at the side if so desired. 
 
 With a slight change in the fire-box, this furnace can burn 
 wood, there being then but a single muffle used. These furnaces 
 take 16 to 24-in. wood as desired. 
 
 Where much work has to be done, furnaces having much larger 
 muffles are used, where as many as four 2O-gm. crucibles can be 
 put in abreast, five rows being melted at a time. 
 
 In firing with coal the bed should be not more than 6 in. deep, 
 a little deeper at the back, using but little coal at a time and feed- 
 
 FIG. 14. BROWN PORTABLE ASSAY FURNACE. 
 
 ing frequently so as to keep up a steady flame. Be careful not 
 to permit any holes through the fuel bed by which cold air can 
 get up to cool the muffle. The fire must be kept clean and free 
 from ashes and cinders. As those latter accumulate in the ash-pit 
 they must be removed. Otherwise, the heat from them softens 
 and melts the grate bars, thus soon destroying them. 
 
AUSTIN'S FIRE ASSAY. 25 
 
 Fig. 14 is a perspective view of the well-known Brown portable 
 furnace for coke. The pipe coming from it is 5 in. diam., and can 
 be carried through the roof of an assay office, so that one can be 
 quickly set up and put in operation. It contains a single muffle. 
 
 The muffle is suited to scorifying and cupelling, but not to 
 receiving crucibles. Crucibles are, accordingly, set among the 
 coke through the top door or lid. The crucibles are set in two 
 rows parallel to the muffle, three pointed E crucibles in a row. 
 They are set to leave 1.5 in. between them and the side of the 
 furnace, and are packed all around with coke to their tops. It 
 is upon the thorough performance of this work that the success 
 of the melting depends. The fire is lit with a little paper and some 
 light kindling followed by heavier pieces at least an inch thick and 
 packed under and up to the muffle. Charcoal is then put in to 
 
 FIG. 15. Bos WORTH ASSAY FURNACE. 
 
 the level of the top of the muffle and then two inches in depth 
 of coke broken to egg size or less. If. left larger, these pieces 
 will not pass down between the muffle and the wall of the furnace. 
 The fire is then lit, and, when the wood has burned away, is 
 poked down with a poker ^ in. diam, by 3^2 ft. long. More 
 coke is then added, and, when burning, is worked down with the 
 
26 
 
 AUSTIN S FIRE ASSAY. 
 
 poker. To do this the poker is thrust down lower than the muffle 
 and is then carried through the coke with a motion tending to lift 
 up the larger pieces while the smaller ones fall by and pack the 
 space below the muffle. Fresh coke is now put in up to the level 
 of the outlet flue and the crucibles are at once set, digging out 
 a hole for them with the poker and tongs and setting them among 
 the cold coke. With a good draft the coke is soon burning and 
 the charge melting down. As soon as the muffle gets sufficiently 
 hot, cupelling or scorifying can proceed. When the crucibles 
 have been poured, the poker is again used to work down the coals, 
 fresh coke is put in and a new batch of crucibles set. After their 
 removal, poke down and replenish the fire before cupelling. 
 
 Even if there are no more crucible-melts to be made, still the 
 fire must be poked down and replenished before cupelling. 
 
 FIG. 16. FRONT ELEVATION OF WIND-FURNACE. 
 
 Fig. 15 represents a Bosworth furnace for a 9 by I5~in. muffle. 
 consisting of a base, a body, and a top, on top of which may be 
 noticed the outlet connecting to the chimney by means of a 5 -in. 
 stove-pipe. The top is covered by an iron door and, when the 
 furnace is in full operation, it is filled with coke to the sill of the 
 feed opening. . The iron door covers the muffle, which latter 
 
AUSTINS FIRE ASSAY. 27 
 
 extends clear to the back of the furnace and is capable of taking 
 j three rows of 2O-gm. crucibles, two in a row, as seen in the cut. 
 The plug immediately above the grate is kept closed except in 
 cleaning out the furnace, and also when it is desired to poke the 
 glowing coke and work it ' beneath the muffles. 
 
 FIG. 17. SIDE ELEVATION OF WIND-FURNACE. 
 
 Fig. 1 6 and 17 represent a wind or natural-draft furnace in- 
 tended for melting only. A thin fire or bed of coke is first 
 started in the furnace. Fresh coke is added and the crucibles 
 are at once set, being careful to place them to their full depth in 
 the coke, and 'to pack it around them. 
 
 GASOLINE ASSAY FURNACE. Fig. 18 represents in perspective 
 a combination gasoline furnace with burner at the rear, and the 
 tank and pump by which the burner is fed. The furnace proper 
 has at the front a muffle for cupelling, while at the top will be 
 noticed an opening with the cover removed and showing crucibles 
 in position for melting. The burner (Fig. 19) just fits into an 
 opening through the rear wall. The lower hand-wheel admits 
 a small quantity of gasoline to a tray shown below the burner, 
 where it is ignited by a match. On opening the upper hand- 
 
28 
 
 AUSTIN S FIRE ASSAY. 
 
 wheel, the gasoline, under pressure from the tank which has been 
 first of all pumped up, is heated and gasified as it enters the 
 furnace from the jet. The issuing vapor is lit by a match imme- 
 
 FIG. 18. GASOLINE ASSAY FURNACE WITH TANK AND BURNER. 
 
 FIG. 19. BURNER FOR GASOLINE FURNACE. 
 
 diately inside the furnace, and as the latter heats, this gasification 
 is kept up. For success in operating, therefore, it is necessary 
 
AUSTIN S FIRE ASSAY. 29 
 
 first to thoroughly heat the burner. The tank should be pumped 
 to 10 or I5~lb. pressure. The crucibles to be melted may be 
 placed within the furnace while it is cold and the heat brought 
 up on them gradually. It is also possible to put them into a 
 heated furnace, letting them take up the heat, and finally to start 
 the burner. These furnaces have the advantage that they are 
 quickly brought up to the full heat, which can then be retained 
 indefinitely with but little further attention. They cost, for the 
 furnace $25, for the burner $11, and for the pump, tank, pipe, 
 and fittings, $20. 
 
 _,_KV 
 
 ' OF THE 
 
 UNIVERSITY 
 
 OF 
 
VI. CRUCIBLES AND SCORIFIERS. 
 
 For ordinary assaying, but few out of the many kinds adver- 
 tised are made use of. Thus, for work in the muffle, but two 
 sizes are used, the 10 and the 2O-gm. crucibles. (See Fig. 20.) 
 These crucibles are short enough to be placed in the muffle of the 
 coal-burning furnaces already described. Where the crucibles 
 are to be set in the coke, the pointed form (Fig. 21) is used, sizes 
 E and F. 
 
 FIG. 20. CLAY CRUCIBLES. (FLAT FORM.) 
 
 FIG. 21. CLAY CRUCIBLES. (POINTED FORM.) 
 
 Crucibles are often used but a single time. If still sound, as 
 when assaying silicious ores, they may be again used. However, 
 for ores or products containing but little of the desired metal, 
 new ones should be employed. Besides the domestic crucibles, 
 Battersea crucibles of English make are very durable. For lead 
 assays, in order to clearly distinguish the crucibles from those 
 
AUSTIN S FIRE ASSAY. 
 
 3 1 
 
 used for a silver assay, the triangular crucible (Fig. 22) is used. 
 These may be of the size 3.25 in. diam. by 3.5 in. high, called U 
 size. 
 
 FIG. 22. NEST OF TRIANGULAR BATTERSEA CRUCIBLES. 
 
 Crucibles are made of fire-clay, white and smooth, or of a simi- 
 lar clay into which has been mixed a good deal of sand, so that 
 they are rougher on the surface and of light-brown color. The 
 latter better resist the action of a silicious charge, the former, 
 or clay crucible, the basic charge. 
 
 In scorifying, the 2^2-in. (outer diameter) scorifier, as shown 
 in Fig. 23, is employed. They are good generally for a single 
 
 FIG. 23. SCORIFIERS. 
 
 time except for a silicious ore, and, even at the first time, may be 
 so corroded as to permit the escape of a charge. The size speci- 
 fied is large enough for assaying one-tenth of an assay ton of 
 ore, and, when larger quantities are to be taken, the crucible assay 
 
 is chosen. 
 
VIL ASSAY BALANCES. 
 
 There are three kinds of balances used in assaying the pulp, 
 button, and gold-balances. In using any of these balances the 
 variable load is placed in the right-hand pan, the fixed quantity 
 in the left-hand one. Thus ore or silver granulations would be 
 placed in the right-hand pan and buttons or fixed weights in the 
 left. The beam is raised to its full height with a sudden move- 
 ment, and, as soon as the pointer begins to move, the assayer 
 judges what change is to be made, and at once drops the 
 beam while he makes the needed change. Only once does he 
 wait for the double swing, and that is when he is satisfied that 
 the weighing is complete. In this way great rapidity can be 
 obtained in weighing. 
 
 Fig. 24 shows a pulp-balance, sensitive to a single milligram, 
 which, with a 2.5-in. pan, will take readily an assay-ton of ore. 
 It is used for weighing ore, or the lead buttons of a lead assay. 
 Price $15. 
 
 FIG. 24. PULP BALANCE WITHOUT GLASS CASE. 
 
 Fig. 25 shows a pulp-balance, enclosed in a case to protect it 
 from air-currents, and sensitive to 0.25 mg. It can therefore be 
 used, not only for weighing pulp and lead buttons, but even for 
 
AUSTIN'S FIRE ASSAY. 33 
 
 chemical determinations, 1 doing away with the necessity of pur- 
 chasing a chemical balance where one wishes to economize in 
 first cost of equipment. Price $30. 
 
 FIG. 25. PULP BALANCE IN A GLASS CASE. 
 
 Fig. 26 represents a so-called button-balance for weighing the 
 small silver buttons obtained in assaying. It has a 6-in. beam, 
 is sensitive to 1-50 mg., the beam being graduated to 5 mg. for 
 a rider of that weight. In the drawers may be kept, on one side 
 the pliers and button-brush, on the other, sheet-lead, copper, and 
 silver- foil for the assay of fine silver bullion. The weights (500 
 mg. down) are conveniently kept on a neatly marked card-board 
 or on a wooden strip, the weights being returned to their com- 
 partments at once from the pan. Never get into the way of 
 displacing them. In handling the weights, the fingers must rest 
 upon the case, and in moving the rider, the fingers must rest 
 against the case to prevent sudden movement. Otherwise the 
 weights are speedily bent and injured, and the rider may be 
 thrown off the beam. Never leave the sliding door of the case 
 open. It should be closed before you begin to use the rider. The 
 balance can be leveled by means of adjusting screws, watching 
 the bubble-tubes in the case. The final adjustment is made by 
 
34 
 
 AUSTINS FIRE ASSAY. 
 
 means of a star-wheel or nut at the middle of the beam. It should 
 be set so that the vibrations are equal on either side of the zero 
 mark. Carry inside the balance a camel's-hair brush with which 
 to brush off the pans before making the final adjustment. Some- 
 times, when the beam seems out of adjustment, it may be no more 
 than a speck of dust in one of the pans, or even on the beam, 
 itself. Price $110. 
 
 FIG. 26. BUTTON BALANCE. 
 
 Fig. 27 represents an unusually fine balance for weighing the 
 small particles of gold which are left upon parting the button 
 of a silver-gold ore. It is sensitive to 1-200 mg. and has a 
 4-in. beam graduated for a i-mg. rider. The knife-edges are of 
 agate. For ease in reading the rider-divisions, pins project on 
 the beam to direct the eye to the divisions. A rider may be 
 carried on either side of the beam, though, since it can travel 
 half the length of the beam, and read from the middle to the 
 
AUSTIN S FIRE ASSAY. 
 
 35 
 
 end, such an addition is hardly needed. The pointer divisions 
 are viewed through a reading-glass. When the beam is dropped 
 in weighing there is no kick or sudden movement of the pointer, 
 and its direction of movement at once indicates the required ad- 
 justment. The beam is light and only 4 in. long, consequently the 
 vibrations are rapid, and weighing is quickly concluded. Such a 
 balance must be carefully shielded from irregular heating, such 
 as the rays of the sun, and even the incandescent globe should 
 be suspended centrally above it. The assayer must not omit test- 
 ing to note whether the balance is in perfect adjustment before 
 weighing. It will help him in the matter of speed if he also 
 knows the value (in tenths of a milligram) of a vibration of the 
 pointer on the divisions of the scale. Price $200. 
 
 FIG. 27. FINE GOLD BALANCE. 
 
 WEIGHTS. The weights used in assaying are gram-weights 
 and assay-ton weights. 
 
 For use at the pulp-scale, for weighing the buttons from the 
 lead assay, a set of weights from 50 gm. to 10 mg., as shown in 
 Fig. 28, is sufficient. 
 
 For use at the button or gold-balances, a set of weights from. 
 500 to i mg., accurately made, is sufficient. 
 
36 AUSTIN'S FIRE ASSAY. 
 
 Assay-ton weights, from one assay-ton to one-twentieth of an 
 assay-ton, are used for weighing out the powdered ore. The 
 assay-ton equals 29.166 gram, and contains, consequently, as 
 many milligrams as there are Troy ounces (by which silver and 
 
 FIG. 28. SET OF GRAM WEIGHTS. 
 
 gold are weighed) in a ton avoirdupois of 2000 Ib. Therefore, 
 if one assay-ton (A. T.) of ore assays one milligram, the ton 
 contains one ounce Troy. The value is thus obtained : A pound 
 avoirdupois weighs 7000 grains and the Troy ounce 480 grains, 
 
 therefore we have ^ - = 14.58 Troy ounces to the pound avoir- 
 400 
 
 dupois, or 2000 by 14.58 = 29,166 ounces per ton of 2000 pounds. 
 In practice, the weights of one gram and above are cylindrical 
 and are handled with the fingers, those less than one gram are 
 flat, and are never touched by the fingers but handled with 
 pincers. Particular care must be taken of the finer weights, and, 
 in placing them, the hand or fingers rest on the case. Where a 
 number of weights are to be removed the pan is taken off with 
 pincers and they are dumped from it upon the floor of the case 
 and there counted. When returned to the box they are counted 
 again as a check on the first reading. In weighing, proceed 
 systematically. Thus if the weight is 378 mg., we would put on 
 500 mg., too much, then 200, too little, 200 more too much, 100 
 too little, 50 too little, 20 too little, 20 too much, 10 too little, 5 too 
 little. The remainder we know to be between 375 and 380, which 
 
AUSTIN S FIRE ASSAY. 
 
 37 
 
 is to be determined by the rider which will weigh five milligrams 
 on the beam. 
 
 The pincers or forceps for these weights are of brass with fine 
 curved points, see Fig. 29, or nickel-plated as in Fig. 30. They 
 
 FIG. 29. PINCERS. 
 
 FIG. 30. NICKEL-PLATED PINCERS. 
 
 are used, not only for handling the weights, but also for pickmg 
 up buttons, for granulations, or other fragments. In preparing 
 the silver button for assay, it is removed from the cupel by means 
 of pointed-nose button-pliers (Fig. 31), and the adherent bone- 
 ash removed with the button brush (Fig. 32). For larger 
 buttons, of over 100 mg., stouter pliers are to be preferred, since 
 
 FIG. 31. BUTTON PLIERS. 
 
 FIG. 32. BUTTON BRUSH. 
 
 the button must be gripped so tightly as to slightly deform it. 
 Small buttons may be removed from the cupel with pincers, and 
 placed on a smooth-faced steel block (Fig. 33) to be flattened 
 with a light hammer (Fig. 34) as in blow-pipe work. The block 
 is placed close to the pan in the balance and the button transferred 
 
AUSTINS FIRE ASSAY. 
 
 from it directly to the pan. In this way one may lessen the chance 
 of accidentally losing the button. 
 
 FIG. 33. STEEL ANVIL BLOCK. 
 
 FIG. 34. BLOW- PIPE HAMMER. 
 
 For handling the lead button, when hammering it into shape 
 and removing the attached slag, plain steel forceps are used, about 
 6 in. long. The hammering of the button should be upon a 
 smooth-faced steel anvil set into a block (Fig. 35) weighing 4 to 
 6 lb., and with a smooth-faced hammer (Fig. 36) reserved for 
 that purpose only, and weighing at least one pound. 
 
 FIG. 35. STEEL ANVIL TO SET IN WOODEN BLOCK. 
 
 FIG. 36. SLAGGING HAMMER. 
 
 SPATULA. For weighing out and mixing ore, a spatula (Fig. 
 37) with a 6-in. blade serves very well. For weighing, an ordi- 
 nary tinned iron teaspoon is even better, being cleaner and 
 quicker. Indeed there are assayers who use this also for meas- 
 uring out their fluxes. 
 
AUSTIN S FIRE ASSAY. 
 
 39 
 
 MOISTURE SCALES. For determining the percentage of mois- 
 ture in an ore, or the weight of fines and metallics in a sample, 
 the moisture scale (Fig. 38) is well suited. It is finished with a 
 
 FIG. 37. SPATULA. 
 
 sliding weight, and weighs to I kg. A scoop can be placed on one 
 pan and its counterpoise on the other, while extra weights beyond 
 i kg. can be added. 
 
 FIG. 38. MOISTURE SCALES. 
 
VIII. FLUXES USED IN ASSAYING. 
 
 At the temperature of the assay furnace most ores are infusible, 
 and in order to make them fusible, and to bring them into a 
 uniform liquid mass in which the precious metals may come in 
 contact with the collecting agent (molten lead) and be taken up 
 by it, fluxes must be. added. These fluxes, where basic, as in 
 the case with soda, will form a fusible slag with silica as a 
 silicate of soda ; where acid, they form a fusible slag with the 
 bases, as when borax is added. In addition, some carbona- 
 ceous substance, as argols or flour, will reduce litharge to 
 metallic form. 
 
 We may divide fluxes into : 
 
 ACID FLUXES, as borax, glass, or sand. 
 
 BASIC FLUXES, as soda, pearlash, and litharge. 
 
 COLLECTING AGENTS, as litharge or test lead. 
 
 REDUCING AGENTS, as argols, flour, or coke dust. 
 
 OXIDIZING AGENTS, as nitre. 
 
 COVER, as salt. 
 
 BORAX. The ordinary crude borax of commerce is the pris- 
 matic neutral variety (Na 2 B 4 O 7 -(- 10 H 2 O) and when pure con- 
 sists of: 
 
 Boracic acid (B 2 O e ) 36.5% 
 
 Soda (Na 2 O) 16.4% 
 
 Water of crystallization 47.1 % 
 
 100.0% 
 
 Borax-glass may be made by melting this crude borax in 
 a crucible, or in an iron dish or ladle, taking care to fill the ves- 
 sel only partly, since the hydrated salt swells to twice or 
 more of its original bulk. Under the action of heat the water 
 of composition is driven off, leaving white and opaque crystals, 
 
AUSTIN S FIRE ASSAY. 4! 
 
 which melt down into a glass. More borax is added as the 
 former portions melt down until the vessel is two-thirds full. 
 The molten mass is now poured into a mold, or preferably 
 upon an iron plate. The discoloration, due to iron, where an 
 iron dish has been used, is a matter of indifference, the amount 
 so dissolved being trifling. When cold, the borax-glass is 
 broken in an iron mortar, and then ground upon the bucking 
 plate until it will pass a flour-screen of i6-mesh. The assayer 
 generally gets it already prepared for use. 
 
 For crucible assays, crude borax may be used. Placed upon 
 the top of the charge, it tends to keep it cool to the last, and until 
 the lower portion of the charge is melted, and tends to prevent 
 the mechanical loss due to rapid escape of the gases. There is no 
 danger to be feared from not mixing it with the rest of the charge, 
 the melting proceeding quite as well. Where borax-glass is to be 
 used it may be (and often is) mixed with the other fluxes and 
 the ore. Borax-glass is most appropriately used in scorifica- 
 tion, where it is placed upon the top of the charge in the 
 scorifier. 
 
 Borax serves as an acid flux to unite with bases contained 
 in the ore, and the quantity to be used depends upon the 
 amount of bases present, as may be estimated either from the 
 appearance of the ore, or better, by actual determination. The 
 common bases are the oxides of the metals as, FeO, ZnO, 
 MnO, CuO, and the alkaline earths CaO, MgO, and BaO. To 
 this may be added A1 2 O 3 , which is, however, such a feeble 
 base that it is quite commonly disregarded. The alkalies and 
 other oxides, not above enumerated, occur to so small an 
 extent that they are not taken into account. Alkalies occur- 
 ring in feldspar, would, in a crucible assay, cut very little 
 figure compared with the soda added as a flux. Where the 
 amount of bases can be determined, it will be found, that for 
 each part of base, we should use crude borax, four parts, and 
 borax-glass, two parts. W T hen so used it makes a fluid easy- 
 pouring slag, and takes up and dissolves the oxidized bases. 
 Such a slag is glassy and brittle. A deficiency of borax is 
 indicated by the slag pouring too stiff or thick, but brittle. 
 
42 AUSTINS FIRE ASSAY. 
 
 Sometimes it will not leave the crucible, or else the pour is 
 lumpy. However, a slag lacking soda may also act in this 
 way, except that the silicious slag will string out and the 
 trouble is to be laid to a shortage in borax, where we know 
 that the amount of soda is right. Where a salt cover is used, 
 and crude borax is employed, the latter is placed above the 
 cover. It intumesces or swells up, and here, when much has 
 to be added, is disposed to overflow before it again fuses and 
 recedes into the crucible. If it appears disposed to thus go 
 beyond bounds it may be pushed back into the crucible in 
 such a way as to melt down. A small loss of borax is a matter 
 of little moment. Crude borax loses half its weight by fusion 
 to borax-glass, and hence twice the weight is needed to per- 
 form the same work as the borax-glass, as has already been 
 indicated. 
 
 GLASS. Glass is a fusible silicate which must be free from lead 
 when used for a lead assay. t It contains 65 to 73% silica, and, 
 being fusible (having been fused in the making) is preferred by 
 many to sand for furnishing silica to a charge. Glass (besides 
 silica) contains in general 20% of soda (Na 2 O) or of potash 
 (K 2 O) together with lime (CaO). Lime, when present, is thought 
 to give a cleaner separation of the button, and hence would be 
 an advantage in assaying. Glass is pulverized to pass through 
 a flour-screen. About one-third of the silica in glass is available 
 as silica, the rest having been satisfied by the alkaline bases. 
 Hence, in using it, three grams should be taken as equivalent to 
 one gram of silica or sand. 
 
 SAND. A clean white sand, ground and sifted through a 60- 
 mesh sieve, will serve to furnish silica to an assay-charge when 
 needed. Both sand and glass protect the crucible from being 
 eaten out under the action of basic fluxes (soda and litharge). 
 We' may reckon the whole weight of a quartz sand as available 
 silica, though it is a pure sand which contains more than 95% 
 SiO 2 . 
 
 Either glass or sand is omitted from the regular assay, since it 
 is found that it will disturb the calculation for reducing power, 
 giving a button of a different size. 
 
43 
 
 SODA. While bicarbonate of soda (NaHCO 3 ) and even the 
 ordinary soda crystals have been used in assaying, the best for 
 such use is the mono-carbonate made by the ammonia process, 
 98% pure and of the composition Na 2 CO 3 . When chemically 
 pure it contains 58.5% caustic soda (Na 2 O). It forms with silica 
 at a red heat, a fusible silicate thus 
 
 Na 2 CO 3 + SiO 2 = Na 2 SiO 8 + CO 2 
 
 the escaping CO 2 gas thoroughly agitating and mixing the melt- 
 ing contents of the crucible. From the formula we find by 
 calculation that one part of silica needs 1.76 part of soda to form 
 the mono-silicate. 
 
 Pearlash, or commercial potassium carbonate, contains 90% 
 K 2 CO 3 together with some sodium carbonate, potassium sulphate, 
 and sodic chloride; so that we have about 60% of caustic potash 
 (K 2 O) available to combine with silica. When the salt is mixed 
 with the sodium mono-carbonate we get an alkaline flux which is 
 more fusible than either of the carbonates taken separately, and 
 this is the principal advantage which may be expected from 
 such a mixture. The disadvantage of using pearlash is the 
 introduction of sulphates into the charge, so that many assayers 
 use sodium mono-carbonate only. Pearlash is disposed to absorb 
 moisture from the air and become deliquescent, or at any rate, 
 lumpy. When used with soda it is mixed in equal parts, or better, 
 in the ratio of 2.7 parts of soda to 2.3 parts of pearlash. Those 
 who use it claim it gives more satisfactory results than soda 
 only, and habitually use it in mixture. 
 
 Litharge (PbO) is not only a basic flux, but that part which 
 is reduced to lead is also a collector of gold and silver. It has 
 the composition 92.8% Pb and 7.2% O. Ordinary commercial 
 litharge contains a small amount of silver, amounting to 0.75 to 
 1.25 oz. per ton. If an assayer is compelled to use this, he should, 
 after mixing the whole lot together to make it uniform, determine 
 the amount of silver present, and deduct this from the assay 
 according to the amount of litharge used. To make such a deter- 
 mination take : 30 gm. litharge, 2>^ gm. argols, and 5 gm. sand. 
 Bring the charge to quiet fusion, which will take 25 minutes. 
 The lead button, weighing 20 gm., is cupelled. 
 
44 AUSTIN S FIRE ASSAY. 
 
 There is now no occasion to use ordinary litharge, since pure 
 litharge is readily obtainable from dealers in assay supplies. 
 
 When unreduced, and entering the slag, litharge is a basic 
 flux, acting upon the silica thus, PbO + SiO 2 = Pb SiO 3 ; but 
 also exerting a solvent action upon the sulphides present. A 
 portion of the litharge, in an ordinary crucible assay, is reduced 
 for the purpose of affording a lead button, into which are col- 
 lected the gold and silver of the ore. Before its reduction, the 
 litharge has exerted its oxidizing effect, and the impurities, thus 
 oxidized, are ready for solution in that which has remained 
 unreduced. As prescribed in the regular assay, the amount of 
 litharge is largely increased where the impurities arsenic, anti- 
 mony, copper, or tellurium are large in amount. The reducing 
 power of various sulphides, or the amount of litharge needed 
 to oxidize them, is, according to Berthier, as follows : 
 
 Manganese sulphide, 28 parts lead or 30 parts litharge. 
 Iron sulphide (FeS) 28 " " " 30 " 
 Ironpyrite (FeS 2 ).. 46.5" * " 5O " 
 Copper sulphide . . . 23 " " " 25 " 
 
 Copper pyrite 32.5 " " " 35 " 
 
 Antimony sulphide . 23 " " " 25 " 
 
 Zinc sulphide 23 " " "25 " 
 
 Arsenic sulphide . . . 46.5 " " " 50 " 
 
 Lead sulphide 2 " " " 2 " 
 
 Red lead (Pb 3 O 4 ) may be used in assaying in place of litharge. 
 It contains 90.5% Pb, and has, because of its greater percentage 
 of oxygen, a slightly greater oxidizing power upon sulphides. 
 
 REDUCING AGENTS. The reducing action of these compounds, 
 such as charcoal, flour, argols, as well as sugar and starch, is 
 due to their contained carbon acting thus, 
 
 PbO + C = Pb + CO (or 2 PbO +" C = 2 Pb + CO 2 ) 
 the carbon monoxide escaping as a gas. 
 
 CHARCOAL. Well-charred wood-charcoal is brought to a red 
 heat in the muffle; it is then withdrawn, the ash blown off, and 
 the pieces broken small in an iron mortar. The pulverizing is 
 finished upon the grinding plate to pass a 4O-mesh sieve. One 
 part of charcoal will reduce 28 parts of lead from litharge and 
 
AUSTIN S FIRE ASSAY. 45 
 
 it is one of the most powerful reducing agents known. Its dis- 
 advantage is its lightness, whereby it tends to float on the surface 
 of the charge into which it has been mixed. 
 
 FLOUR. The flour of wheat or of rye may be used. Its fineness 
 enables it to be mixed intimately with the charge, and it is pre- 
 ferred by many to argols. One part reduces about 12 parts of 
 lead from litharge. 
 
 ARGOLS. Argols is used like flour, being mixed into the charge. 
 One part of it will reduce about eight parts of lead from litharge. 
 When pure it has the formula KHC 4 H 4 O 6 . Crude argols con- 
 tains, besides bitartrate of potash, tartrate of lime, and the 
 amount used in an assay is so small that its fluxing power is not 
 regarded. It is, however, a basic flux. It has an advantage that 
 its color prevents mistakes in using. 
 
 REDUCING-FLUX MIXTURES. Reducing-flux mixtures consist 
 of a mixture of any of the above reducing-fluxes with soda, 
 pearlash, or the two together. The advantage of using such a 
 mixture over that of the ingredients separately, is that because 
 of its dilution, a measured quantity will insure an exact reduction. 
 Two common reducing-flux mixtures are 
 
 80 parts soda to 20 parts argols, and 
 85 parts soda to 15 parts flour. 
 
 The mixture should be intimate and uniform and the determi- 
 nation of the reducing power of a measureful should be made. 
 Pearlash may be substituted in part for soda, or the alkaline car- 
 bonates may be used in ratio of 2.7 of soda to 2.3 of pearlash 
 (in ratio of their atomic weights). 
 
 TABLE OF REDUCING POWERS. One part of each of the follow- 
 ing reducing agents reduces as indicated : 
 
 Charcoal -. 28 parts lead. 
 
 Wheat flour 12 
 
 Argols 8 
 
 Reducing-flux; 80 soda, 20 argols 1.6 " " 
 
 Reducing-flux ; 85 soda, 15 flour 1.8 " 
 
 This reducing power varies, and in any case it is best to deter- 
 mine it when a fresh lot has been prepared as follows : 
 
46 AUSTIN'S FIRE ASSAY. 
 
 Mix in a new crucible 3 gm. reducing-flux, 40 gm. PbO, and 
 10 gm. soda. Melt quickly and pour, just before the time of 
 quiet fusion, into a conical-hole mold and weigh the resultant 
 button. These, and in fact any fluxes, should be passed through 
 a flour-sieve. Where a reducing mixture is made, mixing should 
 be so thorough as to make all piarts appear quite uniform in 
 color. 
 
 OXIDIZING AGENTS. The nitrates of potash (nitre) and of 
 soda, especially the former, are used for oxidizing the sulphides 
 of the charge, where the ore itself would reduce too large a lead 
 button. Their oxidizing power is such that one part will reduce 
 from, litharge: 
 
 Nitrate of soda (NaNO 3 ) 4.9 parts of lead. 
 Nitrate of potash (KNU 8 ) 4.2 parts of lead. 
 
 This is for the pure salt. We reckon the oxidizing power of 
 commercial nitre at 4 even. Neither of them contains water of 
 composition. 
 
 To determine this oxidizing power prepare two crucibles as 
 follows : Charge into each 40 gm. PbO, 10 gm. soda, and 3 gm. 
 argols, and to one of them add 3 gm. nitre. The crucibles are 
 melted quickly and without loss, poured, and the resultant buttons 
 weighed. The difference between the weights, divided by 3, 
 will express the oxidizing power of the salt. As a result of the 
 oxidizing action of the nitrates, their acid portion is removed, 
 leaving the alkaline base. 
 
 SALT. Salt is used as a cover to the charge in assaying, and 
 for this purpose, table salt or any ground salt is well suited. It 
 cannot be regarded as a flux, and when melted, floats as a trans- 
 parent fluid upon the charge. It then slowly volatilizes, giving a 
 white fume. Where the charge is kept for a long time in the 
 fire, it may be quite driven off. It will be found as a white layer 
 upon the top of the poured slag. It acts as a wash, carrying 
 down any particles which cling to the sides of the crucible. The 
 commercial article contains some moisture, which decrepitates 
 on heating. This may be removed by moderately heating the 
 salt in an iron dish before using, where mechanical loss, due to 
 decrepitation, is feared. 
 
AUSTIN S FIRE ASSAY. 47 
 
 Many assayers have given up the use of salt, claiming that 
 equally good results are attainable without it, and that it increases 
 the chance of loss, especially where much nitre is used. Where 
 salt is omitted, dependence is placed upon borax as a cover. 
 
 EFFECT OF EXCESSIVE FLUXING. It is desirable to use no more 
 fluxes than necessary, since as the quantity of slag is increased, 
 so is the silver and lead therein, and no slag is entirely clean. 
 Where the fluxes are in excess the slag is fluid and thin, and is not 
 as clean as the normal slag, especially in the lead assay. 
 
IX. ORES. 
 
 In order to determine what fluxes to use, the assayer must be 
 acquainted with ores from their physical appearance. 
 
 They may be divided into oxidized and sulphide ores. 
 
 Oxidized ores, among which may be classed the carbonates, 
 have a dull earthy look, are whitish or yellowish in color, and, 
 when containing iron, brown and red. The depth of color is 
 intensified where manganese is present, since manganese oxides 
 tend to blacken the ore. At the same time, ores, apparently oxi- 
 dized, may have some sulphides mixed with them and, on close 
 inspection, the particles of zinc, iron, or lead sulphides may be 
 detected. A preliminary determination for reducing power also 
 indicates the same fact. Where there -are shining particles of 
 mica, do not confound them with the sulphides. Ores containing 
 sulphur, as sulphate, are not, however, reducing in their action. 
 
 Sulphide ores have a semi-metallic appearance as shown in the 
 case of galena, blende, and pyrite. Let us take the case of the 
 three common sulphides, we have : 
 
 Galena containing 87% lead, 13% sulphur. 
 
 Blende containing 67% zinc, 33% sulphur. 
 
 Pyrite containing 46% iron, 54% sulphur. 
 It will be noticed that the percentage of metal (or base) de- 
 creases while the sulphur increases in this list. Since the lead 
 in the galena is reduced to metal in assaying, it does not enter the 
 slag, and, if so, from the point of view of the assayer, is not to 
 be regarded as a base. Galena contains but little sulphur as com- 
 pared with the other sulphides. Of the remaining sulphides 
 (blende and pyrite), we may say that they contain much base, 
 and also much sulphur. 
 
 Ores may be divided into silicious ores (having a gritty feel) 
 and non-silicious ores, or quartz at one extreme, and galena or 
 limestone at the other. Generally the non-silicious ores are also 
 
AUSTIN'S FIRE ASSAY. 49 
 
 basic, as lime rock or iron ore, but galena or lead carbonate is 
 norf-silicious but not basic for the reason above given. 
 
 Again we may divide ores into silicious and basic ores, since 
 the one quality is just the opposite of the other. An example of 
 an oxidized basic ore is iron ore, and, among the sulphides, blende 
 and pyrite. 
 
 As regards ores high in lead (leady ores) they are low both in 
 silica and bases. Such ores feel heavy, and, when a portion is 
 weighed out for assay, it occupies but little bulk. Thus the 
 assayer is warned of its nature, and, in case of a lead carbonate, 
 he can judge of the irony base it contains by the color. When, 
 however, the sample has been dried at a rather high temperature 
 this red color is intensified, and this should be taken into account. 
 
 In theory, the ore should be dried at a temperature not to 
 exceed 100 C, since at higher temperatures sulphur on carbon 
 dioxide may be driven off and the sample is in an unstable con- 
 dition. 
 
X. THE SCORIFICATION ASSAY. 
 
 This is the simplest method of assaying, and is suited to ail 
 kinds of ores which come to the assayer. It is used for the 
 determination of gold and silver in ores. 
 
 In making a scorification assay the following is the regular 
 charge : 
 
 30 to 60 gm. test-lead, o.i assay-ton of the ore and I to 5 
 gm. borax. 
 
 In a 2^2 -in. scorifier measure one half the specified amount of 
 test-lead, weigh out the ore, and mix with a spatula in the 
 scorifier. Sprinkle on the remainder of the test-lead, completely 
 covering the mixed portion, then add the required amount of 
 borax. The quantity of borax will depend upon the amount of 
 bases in the ore. When there is much base, add much borax, and 
 for little base use little borax. For example, the pyrite or the 
 blende ores, already mentioned, need 4 gm. borax, while a silicious 
 non-basic ore needs but I gm., and that only as a safeguard. 
 
 The scorifier, thus prepared, and marked with red chalk or 
 riddle to distinguish it from other assays to be done at the same 
 time, is placed at the hottest part of the muffle, and the door of 
 the muffle closed. The charge soon melts down, and, when in 
 complete fusion, the door is opened and the scorifier brought 
 toward the front of the muffle to scorify or 'drive.' Here it is 
 exposed to the air, and the molten lead, at that high heat, begins 
 to oxidize, giving off fumes. As the operation proceeds litharge 
 forms, and, together with the borax, slags with the gangue of the 
 ore. The slag accumulates upon the borders of the scorifier, and 
 finally there is left exposed to the action of the air the middle of 
 the bath only. This centre or eye diminishes to, say, ^ in. diam., 
 when this stage is complete. The scorifier is again moved back 
 to the hot portion of the muffle and the door closed. As soon as 
 the slag is melted, the contents of the scorifier is poured into a 
 slag mold. The scorifier, if the work is well performed, should be 
 
AUSTIN S FIRE ASSAY. 51 
 
 quite free from any imdecomposed ore and smooth upon its whole 
 interior surface. 
 
 It will be noticed that the scorification proceeds by three stages : 
 
 1. The ore is melted down (about five minutes). 
 
 2. The muffle is opened and scorification proceeds (about 
 twenty minutes). 
 
 3. The scorifier is again heated (about five minutes) and 
 is ready for pouring. 
 
 To return to the slag left cooling in the mold. When the con- 
 tents of the mold is turned out, there will be found a lead-button, 
 preferably of 10 or 12 gm., at the point of the mold, and above 
 it the adhering slag. This slag is removed by hammering the 
 button as it rests upon an anvil, so that the slag is thoroughly 
 cleaned off and the button is ready for cupelling. It is to be 
 understood that this button has gathered up all the gold and silver 
 which was in the ore. This lead is removed in the operation of 
 cupelling, and there remains behind a bead or button of silver 
 (and jgold). 
 
 CHEMICAL REACTIONS IN SCORIFICATION. In the first stage, 
 as the charge (containing not quite three grams of ore) melts 
 down, air comes in contact with it, tending to oxidize or roast 
 any sulphides present according to the reaction : 
 PbS + 3 O =PbO + SO 2 
 ZnS + 30 =ZnO + SO 2 
 FeS 2 + heat = FeS + S 
 The sulphur, on contact with air, burning to SO 2 . Then : 
 
 FeS + 30 == FeO + SO 2 
 The molten lead is also oxidized to litharge as follows : 
 
 Pb + O = PbO 
 
 But any PbO, thus forming, eagerly takes up any silica in the 
 gangue of the ore. The borax is busy dissolving the oxides as 
 they form thus : 
 
 FeO + Na 2 B 4 7 = Na 2 FeB 4 O 8 
 ZnO + Na 2 BO 4 O 7 = Na 2 ZnB 4 O 8 
 
 By the time the door has been opened, enough PbO has been 
 formed to dissolve the silicious gangue, and if not, the increased 
 access of air at this time speedily produces it. At the period of 
 
AUSTINS FIRE ASSAY. 
 
 driving, oxidation of litharge proceeds, which, reacting upon the 
 partly decomposed sulphides, completes their roasting, thus : 
 
 PbS + 2PbO = 3?b + SO 2 
 
 ZnS + 3PbO = sPb + ZnO + SO, 
 
 FeS + 3?bO = 3?b + FeO + SO," 
 
 CuS + 3?bO = 3?b + CuO + SO~ 2 
 
 As 2 S 3 + QPbO = As 2 O 3 + 3SO 2 + pPb 
 
 That is to say, oxidation may be brought about by contact with 
 
 the forming litharge as well as by the action of the air. As 
 
 oxidation proceeds, the slag which is formed gradually covers 
 
 the surface of the metal, leaving a continually smaller eye at the 
 
 centre exposed to the air. When the eye at the centre becomes 
 
 small, oxidation necessarily proceeds slowly ; so that, to save time, 
 
 it is better to pour the assay. The resultant button should weigh 
 
 preferably 10 to 12 gm., although we often have it of 20 gm. 
 
 weight, and sometimes no more than 5 grams. 
 
 GENERAL REMARKS ON SCORIFICATION. It will be noticed that 
 the quantity of test-lead prescribed for an ordinary assay is 10 
 times the weight of the ore. When, however, the ore is 'dirty' 
 (containing arsenic, antimony, copper, or tellurium) the lead is 
 increased proportionately in accordance with the following table : 
 
 TABLE SHOWING THE QUALITY OF TEST-LEAD NEEDED PER y lfl A. T. OF 
 
 MINERAL. 
 
 MINERAL 
 
 GRAMS 
 TEST LEAD 
 
 REMARKS 
 
 Antimonial ore 
 
 45 
 
 
 Arsenical ore 
 
 45 
 
 Needs high heat in scorifying 
 
 Milling- silver ore 
 
 30 
 
 
 Cupriferous ore 
 
 30 to 60 
 
 
 Zinc-bearing ore 
 
 30 to 45 
 
 Needs high heat and much borax 
 
 Leady ore 
 
 30 
 
 
 Silicious ore 
 
 30 
 
 
 Irony ore 
 
 30 
 
 
 Telluride ore 
 
 30 to 60 
 
 
AUSTIN'S FIRE ASSAY. 53 
 
 In scorifying, the air, as it passes over the contents of the 
 scorifier, oxidizes the contained molten lead to litharge. The lith- 
 arge reacts on the impurities of the ore, oxidizing them, and they 
 enter the slag (formed by the litharge uniting itself to the gangue 
 of the ore and to whatever borax has been added). The ore 
 also tends to float on the surface of the lead, and is thus rapidly 
 roasted in contact with the air. Borax is added to the charge 
 according to the quantity of base present. Much base needs much 
 borax. Thus iron ore, limestone, pyrite, and blende need much 
 borax (especially the last, which tends to form a stiff slag). On 
 the contrary, a silicious ore, a lead carbonate, or galena, needs 
 but little. The base most commonly present is iron, and, in case 
 of oxidized ore, the color gives some idea of its quantity. 
 
XLCUPELLING. 
 
 PREPARATION OF THE CUPEL. Ground bone-ash is mixed in a 
 pan with a little water until there is a coherence of the material 
 when grasped in the hand. This is determined by experience. 
 The bone-ash is placed in the mold, and the plunger brought 
 squarely upon it. A. blow or two of the mallet compresses the 
 bone-ash in the ring. The cupel is ejected and placed upon a 
 board, and the accumulated supply of cupels is placed in a warm 
 place to dry. They can be used after thorough drying. 
 
 OPERATION OF CUPELLING. The cupel is warmed upon some 
 portion of the furnace, is freed from dust by blowing upon it, 
 and is then put into the hot part of the muffle. After remaining 
 there until it has attained the heat of the muffle, the assay-piece 
 is placed in it by means of cupel tongs. These must grip the 
 cupel so as not to injure its top edges. The door of the muffle 
 is closed, and the lead brought to its full heat until it clears. This 
 is due to the melting of film of oxide which covers the assay- 
 piece and which goes to the edge of the cupel leaving the surface 
 exposed to the oxidizing influence of the air. The door is opened 
 and the cupel brought further forward toward the front of the 
 muffle to 'drive.' When the muffle is not so hot as to start this 
 operation readily, it may be done by putting in a small piece of 
 wood the size of the finger, which, laid in front of the cupel on 
 the floor of the muffle, burns, producing a reducing atmosphere, 
 while, at the same time, the door of the muffle is closed. This 
 reduces the lead film already mentioned, so that when the door is 
 opened, and the wood removed, cupellation at once proceeds. The 
 temperature should be controlled between rather narrow limits. 
 If it is permitted to become too low there is a liability of the 
 button 'freezing' or solidifying and the assay is ruined. Even if 
 the work can be pushed to the hottest part of the muffle and the 
 assay started again, results will come out low, so that a button 
 once frozen, especially toward the last of a cupellation, is useless. 
 If the temperature is too high there will be a volatilization loss of 
 silver, and, even when the operation is well performed, this loss 
 
AUSTIN S FIRE ASSAY. 55 
 
 will be as much as 2.5%. The proper temperature can only be 
 learned by experience. The rule should be that the heat must be 
 such that a slight ring of litharge crystals (feather litharge) 
 forms around the sides, and especially at the front interior side of 
 the cupel while the litharge-smoke is seen rising, and the top sur- 
 face of the molten metal shows a slightly darker ring around its 
 borders. As cupellation proceeds, the lead is oxidized and at 
 once mostly absorbed by the bone-ash of the cupel, while a little 
 passes off as a litharge-smoke. The button becomes gradually 
 smaller and, at the expiration of perhaps 20 minutes, begins to 
 round up, while, if the button is large enough, a play of colors 
 may be seen passing over it as the last traces of lead are removed. 
 At this time the temperature should be increased by pushing the 
 cupel back to the hotter portion of the muffle. When the lead is 
 all gone the play of colors ceases, the button brightens or 'blicks' 
 suddenly, and the cupellation is complete. The cupel may be left 
 for a little while longer in the muffle, since there can now be no 
 further loss of silver. It is then removed, placed on a mold or 
 an iron plate, and allowed to cool previous to weighing. If the 
 button is large it is better to cover with a hot cupel before re- 
 moving from the muffle and to remove it gradually, since other- 
 wise it is liable to 'spit' or sprout, which may result in a loss 
 through the projection of microscopic particles of silver from 
 the substance of the button, which may be observed under the 
 microscope at the borders of the cavity of the cupel. The reaction, 
 Pb + O = PbO = 51,000 calories, indicates that much heat is 
 evolved as the result of the burning of the lead to litharge. 
 
 The following table shows the losses which occur in cupelling 
 a button of 10 gm. at different temperatures and the importance 
 of performing this operation at the correct temperature : 
 
 TEMPERATURE 
 DEGREES C 
 
 Loss 
 
 % 
 
 REMARKS 
 
 700 
 
 1.05 
 
 Feather-litharge about button 
 
 775 
 
 1.18 
 
 Feather-litharge on cooler side 
 
 850 
 
 1.70 
 
 No more crvstals 
 
 925 
 
 2.59 
 
 
 
 1000 
 
 4.78 
 
 <t 
 
XII. PARTING. 
 
 The silver-gold button or bead, resulting from the cupellation, 
 has now to be weighed upon the assay balance to the nearest 
 tenth of a milligram, and the result at once written in the note 
 book in ounces. (The weight in milligrams is not to be written 
 in.) The button is taken from the cupel, being gripped strongly 
 in the pliers (Fig. 32), and its lower surface brushed clean from 
 bone-ash with a button-brush. The button is then placed on the 
 table, the forceps seize it at right angles to the former position, 
 and it is again brushed. Squeezing the button hard in the pliers 
 should deform it somewhat, and thus tend to loosen any particles 
 of bone-ash sticking to it.* The button may now be flattened 
 with a hammer, after which it is ready for parting. 
 
 PARTING IN A PORCELAIN CAPSULE. A Royal-Berlin glazed 
 porcelain capsule of i^ in. diam. and 1^/4 in. high, is preferably 
 used in parting the button. Upon this is poured the pure dilute 
 nitric acid (i of acid of sp. gr. 1.42 to 3 of water), covering the 
 button to say, J4 in. depth. This is brought to the boiling point 
 on the hot plate, taking care the ebullition is not violent, since 
 such action tears apart the newly released gold. When action 
 ceases, manipulate the particles of gold so as to bring them 
 together. This can be effected with the aid of a glass rod, stirring 
 the liquid in vortex fashion, and tapping the outside of the 
 capsule to work the particles together. If a small particle of gold 
 shows itself, floating on the surface of the liquid, it may be made 
 to sink by touching it with the glass rod. It is always necessary 
 to watch that no such particles are overlooked. , The dilute 
 acid is poured off into a jar or wide-mouthed bottle, leaving the 
 gold behind. A few drops of stronger acid (7 of acid to 4 of 
 water) are then added to the capsule, brought to boiling, until 
 *Let the beginner weigh the button without completely cleaning it, 
 and then, after proper cleaning, weigh again. The difference in weight 
 will be observable. 
 
AUSTIN S FIRE ASSAY. 
 
 57 
 
 action of the gold has ceased, and then decanted. The capsule 
 is now nearly filled with hot distilled water from a wash-bottle, 
 directing the stream tangentially on the border of the capsule 
 so as to disturb the gold as little as possible. This wash water 
 is now poured off and the operation repeated. 
 
 Where distilled water cannot be obtained, the common water 
 will form a precipitate of silver chloride, and, when this happens 
 with the first addition of water, a few drops of ammonia, also 
 added, will dissolve the chloride and clear the solution. Where 
 common water has to be used to dilute the nitric acid, a small 
 button of silver placed in the acid diluted with it will precipitate 
 all chlorides present. This precipitate is filtered off before using 
 the dilute acid. 
 
 The capsule is dried on a hot-plate, and ignited at a visible-red 
 heat in the muffle. It is then removed, cooled by setting on a 
 cold iron surface, and is ready for weighing. In weighing the 
 particle of gold, first see that the balance is exactly adjusted. 
 Wipe any particles of bone-ash from the exterior of the capsule, 
 loosen the gold in it with the point of the pincers, and tap the 
 capsule with the pincers so as to cause the particle to enter the 
 pan, which latter has been removed and placed on the clean floor 
 of the balance in readiness to do this. Weighing the gold is done 
 to the one-hundredth of a milligram. A camel's-hair pencil brush 
 may sometimes be used to advantage for removing final traces 
 of gold. The results are registered thus: 
 
 1900 
 
 Jan. 
 
 Ontario No. 27 (Irony ore.) 
 Ag and Au 15.8 
 
 Au 3.10 
 
 Ag 
 Pb 17.2 
 
 3.10 
 12.7 
 
 PARTING WITH A FLASK. The button is dropped into a part- 
 ing-flask, preferably one of i-oz. capacity (See Fig. 39), which 
 
58 AUSTIN'S FIRE ASSAY. 
 
 can stand on the hot-plate. Acid in moderate quantity, say to 
 one-quarter inch in depth, is added, and the acid brought to the 
 boiling point on the hot-plate. The action should be gentle, to 
 avoid mechanical loss, and should go on until the nitrous acid 
 fumes are driven from the solution. The acid is poured off, 
 
 FIG. 39. PARTING-FLASK. 
 
 and some of the stronger acid added in order to remove the last 
 traces of silver. This is in turn brought to the boiling point and 
 then poured off. Note, that in all cases the assayer is to see that 
 the acid has completed its work before pouring it from the flask. 
 Turn the flask so that the gold adheres to the upper side of it, 
 and carefully pour in hot water, or add it from the wash-bottle, 
 to half fill the flask. Turn back the flask to bring the gold be- 
 neath the water. Be careful not to break up the gold by violent 
 action of the water. This wash-water having been removed, the 
 flask is completely filled, a dry-cup of white clay, ij4 m - diam., 
 is inverted on the flask and the whole suddenly reversed. The 
 gold is seen to fall through the water to the bottom of the cup. 
 Cant it a little to one side, and quickly and quietly withdraw the 
 flask, letting the water run out of it into a convenient jar. The 
 gold in the cup is tapped together, if necessary, and the water 
 poured from it. This work should be done promptly, or other- 
 wise the dry-cup becomes soaked with water, and, when pouring 
 off, the water will not be absorbed from the gold. The cup is 
 dried, and ignited at a low-red heat, and the gold, now showing 
 its characteristic yellow color, is weighed. 
 
 Any carelessness in parting may result in a loss of particles 
 of gold giving a low result. Or again particles of dust may get 
 
AUSTIN'S FIRE ASSAY. 59 
 
 into the parting capsule and if weighed up with the gold would 
 give a high result. Where there are small amounts of gold, as 
 two or three hundredths of a milligram, the assayer should ad- 
 just his balance just prior to weighing. He should never report 
 small particles as a 'trace' until, after weighing, he has found 
 that there is actually no change in weight. 
 
 Where the amount of gold in the button is high compared with 
 silver, care must be taken that the silver is completely dissolved, 
 using the stronger acid after the weaker, otherwise results will 
 be reported too high. If you find upon careful treatment that not 
 less than one-fourth of the original button is gold, you may be 
 pretty sure that you have not dissolved out all the silver. In 
 such a case it is necessary to add to this button, at least three 
 times its weight of silver, wrapped in a small piece of lead-foil, 
 and cupel off the lead in a small cupel. Or the same operation 
 may be performed on charcoal, using but little lead-foil. The 
 remaining button can then be parted properly and completely. 
 
 Right here the young assayer must be cautioned not to commit 
 the blunder of reporting as gold a residue which may contain 
 silver. No one will pay $20 per ounce for that which is worth 
 but 7oc, and an assayer who reports in this way is liable (and 
 rightly) to lose his job. 
 
XIII. THE CRUCIBLE ASSAY. 
 
 While any ore may be assayed by scorification, the amount in 
 which the determination is made is but one-tenth of an assay-ton, 
 and consequently, whatever the error is in weighing, it is multi- 
 plied by ten in reporting the ounces per ton. Any variation in 
 the sample is also exaggerated in the same way. In the crucible 
 assay, however, one-half an assay-ton is commonly used, and 
 results become more accurate. The difficulty may also be over- 
 come by making several scorifications and uniting the silver 
 buttons into one for the determination of the gold. This implies, 
 however, a good deal of extra work, and the same result is 
 attained by a single crucible assay. 
 
 Assays by crucible are made by one of two methods. The 
 first is called the 'nail assay,' adapted to the treatment of clean 
 ores, and the second, the 'regular assay,' also called litharge 
 assay, intended for dirty ores, which carry impurities to be got 
 rid of, such as copper, arsenic, antimony, or tellurium. 
 
 On the Pacific Coast as much as two to four assay-tons of ore 
 are fused in large crucibles in a coke-furnace or in a crucible- 
 furnace intended for melting purposes only. The method has 
 the advantage that the determination is made in a large portion 
 of ore, and consequently, on low-grade ores, with much accuracy. 
 
 THE NAIL ASSAY (for clean ores). The charge is as follows: 
 25 gm. soda, 20 gm. litharge, 3 or 4 gm. argols, 5 to 15 gm. 
 borax-glass, I to 5 nails, and 0.5 assay-ton of the ore. 
 
 The various amounts are added by measure, there being a 
 small cylindrical tin cup provided for each of the fluxes of the 
 following dimensions: 
 
 For soda, i^ in. diam. by 1^4 in. high. 
 
 " litharge, # " " " % " " 
 
 " argols, Y 4 " " " 7/s '" " 
 
 " borax-glass, I " " " i# " " 
 
 " test-lead, # " " " # " " 
 
AUSTIN'S FIRE ASSAY. 61 
 
 These measures can be made by any tinner. They must all be 
 tried for their capacity when level full. In preparing the charge, 
 the soda, litharge, and argols are measured at once into a 2O-gm. 
 crucible. The ore is then weighed out accurately upon the pulp- 
 scales, and poured on top of these fluxes. A spatula with a 6-in. 
 blade is now used to mix the contents of the crucible until they 
 appear of a uniform color. The side of the crucible is tapped 
 with the handle of the spatula to settle its contents, the borax is 
 placed on top as a cover, and the nails are thrust into the charge, 
 points downward. 
 
 The crucibles, thus prepared, may be placed at once in the hot 
 muffle, though it is well to place them in the muffle when cold and 
 allow the heat to come up gradually. Where the crucible is to be 
 placed in the coke, fresh coke is added to the fire before so doing, 
 so that the top part of the crucible is cold. This diminishes the 
 chance of boiling over and consequent loss. The crucible remains 
 in the fire until its contents are in tranquil fusion, the time vary- 
 ing, in general, from twenty minutes to half an hour. Sometimes, 
 however, the time is greatly exceeded, as when an occasional 
 bubble is seen rising through the molten charge. In any case 
 the charge should be hot, well fused, and liquid, and should pour 
 smoothly, without lumps and yet not be too thin or easily flowing. 
 After pouring, the assayer should look into the crucible to see 
 that it is perfectly smooth and with no unfused matter sticking 
 to its sides. Sometimes scale from the nails, or a piece of one of 
 them, may adhere to the crucible, but this can be scraped out with 
 the cupel tongs. Upon removal of the crucible from the furnace, 
 hold it by the tongs in the left hand, and, with the right, use a pair 
 of short tongs or forceps (Fig. 4) to take out the nails. Now, 
 still holding the crucible, tap it upon the table or plate where you 
 are doing the pouring, and then neatly pour into a conical mold 
 (Fig. 10). After pouring, do not disturb the mold for five 
 minutes or more, or until the contents have cooled and set. If 
 you have to move the mold, keep it perfectly level. Otherwise 
 you have the lead forming a fin at the side of the mold. When 
 set, the cone of slag and the button are slid out on the grinding 
 plate, or on an anvil, the button is detached from the slag and 
 
62 AUSTIN'S FIRE ASSAY. 
 
 hammered into the form of a cube. The object of the hammering 
 is to remove the slag, and, when the button is thus cleaned, it is 
 ready for the next operation of cupelling. Any slag left adherent 
 to the button shows itself when cupelling as a slight crust at the 
 edges of the cupel, and many retain a little lead, thus causing a 
 loss. 
 
 Cupelling should be done at a low temperature so as to make 
 feather-litharge upon the surface of the cupel, otherwise there 
 is an increased cupellation-loss by volatilization. For gold, this is 
 less important. Toward the end of the cupelling, and as the silver 
 button begins to show up, the cupel should be shoved back an inch 
 or two to where the muffle is hotter. This is to insure that the 
 last traces of lead are removed. By close watching you may see 
 the 'blick' or sudden brightening of the silver bead which remains 
 upon the cupel. Some assayers think that at this stage it is 
 essential to remove the cupel from the furnace, but, evidently, 
 when the silver button has become solid, as it now does, no further 
 silver loss can occur. 
 
 The silver bead, when brushed to free it from bone-ash, may be 
 weighed, then flattened, dropped into hot dilute nitric acid in the 
 parting capsule, and parted, as already described under the head 
 of scorification. If, however, the amount of gold by weight 
 approximates one-fourth of the button, this treatment, with 
 weaker acid, must be followed by one with acid of 1.3 sp. gr. 
 (7 parts strong acid to 4 parts water). When the gold is in 
 excess of this specified amount, parting proceeds slowly and 
 imperfectly, and the silver cannot all be removed. In such a 
 case the button is wrapped up in two or three grams of lead 
 foil together with a little silver foil, in quantity such that the 
 total silver present shall be at least three times that of the con- 
 tained gold. This is quickly cupelled on the old cupel or on 
 another small one of one inch diameter. The resultant button 
 can now be easily parted. Where it is not desired to determine 
 the silver, but the gold only, a small piece of silver foil may be 
 added to the lead button just before cupelling, or it may be added 
 in the crucible. 
 
 ACTION OF THE FLUXES. As regards the action of the soda and 
 borax, an easy aid to the memory is to say: S equals S and b 
 
AUSTIN'S FIRE ASSAY. 63 
 
 equals b, or that the soda is intended to flux the silica and the 
 borax the bases. In an extremely silicious ore the slag, even with 
 a good heat, may be stiff, so that when poured it runs with dif- 
 ficulty (sometimes not at all) from the crucible. Now, the 25 gm. 
 soda already prescribed is sufficient for all ores except the very 
 silicious ones, and when, as in the above case, this trouble threat- 
 ens, it may be got over by adding another 10 gm. soda direct to 
 the crucible while still in the furnace. The addition of borax 
 will not help. When grinding up the ore on the plate, its gritty 
 feel is a sign that the ore is silicious. An ore consisting of lime- 
 stone, dolomite, or heavy spar, while of a light color, would not 
 have a gritty feel under the muller. 
 
 On the other hand an ore which the assayer knows to be basic, 
 if acting in this way, would need more borax (but not more soda) 
 added to the charge, in order to improve its fluidity. 
 
 The way the scarcely fused slag acts is significant. When it 
 can be strung out in threads like molten glass, it is silicious, but 
 when it breaks short, as though brittle, in a way to be learned 
 by observation, it is too basic. 
 
 Referring to the table and the comments thereon (Chapter IX), 
 it will be noticed, that in the nail-assay, the lead is not to be 
 regarded as a base which will go into the slag, and hence no 
 borax is to be provided for because of its presence. With zinc 
 and iron present, as in blende or pyrite, in calamine or in iron 
 ore, or with the alkaline earths as in heavy spar and limestone, 
 it is a different matter, since these bases enter the slag and they 
 need borax to flux them. The amount of borax prescribed is from 
 5 to 15 gm. for the half assay-ton of ore taken. The smaller 
 amount of 5 gm. is no more than enough to form a cover to the 
 charge, though enough for small amounts of bases. The maxi- 
 mum amount of 15 gm. is to be used where there is much base 
 present. In particular, when there is much blende, we would use 
 the maximum amount of borax, since blende is disposed to make 
 a stiff slag. 
 
 In the nail-assay, enough litharge is used to form a lead button 
 of 1 8 to 20 grams, which is large enough to remove or collect all 
 the silver or gold which the ore contains. Since an excess of 
 
64 AUSTIN'S FIRE ASSAY. 
 
 argols is also used, this lead is all reduced. Should there be lead 
 in the ore, this would also go into the lead button, and hence the 
 assayer must take care correspondingly to lessen the litharge. 
 For example, in the case of a galena carrying say 70% lead 
 there would be 10.5 gm. lead, and hence only half the usual 
 quantity of litharge would be needed. There is considerable 
 lee- way allowable in the size of the button. As little as 15 or as 
 much as 25 gm. will equally serve, but when below the former 
 amount, it is doubtful whether all the gold and silver is collected, 
 while, where the button is so large as 25 gm., it is longer in 
 cupelling, and volatilization loss increases. 
 
 Since the reducing power of a gram of argols is such as will 
 reduce approximately 8 gm. lead, the amount of 3 or 4 gm. pre- 
 scribed will be in excess of all requirements, so that all the lead 
 present in the charge is bound to be reduced. 
 
 The limits in the number of lo-penny nails are from one to six, 
 and the number is dependent upon the quantity of sulphur present. 
 Where the ore consists of pyrite, or of blende, the larger number 
 of nails is used, in the case of galena two, while those ores con- 
 taining little or no sulphur need but a single nail. One nail at 
 least is used, since it may be possible, even in a supposedly oxi- 
 dized ore, that there is still a little sulphur. At any rate a nail can do 
 no harm. It cannot be said that too many nails are an objection, 
 except as they take time to remove from the crucible before 
 pouring, and that a drop of lead from the button might become 
 adherent to one of them. 
 
 To show how much borax and how many nails are needed in 
 assaying sulphide ores, let us take the case of the following 
 sulphides. They contain, when pure, as follows : 
 
 Galena PbS .87% Pb 13% S 
 
 Blende ZnS 67% Zn 33% S 
 
 Pyrite FeS 47% Fe 53% S 
 
 It will be seen, that starting with blende, its sulphur is 33%. 
 'The zinc makes up the remaining 67% and is twice the sulphur, 
 that is, in blende, the sulphur makes up one-third, the zinc two- 
 thirds. Next we find the quantities of base varying by 20% up 
 
AUSTIN'S FIRE ASSAY. 65 
 
 and down. Thus all the numbers are deduced if we know the 
 sulphur in blende. 
 
 Now as to the action of borax. Referring to our data we need 
 not consider the lead as a base, since it enters the button, not 
 showing up in the slag at all. Of the two remaining bases the 
 zinc is more basic, blende needing more borax than pyrite. For 
 the sulphur contents we would conclude, that the pyrite and 
 blende need the maximum amount of six nails, the galena two. 
 As a matter of fact we use many for both blende and pyrite, and 
 few for galena. 
 
 CHEMICAL REACTIONS DURING FUSION. Lead oxide is acted 
 on by the carbon in the argols, the reaction being as follows : 
 
 PbO + C = Pb + CO 
 CO + PbO == CO 2 + Pb 
 
 The button, thus reduced, carries down with it the gold and 
 silver contained in the ore. If copper and antimony are present, 
 they are reduced also, making a hard button, if there is much of 
 either metal. When such a lead button is cupelled, some gold 
 or silver is carried into the cupel, so that the nail-assay is suited 
 only to clean ores. 
 
 Silica, itself infusible at high temperatures, is fluxed by soda 
 thus: 
 
 SiO 2 + Na 2 CO 3 = Na 2 SiO 3 + CO, 
 
 the carbon dioxide passing off as a gas. By calculation, there 
 must be 26 gm. sodium carbonate to form the mono-silicate where 
 there is present 15 gm. (nearly one-half assay-ton) silica as in the 
 case of a pure quartz ore. 
 
 If crude borax is used in the assay, its water of composition is 
 first given off thus: 
 
 Na 2 B 4 O 7 + ioH 2 O + heat = Na 2 B 4 O- 
 At first it swells up, sometimes to the extent of overrunning 
 the side of the crucible, then melts down into borax-glass, hence 
 the advantage of using this latter. Borax has a solvent e.ffect 
 upon bases, much as follows: 
 
 FeO + Na 2 B 4 O 7 = Na 2 FeB 4 O 8 
 CaO + Na 2 B 4 O 7 = Na 2 CaB 4 O 8 
 
66 AUSTIN'S FIRE ASSAY. 
 
 forming fusible berates. Borax, however, dissolves bases in all 
 proportions. 
 
 The nails act upon the sulphides much as follows : 
 
 FeS 2 + Fe = 2FeS 
 
 The iron sulphide or matte thus formed is dissolved by an 
 alkaline carbonate, hence the necessity of having plenty of soda 
 at all times in the charge. Galena is decomposed as follows : 
 
 PbS + Fe = Pb + FeS 
 and for lead in oxidized form 
 
 PbO + Fe = FeO + Pb 
 
 so that, even were there no argols present, the lead would be 
 reduced to metallic form. Blende is decomposed according to 
 the equation : 
 
 ZnS + Fe -f PbO = FeS + ZnO + Pb 
 
 THE REGULAR OR LITHARGE ASSAY (also called 'nitre assay'). 
 This is used for impure ores, or those containing certain im- 
 purities as : -As, Sb, Fe, and Cu, which must be eliminated from 
 the lead button before cupelling. To do this the following charge 
 is made up: 
 
 25 gm. soda (mono-carbonate), 30 to 120 gm. litharge, 0.5 
 A. T. ore, 3 to 15 gm. borax-glass, with argols or nitre to suit. 
 
 The quantity, either of argols or nitre, which is to be used will 
 depend upon the reducing power of the ore, and is so regulated 
 that a lead button of 20 gm. is reduced from the litharge in the 
 charge. 
 
 To ascertain this reducing power we take a preliminary charge 
 as follows: 
 
 o.i A. T. ore, 40 gm. litharge, and 10 gm. soda. 
 
 This is melted down rapidly in a lo-gm. crucible, withdrawing 
 it, even before quiet fusion, and pouring into a conical mold. 
 The lead button obtained will vary, from nothing if it is an 
 oxidized ore, to upward of 30 gm. if the ore is an iron pyrite, 
 the reducing action being due to the presence of sulphur in 
 sulphides. If the reducing power of this one-tenth be multiplied 
 by five we have that of a half assay-ton. We may have either: 
 
AUSTIN'S FIRE ASSAY. 67 
 
 1. The reducing power of one-tenth A. T. less than 4. In 
 this case multiply by 5, subtract the product from 20, and divide 
 the remainder by the reducing power of the argols, say 8. This 
 will give us the number of grams of argols to be used. For 
 example, let the reducing power be 2. Then 
 
 20 (2 X 5) = I - 2 S g m -> 
 
 8 
 the quantity of argols to add in the regular assay. 
 
 2. In the second case, the reducing power being greater than 
 4, we multiply by 5, subtract 20 from it, and divide the remainder 
 by 4, the oxidizing power of nitre. Thus suppose the reducing 
 power to be 22 ; then we have 
 
 (22 X 5) 20 22.5 gm. 
 
 4 
 to be added to the charge in order to secure a button of 20 grams. 
 
 Returning to a consideration of the regular charge we may 
 note: 
 
 The soda is added, as in the nail-assay, to flux the silica. 
 
 The litharge varies according to the impurities in the ore, 
 which may be estimated from the appearance of the ore, or from 
 experience with it. For an ore containing but little of them, we 
 may use 30 gm., while for ores carrying much arsenic, antimony, 
 tellurium, or copper, we use to the limit of 120 gm. PbO. It will 
 be seen, that since but 20 gm. is reduced from the litharge, there 
 will be left anywhere from 10 to 100 gm., which goes into the 
 slag. It is this litharge which acts on these impurities, oxidizing 
 and dissolving them. 
 
 The borax-glass acts precisely as in the nail-assay, forming 
 double berates with the bases. 
 
 This charge is placed in a 2O-gm. crucible, is mixed there with 
 a spatula, and the borax sprinkled upon it to form a cover. It 
 is placed, either in the muffle or among the coals of a wind- 
 furnace, brought to fusion, but at such a rate as not to permit 
 any of its contents to escape from the crucible or to boil over, 
 and finally brought to quiet fusion. This work will take from 
 one-half to one hour. There is no objection to retaining the melt 
 
68 AUSTIN'S FIRE ASSAY. 
 
 in the furnace even after this time. On the other hand, if the 
 crucible is withdrawn before reduction is complete, not all the 
 silver is obtained. In pouring, rest the lip of the crucible upon 
 the mold, and pour at an even rate. If the slag appears to be 
 stiff, when melting down an ore known to be silicious, it may 
 be made more fluid by the addition of soda. For a known basic 
 ore, on the contrary, we must add borax to produce the same 
 effect. This addition may be made by putting in the amount by 
 means of a long-handled spoon forged on the end of an iron rod 
 two feet long, or, in an off-hand way, by means of a scorifying 
 dish held by the crucible tongs. The button which results from 
 this assay should be soft, and when cupelled, show freedom from 
 impurities. If not in this condition, the assay should be repeated, 
 using a much larger quantity of litharge. 
 
 CHEMICAL REACTIONS OF THE REGULAR ASSAY. A sulphide 
 ore, having a high reducing power, is permitted, in the regular 
 assay, to act upon and reduce from litharge 20 gm. lead, much as 
 follows : 
 
 FeS 2 + 5?bO = 5?b + FeO + 2SO 2 
 PbS + 2PbO = 3?b + SO 2 
 ZnS + 3?bO = 3?b + SO 2 + ZnO 
 
 The sulphur dioxide escapes as a gas, and the oxides are dis- 
 solved by the borax. But there remains a further amount of the 
 sulphides, for the oxidation of which nitre has been provided, 
 and which, at a high temperature, is decomposed. 
 
 2KNO 3 = K 2 O + 2NO + 30 
 
 the oxygen thus furnished acting directly upon the sulphides as 
 for example : 
 
 FeS 2 + 5O = FeO + 2SO 2 
 CuS + 30 = CuO + SO 2 
 Also 
 
 SO 2 + O = SO 3 
 K 2 + S0 3 = K 2 S0 4 
 and this floats on top of the charge. 
 
 We may note here, that where, in the nail-assay, copper is re- 
 duced to metallic form entering the button, it is here oxidized 
 
AUSTIN'S FIRE ASSAY. 69 
 
 and goes into the slag. Arsenic and antimony sulphides are also 
 oxidized according to the formula : 
 
 As 2 S 3 + 90 = As 2 3 + 3 S0 2 
 
 Sb 2 S 3 + 9 = Sb 2 3 + 3 S0 2 
 and enter the slag, being dissolved by the litharge. 
 
 Oxidized ores have no reducing power, and to them we add 
 argols, which act as in the nail-assay: 
 
 PbO + C = Pb + CO 
 
 We are careful in this case to add no more (as may be determined 
 by a preliminary assay) than will reduce the 2O-gm. button, since 
 it is essential that there be litharge to enter the slag to dissolve 
 impurities. 
 
XIV. ROASTING OF ORES. 
 
 Where an ore contains but little gold, and we wish, therefore, 
 to determine it in a large quantity, the ore may be first roasted. 
 To do this, weigh out one assay-ton of the ore and put it in a 
 4-in. roasting-dish. The dish is first carried at a low heat at the 
 front of the muffle. Where there is much pyrite present it will 
 snap and fly, causing mechanical loss. This may be guarded 
 against by partly covering the dish with a crucible cover, leaving 
 the front side of the dish exposed. The dish is gradually ad- 
 vanced into the muffle, finishing the ore at a red heat, but so that 
 it is not fused together. During the roasting the contents of the 
 dish are stirred several times to expose fresh surfaces to the 
 action of the air. Roasting may be regarded as complete when 
 no odor of sulphur is observable upon removing the dish and 
 smelling of its contents. 
 
 CHEMISTRY OF ROASTING. The ore may consist of a silicious 
 gangue together with sulphides of iron, copper, zinc, and lead. 
 Of these, pyrite is the most easily decomposed, the first equivalent 
 of sulphur being driven off at a low temperature thus : 
 
 FeS 2 + heat = FeS + S 
 
 The sulphur, as it is evolved, encounters the air and is burned 
 to SO 2 . With an abundance of air, and at an increased tem- 
 perature, the iron sulphide is roasted as follows : 
 
 3FeS + iiO = 2SO 2 + Fe 2 O 3 + FeSO 4 
 
 The odor from the escaping sulphuric dioxide is quite evident. 
 We also have when there is chalcopyrite, 
 
 FeCuS 2 + 6O = CuO + FeO + 2SO 2 
 
 As the heat is increased to 590, the iron sulphate becomes 
 decomposed, transferring its SO 3 to the copper thus, 
 
 FeSO 4 + CuO = FeO + CuSO 4 
 
AUSTIN'S FIRE ASSAY. 71 
 
 The ferrous oxide is then further oxidized, giving, where much 
 of it is present, a reddish color to the roast, namely, 
 
 2FeO + O = Fe 2 O 3 
 Zinc and lead sulphides are roasted as follows : 
 
 ZnS + 30 = ZnO + SO 2 
 PbS + 30 == PbO + S0 2 
 
 and at a still higher temperature (750) the copper sulphate is 
 decomposed, transferring its sulphuric anyhydride to the zinc 
 and lead compounds, and forming sulphates, 
 
 CuSO 4 + ZnO = ZnSO 4 + CuO 
 CuSO 4 + PbO == PbSO 4 + CuO 
 
 These newly formed sulphates are hard to decompose (above 
 1000 being needed) by the action of heat, especially since the 
 finishing heat must be high, silver will be volatilized. When so 
 roasted, these sulphates do not affect the reducing power of the 
 ore. Antimony and especially arsenic are oxidized, forming 
 volatile compounds, and are thus, in part at least, removed. 
 As 2 S 3 + 90 = As 2 O 3 -|- 3SO 2 
 
 Ores, because '-of the time and attention needed, are seldom 
 roasted; the assayer generally takes the shorter method of a 
 nail or regular assay, making them in duplicate, and combining 
 the cupelled buttons when the amount of gold is but small. 
 
 THE ASSAY OF ROASTED ORE FOR SILVER AND GOLD. One assay- 
 ton of ore has been roasted so that it is oxidized and has lost 
 its reducing power. We accordingly take : I A. T. ore, 40 gm. 
 soda, 40 gm. litharge, 20 gm. borax, and 2.5 gm. argols. 
 
 This is placed in a No. F crucible, if it is to be done in a wind- 
 furnace, the borax being used as usual as a cover. The excess 
 of litharge, over that needed to form the lead button, is intended 
 to take care of the impurities present. Where there is much of 
 these, the roasting method would not be followed. 
 
 The charge is melted down to quiet fusion, and is then poured 
 into a conical mold. The quantity of slag may be so much as to 
 overflow the mold, but this does not matter. A 2O-gm. button 
 should result. In other respects the assay is conducted as in the 
 regular assay. 
 
XV. ASSAY OF MATTE. 
 
 This is an artificial iron-copper sulphide, a furnace product 
 consisting of iron and sulphur, the iron being more or less re- 
 placed by copper and perhaps lead. It contains silver and gold 
 taken from the ore-charge, and is assayed by scorification using 
 45 to 60 gm. test-lead, o.i A. T. the ore, and 2 to 3 gm. borax. 
 
 The higher the matte in copper, the more test-lead we add. In 
 cupelling one can tell by the appearance of the cupel whether the 
 copper has been entirely removed from the button, since at the 
 finish the surface of the cupel immediately around the button is 
 a little lighter in color than the rest. Also, upon dissolving the 
 button, if the solution is neutralized with a little ammonia, the 
 absence of blue color denotes freedom from copper. 
 
XVI. HIGH-GRADE SILVER-SULPHIDE ASSAY. 
 
 The assayer may be called upon to assay ores rich in silver, 
 or silver sulphides of several thousand ounces value, the product 
 of a silver-lixiviation process. The sulphides may contain much 
 copper, which has been precipitated at the same time with the 
 silver, so that we scorify the sulphide, making a charge as fol- 
 lows : o.i A. T. ore, 45 gm. test-lead, and 2 gm. borax-glass. 
 
 It will be noticed that the usual quantity of 30 gm. test-lead 
 has been increased because of the presence of copper. A clean 
 ore, no matter how rich, can be quite as well assayed with the 
 usual 30 gm. The borax-glass is added to take care of the copper 
 oxide. 
 
 .When the button resulting from thfs scorification has been 
 cupelled, the litharge-soaked part of the cupel is saved, together 
 with the slag of the scorification. These two are then coarsely 
 powdered in a mortar, or on the grinding plate, to about 2O-mesh 
 and assayed with the following charge : Slag and cupel, 30 gm. 
 litharge, 20 gm. soda, 2.5 gm. argols, and 30 gm. borax. 
 
 The bone-ash of the cupel, being chiefly calcium phosphate, 
 needs much borax to flux its lime base. The material, being quite 
 oxidized, needs just enough argols to give the .usual 2O-gm. 
 button. A small amount of silver is recovered, which is added 
 to the other for the total gold-silver button, and the result in 
 milligrams multiplied by 10. The buttons are then parted for 
 gold. Generally four to six results are made from as many tenths 
 of an assay-ton. The buttons may be parted two and two, so that 
 the gold is weighed in one-fifth of an assay-ton, and the results 
 must check one another. 
 
XVII. ASSAY OF CYANIDE SOLUTIONS. 
 
 Cyanide solutions, containing silver and gold, may be either the 
 solution containing the precious metals before extraction, or the 
 solution after it has passed through the zinc-boxes, having had 
 its values mostly removed (barren solution). In the latter case 
 a much larger portion of the solution must be taken for assay 
 than in the former. 
 
 Of the numerous methods devised we give two. 
 
 FIRST METHOD. This is accurate and simple, but slower than 
 the second. It consists in evaporating to dryness upon a hot- 
 plate, say 2 A. T. solution in a lead-tray made by turning up the 
 edges of lead-foil such as is used in assaying. This is formed over 
 a wooden block and may be 2 by 2 by ^4 m - deep. The evapora- 
 tion being finished the tray is then bent into a compact form and 
 directly cupelled. If, however, the residue is considerable, it may 
 be advisable to scorify the lead before cupelling. The results are 
 to be reported in ounces per ton of solution. 
 
 SECOND METHOD. This method is more rapid than the first, 
 and can be used with a larger bulk of solution, and is performed 
 as follows : 
 
 Measure into a No. 4 beaker 5 A. T. or 146 c.c. of the solution. 
 (If the solution is a 'barren one,' 10 A. T. should be used.) Add 
 enough KCy to the solution to bring it up to 0.5%, KCy.* Thus 
 if the strength is 0.2% we shall need to bring it to 0.5% the addition 
 of 0.3% of 146 gm., or 0.43 gm. of the salt. Stir into the solution 
 10 c.c. of a 10% solution of lead acetate (PbA) slightly acidified 
 with acetic acid. Add from the point of a spatula about 0.4 gm. 
 zinc dust, weighing it until able to judge of the quantity needed. 
 This amount of zinc is amply sufficient to precipitate not only all 
 the lead, but also the gold and silver in the solution. Thus we 
 have : 
 
AUSTINS FIRE ASSAY. 75 
 
 Zn + PbA =ZnA + Pb 
 
 Zn + 2KAuCy 2 = K,ZnCy 4 + 2Au 
 Zn + 2KAgCy 2 = K 2 ZnCy 4 + 2Ag 
 
 Stir it and bring to boiling. Add 15 c.c. strong HC1 to dissolve 
 the excess of zinc. Leave the solution on the hot plate until the 
 zinc is completely dissolved, and the lead has gathered into a 
 sponge. As soon as thus dissolved, remove from the heat to 
 prevent the lead from being at all dissolved. Holding the lead 
 with a flattened glass rod, decant the solution. Wash the sponge 
 with water, squeeze it into a clot with the fingers, and place it 
 upon a piece of lead-foil weighing 5 to 6 gin. Fold the foil so 
 as to leave an outlet for steam, then cupel it. The resultant 
 button will give the silver and gold in the 5 A, T. solution. 
 
 * To determine the strength of the KCy solution we may proceed as 
 follows : Have 50 c.c. burettes. Fill one with AgNOs solution consisting 
 of 17 gm. of the pure salt in one litre of water. Fill the other with the 
 solution to be tested. Measure into a small beaker, 13 gm. of the KCy 
 solution. Run in cautiously the standard AgNOs solution until the white 
 precipitate, which has appeared, ceases to dissolve where the contents of 
 the beaker has been stirred, that is, when only a faint opalescense remains. 
 Ten c.c. of the standard solution equals \%, so that the reading is to be 
 divided by 10. Thus suppose our reading was 14.5 c.c., then we will 
 have 1.45% KCy present. For very dilute solutions we may take 130 c.c., 
 and then the burette reading must be divided by 100. 
 
XVIII. ASSAY OF BASE-BULLION. 
 
 Base-bullion, or work-lead, is the product of the silver-lead 
 blast-furnace, and may be considered as an unrefined lead con- 
 taining upward of one per cent of silver together with a few 
 per cent, at most, of impurities, as arsenic, antimony, and copper. 
 
 The sample may come into the assayer in the form of cylin- 
 drical 'chips' 2 in. long by l /s in- diam. Since two of these pieces 
 are taken from each ingot or bar, there would be 800 of them for 
 a 4OO-bar carload, and amounting in quantity to a quart. A 
 plumbago crucible is made red-hot in the wind-furnace, and the 
 chips are shot into it by means of a scoop. The lead quickly melts 
 down, and is stirred with a wooden stick. It is then lifted by 
 
 J 
 
 8 
 
 FIG. 40. BASE-BULLION SAMPLE. 
 
 the tongs, given a swirling motion to complete the mixing, and 
 quickly poured into a mold 8* in- long by 2.5 in. wide, where it 
 forms a bar about 0.3 to 0.4 in. thick. This bar is cut longi- 
 tudinally, by means of a chisel, into two, each half being num- 
 bered; one of these is sent to the refinery and one retained by 
 the works producing the base-bullion. Assuming that the lower 
 half is retained, four pieces marked in as i, 2, 3, and 4, each of 
 a little more than 0.5 A. T. are cut from the bar. They represent 
 a fair average of the bar, since the exterior runs a little higher 
 than the central portion at the time it solidifies. 
 
 In case of an impure bullion, each sample should be scorified 
 with extra test-lead and a little borax before cupelling. Where 
 the sample is sufficiently pure, direct cupellation is satisfactory. 
 
AUSTIN S FIRE ASSAY. 77 
 
 Some assayers prefer always to scorify before cupellation, con- 
 tending that the loss is less in scorifying than in cupelling. 
 Cupellation should proceed with the formation of feather-litharge, 
 and, just prior to brightening, the cupel should be pushed back a 
 little to the hotter part of the muffle. Where the buttons are 
 large it is a good plan to have cupels set just behind in the muffle 
 to cover them when finished as in the assay of silver bars. The 
 buttons are parted as in ores. 
 
XIX. ASSAY OF SILVER BARS OR INGOTS. 
 
 Silver bars are assayed to determine their fineness, or parts 
 in a thousand. Thus U. S. coin silver is 90% or 900 fine, the 
 residue being copper. Ordinary ingots or bars from stamp-mills 
 may vary from 250 to 900 fine in silver and gold, the impurity 
 being copper. By fine silver we mean pure silver bars of 1000 
 fine, though the term is also applied to bars of over 900 fine. 
 
 Mill-bars may be assayed from granulations or from chips. 
 To make granulations, when the silver has been melted and 
 refined in a crucible, and just before it is poured into ingots, 
 some of the molten metal is removed by means of a heated ladle 
 or scorifier, and poured from a height of three feet into a bucket- 
 ful of water. The metal is broken into fragments and shot of 
 all sizes from which may be selected particles of any desired size. 
 Where the ingots come to the assayer ready cast, he takes his 
 sample by chips cut from opposite edges of the bar, the chisel 
 cutting in a converging way into it to take out a wedge-shaped 
 chip of from 2 to 5 gm. These are placed in a scorifying dish 
 in the muffle, and brought to a just-visible red heat in order to 
 anneal them. They are then hammered out upon an anvil, again 
 annealed, and put through rolls (Fig. 41) where they are brought 
 down thin enough to cut readily with scissors. Before use this 
 sheet-metal is cleaned by rubbing with emery cloth. 
 
 The operation is as follows: 
 
 TRIAL ASSAY. To determine approximately the fineness of the 
 sample, we weigh out 500 mg. upon a button-balance, selecting 
 from the granulations larger pieces, and finally the finest particles, 
 to obtain the exact amount. This weighing need be exact but to 
 o.i mg. Where the sample is taken from a rolled sample, we may 
 cut out a longitudinal strip 0.3 to 0.4 in. wide and cut this trans- 
 versely into various sizes, down to mere shavings, in quantity more 
 than sufficient for an exact weighing. This is wrapped up in 20 gm. 
 
AUSTINS FIRE ASSAY. 
 
 79 
 
 of silver-free lead-foil, folding it in both directions so as to form 
 a cornet. The sample is dumped into it from the pan of the 
 balance, and the foil folded up ready for cupelling. Cupellation 
 is performed at a moderate temperature to form feather-litharge. 
 Near the end, the cupel is pushed back into the muffle and finished 
 at a higher temperature. A hot cupel should be standing ready 
 
 FIG. 41. METAL ROLLS. 
 
 in the muffle and, at the moment of blicking, it should be put on 
 the other, thus excluding air-currents which tend to cause so 
 large a button to sprout or spit. This is, as its name indicates, 
 
8o AUSTIN'S FIRE ASSAY. 
 
 the formation of excretions or projections upon the surface of 
 the button due to the pressure of escaping oxygen which has 
 just been absorbed by the molten metal, and which, upon solidi- 
 fication is seeking to escape. The cupel is gradually moved 
 forward to the front of the muffle and as soon as the silver gets 
 solid, is' removed. The resultant button is weighed, and the 
 number of milligrams multiplied by 2 will give the approximate 
 fineness. The real fineness is greater than this by 10 to 15 fine. 
 PREPARATION OF THE PROOF-BUTTON. The button, which has 
 been carefully weighed, is put in the right-hand pan of the bal- 
 ance, and a 5oo-mg. weight in the left one. Enough copper, cut 
 from copper foil, is added with the silver button to make up the 
 weight to 500 mg. This is transferred from the pan to a cornet 
 of lead-foil in the manner already indicated, the quantity of foil 
 to be taken being determined from the table herewith. 
 
 WEIGHT OF LEAD 
 FINENESS 
 
 IN GRAMS 
 
 950 to looo 
 800 to 950 
 600 to 800 
 600 and less 
 
 5 
 10 
 
 15 
 20 
 
 Thus, suppose we had found by the trial-assay a button of 312 
 mg. equal to an approximate fineness of 624, we would see from 
 the table that 15 gm. sheet-lead would be enough. Two other 
 portions of the sample are taken, each of 500 mg., and wrapped in 
 the same quantity of lead-foil as the proof. The three are 
 cupelled abreast in the muffle as in Fig. 42, taking care to cupel 
 them all at the same temperature and in such a way as to form 
 feather-litharge. The cupellation loss of each of these may be 
 considered as being the same. The fineness is, therefore, the sum 
 of the two outer buttons plus twice the loss in cupelling. Thus 
 suppose the results are as given in Fig. 42. Then, 
 
 3i8 + 3i7 + (2 X 6) == 647, 
 which will be the actual fineness. It will be noticed that the proof 
 
AUSTIN'S FIRE ASSAY. 
 
 8 1 
 
 is made up to imitate the sample. Strictly speaking, 5 to 10 mg. 
 of silver should be added to the silver button to allow for the 
 cupellation-loss, or we may take the same weight (in this case 
 512 -f- 6 = 518 mg.) of pure silver- foil. The rest of the proof 
 
 3/7 
 
 FIG. 42. CUPELLING FINE SILVER. 
 
 is assumed to be copper ; but its exact amount may be determined 
 by assay in the wet way. The duplicates should not vary from 
 one another more than one milligram. 
 
 The gold is determined by parting the duplicates together 
 the weight in milligrams being the fineness, and this may be 
 checked by parting the proof-button. 
 
XX. ASSAY OF BLISTER OR PIG-COPPER 
 CONTAINING SILVER AND GOLD. 
 
 The losses of silver in attempting to assay copper matte and 
 blister-copper by scorification are quite large, and it is accordingly 
 best to remove the copper by a wet method before fire-treatment. 
 
 The Whitehead method of blister-copper assay is as follows : 
 One-half of an assay-ton of the granulations is introduced into a 
 No. 4 beaker with 50 c.c. water. To this we add 25 c.c. nitric 
 acid of 1.42 sp. gr. and, when action has nearly ceased, another 
 25 c.c. is put in. The solution is allowed to stand in a warm 
 place until the red fumes have disappeared, after which it is 
 diluted to 300 c.c. The copper and silver have gone into solution 
 thus : 
 
 3 Cu + 8HN0 3 = 3 Cu (N0 3 ) 2 + 4 H,O + 2NO 
 3 Ag + 4 HN0 3 = 3 AgN0 3 + 2 H 2 O + NO 
 while the gold remains as an undissolved residue. 
 
 We now run into the liquid, drop by drop, a 10% solution of 
 common salt (NaCl) in slight excess to precipitate the silver as 
 chloride. 
 
 AgNO 3 + NaCl = AgCl + NaNO 3 . 
 
 The solution is well stirred, and 10 c.c. of a 10% solution of 
 lead acetate added and, still stirring, 2 c.c. of sulphuric acid 
 (i part acid to I part water), after which it is allowed to stand 
 until settled. The sulphuric acid precipitates the lead thus: 
 
 PbA + H 2 SO 4 = PbSO 4 + H 2 A, 
 
 and the lead sulphate, in falling, carries down the gold mechani- 
 cally. Finally the solution is filtered off, and the precipitate 
 washed on the filter to remove copper salts. The filter paper is 
 removed, folded over the precipitate, placed in a scorifying dish 
 on top of 10 gm. of test-lead, and then gradually burned at the 
 front of the muffle. A cover of 10 gm. more of test-lead, and 
 
AUSTIN'S FIRE ASSAY. 83 
 
 I gm. of borax is added, and the whole started to scorifying. 
 When scorified the lead buttons, weighing 6 to 8 gm., are cupelled 
 with the formation of feather-litharge and the gold-silver button 
 weighed and parted. The work should be done in duplicate, and 
 the results should agree almost exactly. 
 
 The method may also be used for the determination of gold 
 and silver in base ores, such as gray copper or arsenical pyrite, 
 or of high-grade copper matte. 
 
XXI. ASSAY OF GOLD BULLION. 
 
 The sample is generally taken by two chips from opposite edges 
 of the bar as in the silver assay. For a small bar, as little as 
 100 mg., and for a large one, 500 mg. may be taken. 
 
 The base metals are first removed by cupellation, wrapping the 
 500 mg. in 10 gm. of lead-foil. There remains behind the pure 
 gold-silver button, which is weighed to determine the fineness of 
 gold and silver together. 
 
 The silver is removed from the gold by inquartation, taking 
 500 mg. of the original sample, to which is added three times the 
 weight of the pure silver-gold button. This is cupelled with 5 gm. 
 of lead-foil, since it is not desired to remove all the copper. If no 
 copper is present, 10 mg. of pure copper should be added. The 
 resultant button is weighed, then flattened out under the hammer. 
 It is annealed and then passed through the rolls (Fig. 43), which 
 draws it out into a piece about four inches in length. When cold 
 it is rolled around a lead pencil into a spiral coil, and parted in a 
 porcelain capsule. It is boiled first for ten minutes with I to I 
 acid (1.20 sp. gr.), then with stronger acid of 7 parts acid to 4 
 of water (1.30 sp. gr.) for the same period for the removal of the 
 last traces of silver. The acids are decanted, and the residue, 
 which is quite coherent, washed with pure hot distilled water. 
 The gold is now dried, ignited at a low-red heat, let cool, and 
 weighed. This weight, subtracted from the first one, gives the 
 contained silver. The results, as before, are doubled to express 
 the fineness. 
 
 This method is sufficiently exact for ordinary commercial 
 assays. For the more precise methods of the government assay 
 offices, see Furman's 'Manual of Practical Assaying/ page 246. 
 
XXII. ASSAY OF ORES CONTAINING METALLICS. 
 
 Ores containing native gold, silver, or copper, and by.-products 
 of metallurgical operations having metallic particles, prills, or 
 scales, are assayed as follows : 
 
 The ore is pulverized upon the grinding plate, the metallics 
 flattening out into scales which will not go through the screen. 
 We accordingly have the screened material, which must be thor- 
 oughly mixed before assaying it, and the metallics remaining on 
 the screen. Both the fines and the metallics are weighed sep- 
 arately. 
 
 Either one of the following methods may now be followed : 
 
 Method i. In the total quantity, determine the percentage of 
 metallics. Where a half-assay-ton is to be taken, figure this per- 
 centage in grams. Place that quantity on the pan of the pulp- 
 balance, and make up to the half-ton with fines. We thus assay 
 the whole sample in the proper relative proportions, and easily 
 estimate the average result. 
 
 Method 2. Where the metallies are irregular in value and 
 small in quantity, containing, say, silver, copper, and gold par- 
 ticles, it is better to assay the whole and determine the value in 
 ounces per ton. Thus if the metallics weigh 7.3 gm., giving in 
 that quantity 156 mg., we would have 29.166 X 156 623.3 oz - 
 
 7-3 
 
 per ton. Let us say that the fines weighed 720 gm. and assayed 
 1 15.0 oz. per ton. Then, to get the average value, we would have : 
 
 720.0 X 115.0 = 82800.0 
 7.3 X 623.3 = 4550-1 
 
 727.3 into 87350.1 = 1 20. i oz. per ton. 
 Where the fines preponderate so greatly, we may expect this 
 result to be but little more than the lower figure of 1 15 oz. per ton. 
 
86 AUSTIN'S FIRE ASSAY. 
 
 _% : "- 
 
 An example of an ore carrying native copper would be as 
 follows : 
 
 The whole sample, consisting of pieces 1.25 in. diam. and 
 smaller, weighed 142 oz. and contained 2.75 oz. of metallics which 
 were picked out, freed from adherent gangue by hammering 
 them, and were, by the Whitehead process (Chapter XX) found 
 to contain 54 oz. Ag per ton. 
 
 The residue was regularly sampled, but, when finally ground 
 to pass an 8o-mesh sieve, gave 0.8 gm. metallics and 124 gm. fines. 
 
 On determination the fines gave 0.3 oz. Ag per ton and the 
 metallics were assumed to have the same value as in the first 
 metallics (in the 2.75 oz.). 
 
 Hence we have : 
 
 124.0 X 0.3 = 37.2 
 0.8 X 54-0 = 43- 2 
 
 124.8 into 80.4 = 0.64 oz. Ag per ton. 
 Then, referring to the whole sample, we have: 
 139.25 X 0.64= 89.12 
 2.75 X 54-0 = 148.50 
 
 142.00 into 237.62 = 1.67 oz. Ag per ton. 
 
XXIII. THE LEAD ASSAY. 
 
 Any lead-bearing ore may be assayed for lead, the process 
 being a simple reduction fusion, similar to the nail-assay. Where 
 the ore contains much pyrite it may be first roasted, as described 
 in Chapter XIV, using 10. gm. for the roast. Ores are assayed 
 for lead, using the following charge : 
 10 gm. ore, 
 20 gm. soda, 
 
 5 gm. argols, 
 
 4 to 15 gm. borax fused as a cover), and 
 i to 5 ten-penny nails. 
 
 This charge, in a lo-gm. crucible, is melted down gradually so 
 as to avoid boiling over and is finished at a high heat. As soon 
 as it is in quiet fusion it is poured. It should not be kept long 
 in the furnace as one may do in a silver assay, but must be 
 poured as soon as reduction is complete. 
 
 Drops of lead, still clinging to the nails as they are withdrawn 
 from the crucible before pouring, indicate incomplete reduction. 
 The melt is poured into a conical mold, and, as soon as cool, the 
 lead button at the point of the cone is removed and hammered to 
 remove adherent slag. The button is weighed to the nearest tenth 
 of one per cent. The average of duplicates may be taken. They 
 should agree to within 0.5 to \%. This fire method of lead assay- 
 ing gives results a little lower than the true one, unless there are 
 impurities as copper or antimony in the ore, which being reduced, 
 go into the lead button. Its execution requires the best skill of 
 the assayej, and depends on care and skill in the fire-work. The 
 slag from such an assay pours thicker, or is stiffer, than from a 
 silver assay. The same precautions in regard to fluxing should be 
 taken as in a nail-assay for silver. The soda fluxes the silica and 
 dissolves any matte which may be formed. The quantity of borax 
 must be varied according to the amount of bases present. Lead, 
 
AUSTIN'S FIRE ASSAY. 
 
 since it is reduced, does not enter the slag and hence is not a base 
 that must be taken care of by borax. The nails are put in accord- 
 ing to the needs of the sulphur, the maximum number of five 
 being used for a pyrite or blende ore, while a single one only 
 would be put in for an oxidized ore. 
 
 The fluxes mentioned above are often put in in mixture, the 
 reducing mixture (Chapter VIII) having borax added to it as 
 follows : 
 
 20 parts monocarbonate of soda, 
 5 " flour, 
 5 " borax. 
 They are thoroughly mixed and put through a lo-mesh screen. 
 
 When thus used, the charging of the crucible is quite simple. 
 The measured amount, say 30 gm., is put in, then 10 gm. of the 
 ore, and both mixed with a Spatula. A cover of salt, one-fourth 
 inch deep, follows, then the needed number of nails. Additional 
 borax may also have to be added. 
 
 Some assayers claim that in the fire-work the assay should be 
 melted rapidly and brought to a high heat before pouring. It is 
 well for the assayer to follow the method that gives him the best 
 results. 
 

 4950 
 
 
 179709