The Cyanide Process for the Extraction of Gold and Its Practical Application ...

Manuel Eissler · 1898 · Historical

The Cyanide Process for the Extraction of Gold and Its Practical Application ... by Manuel Eissler (1898). Full text and reference in the Mountain Man Mining…

Public-domain full text preserved in the Mountain Man Mining Library. Original source: archive.org.

at | : . com/

Digitized

16

The Cyanide Process For The Extraction Of Gold

By The Same Author.

THE METALLURGY OP GOLD. A Practical Treatise on the Metallurgical Treatment of Gold-bearine Ores. IncludinfiT the Processes of Concentration, Chlorinatinn, and Extraction by Cyanide, and the Assaying, Melting, and Ri fining of Gold. ByM. EissLKR, M.E. Fourth Edition, Enlarged. With about 250 Illustrations and numerous Folding Plates and Working Drawings. Large crown 8vo, 168. Cloth.

The work is a storehouse of information and valuable data, and we strongly recommenn it to all professional men engaged in the gold-mining industry." — Mining Journal,

THE METALLURGY OP SILVER. A Practical Treatise on tho Amalgamation, Roasting, and Lixiviation of Silver Ores. Including the Assaying, Melting and Kefining of

Silver Bullion, xos. 6d. Cloth.

Third Edition. Crown 8vo,

'A practical treatise, and a technical work which we are convinced will supply a long*felt want amongst pracical men, and at the same time be of value to students and others indirectly connected with the industries." — Mining JoumaL

THE METALLURGY OP ARGENTIFEROUS LEAD. A Practical Treatise on the Smelting of Silver-Lead Ores and the Refining of Lead Bullion. Including Reports on vaiious Smelting Establishments and Descriptions of Modern Smelting Furnaces and Plants in Europe and America. By M Eisslbr, M.E. Crown 8vo, 400 pp., with 183 Illustrations, 12s. 6d. Cloth.

" The numerous metallurgical processes, which are fully and exten- sively treated of, embrace all the stages experienced in the passae of the lead from the various natural states to its issue from the refinery as an article oi commerce."'Praciical Engineer.

A HANDBOOK ON MODERN EXPLOSIVES. A Practical Treatise on the Manufacture and Use of Dynamite, Gun- Cotton, Nitro-Glycerine and other Explosive Compounds, including Collodion-Cotton. With Chapters on Explrsives in Practical Appli- cation. By M. EissLBR, M.E. Second Edition, Enlarged. With 150 Illustrations. Ciown 8vo, X2s. 6d. Cloth.

" Useful not only to the miner, but also to officers of both services to whom blasting and the use of explosives generally may at any time become a necessary auxiliary." — Nature.

London : Crosby Lockwood And Son.

The

Cyanide Process

For The

Extraction Of Gold

And

Its Practical Application On The

Witwatersrand Goldfields

And Elsewhere

By

M. Eissler

MINING ENGINBBR ; A.I.M.B. ; MEMBER OP THE INSTITUTE OF MINING AND METALLURGY AUTHOR OF '' TUB METALLURGY OF GOLD/' ETC., ETC.

London

Crosby Lockwood And Son

7, Stationers* Hall Court, Ludgate Hill

\AU rights reserved']

7'-=0iif

&B,

Preface To The Second Edition.

The continued success of the Cyanide Process for the Extraction of Gold, with the ever-increasing number of mines at which it is employed, has naturally led to a large increase in the number of people interested in the process, and hence the demand for a new and enlarged edition of this work.

Mr. Eissler, who is busily engaged in mining operations in Australia, could not himself uniler- take the task of revision, but as he has supplied the bulk of the new matter which has been incorporated in this edition, my task as editor has been compara- tively an easy one. In the new matter which has thus been added will be found descriptions of the working of the Cyanide Process in other parts of the world besides the Witwatersrand gold fields — notably at the Gibraltar Cyanide Works, New South Wales, and at Deeble's Cyanide Works, Victoria.

An interesting account of the Cyanide Plant re-

Vi Preface.

cently constructed at Mercur, Utah, U.S.A., has come to hand in time for insertion in an Appendix (p. 127). I need hardly add that the task of preparing the present volume for the press, although light, has been an agreeable one, and I sincerely hope that this new edition will meet with a reception as favourable as that accorded to its predecessor.

E. Henry Davies, M.E., F.G.S.

6, Gkeat Winchestm Street, lokdon, e.c.

July, 1898.

P.S. — For the photograph which has been re- produced in the frontispiece to the present edition, illustrating a portion of the Cyanide Plant of the May Consolidated Gold Mining Company, I am indebted to Mr. Ad. Goerz, of London, the Chairman of the Rand Central Ore Reduction Company, who' has also kindly contributed a brief article on the Treat- ment of Battery Slimes. A description of the Cyanide Plant mentioned above has been kindly supplied by Mr. Charles Butters.

It is right that I should acknowledge here the service Mr. Davies has rendered in carefully carrying out the revision of the book for the present edition in accordance with my wishes.

M. EISSLER. Sydney, New South Wales.

Introduction To The First Edition.

It is a great satisfaction to me to be in a position to place before the public, in the present volume, an account of the Extraction of Gold by the Cyanide Process, which I believe will be found by Metal- lurgists who desire to avail themselves of the pro- cess to be sufficiently full and complete for practical purposes. That I am in a position to do this, is owing to the fact of my having made a lengthened stay in the Witwatersrand gold fields — whence I have just returned — and to my having enjoyed there unique opportunities of studying the process in actual operation.

Before entering upon an account of the process, some further introductory remarks will be advisable.

Metallurgy, as applied to preparing and winning the gold from its ores, resolves itself, broadly speak- ing, into two methods, in )ne of which mechanical

viii INTRODUCTION.

means, and in the other chemical means, are employed. I consider the stamping or crushing of the ore in a battery, and the catching of the gold on amalgamated copper plates, a mechanical process, although the recovery of the gold in chemical combination with mercury might entitle it to be considered otherwise. The whole process is nothing more or less than a simple system of concentration, as the gold, owing to its great specific gravity, separates from the lighter gangue in which it is enclosed; and when the pul- verized ore particles are washed away, the gold grains sink and are arrested on the amalgamated smooth surfaces of the plates over which the stream carries them, or get entangled in the hairs of the blankets, or are stopped in the riffles of the sluice box, or what- ever other method is used in the battery to save the free gold.

When gold is in combination with other minerals, more complicated methods have to be employed to collect it by mechanical means, but these it will be unnecessary to describe in the following pages.

There are very few chemical means by which gold can be won from its ores, leaving aside the smelting process by which the gold is collected in the smelting furnace by means of, and in combination

Introduction. Ix

with, Other metals. The reason why gold is not easily won by chemical means is because it is a non* oxidizable metal, which maintains its purity, and is therefore mostly found in its native or metallic state. Owing to its rarity it has become of such great value, and the medium of interchange since the remotest times.

Its solvents are aqua regia, chlorine, and potas- sium cyanide. The application of such a corrosive agent as aqua regia for the treatment of ores on a large scale is out of the question, and of the chemical means at our command there remain only the last two. By exposing the gold ores, or the concentrated portion of the same, to the action of chlorine gas, the gold is converted into a soluble chloride of gold by the well-known chlorination methods ; and by the other method the gold is dissolved in solutions of cyanide of potassium.

That gold when in a fine state of division was soluble in cyanide of potassium, was already known in the middle ages, and the gilding of metals was carried out in those remote days by jewellers and alchemists, by the use of gold in cyanide solutions. Of course, gilding by means of fire was usually employed.

X Introduction.

Several scientific books, dating back to the begin* ning of this century, mention the solubility of gold in potassium cyanide solutions. The application of this solvent for the treatment of auriferous ores was first thought of and patented in the United States in 1867 ; and although the process was tried and experimented with by some eminent metallurgists of that country, no practical or commercial results were obtained.

Now, how is it — in spite of repeated failure in America and elsewhere — that within the last few years the cyanide process has come into such pre- eminence, and as such a grand success, before the metallurgical world ?

In answer to this question, I believe I am safe in stating that the ores of the Witwatersrand gold fields, where the cyanide process was first introduced, carry the gold in a pure and metallic state, in an extremely fine state of division (and even in the pyrites, the gold does not occur in chemical combination, but in a free state) ; and, therefore, all the conditions existed there to make the application of the cyanide process a perfect success. These facts are not only of interest to the metallurgist, but they should also throw some light on the geological features and conditions under which these peculiar conglomerate beds were formed.

INTRODUCTION. xi

The mines of this district, it should be said, will for ever have to acknowledge the immense services which the MacArthur-Forrest Company have rendered by developing and introducing the process, as they spent large sums to bring it into practical shape, and to demonstrate its commercial advantages. But for those exertions, enormous values would have remained practically unavailable, preventing many mines from working at a profit, and the production of these fields would not be what it is to-day by 50 per cent.

In this connection I can mention that out of a total gold production of 1,478,470 oz. in 1893, there were produced from the tailings 330,510 oz. by the cyanide process ; and in August, 1894, out of the monthly pro- duction of 174,077 oz., nearly 58,000 oz. were won by the cyanide process. [See Postscript on p. xiv.]

The great future and potentiality of development which may be anticipated for these gold fields is strikingly illustrated by the opinion of such an emi- nent authority as Mr. Hamilton Smith, who (in his report to the house of N. M. Rothschild & Sons), has stated that that portion of the Witwatersrand which lies between the Langlaagte Block B and the Glen- cairn Mine— or about ii miles in length, along the strike of the main reef series, — if exploited to a verti-

Xii INTRODUCTION*

cal depth of 3,000 ft., would yield in gold 215 million pounds sterling; while Bergrath Schmeisser (in his report made to the German Government), taking for estimate a depth of 800 metres, arrives at the yield of 208 million pounds, and for a depth of 1,200 metres, the yield of 349 millions. And when we consider that at least one-third of these prodi- gious amounts will be won by the cyanide pro- cess, one can hardly overestimate the importance of the work which the MacArthur-Forrest Company has done in bringing that process into its pre- sent position of prominence in the Witwatersrand gold fields. Certainly it cannot be insaid that they are entitled to a fair reward for their labours.

When I arrived in 1890 on the Witwatersrand gold fields, I undertook numerous experiments for the treat- ment of tailings by pan amalgamation, but with un- satisfactory results. The difiiculty I had to contend with was the formation of large quantities of iron amalgam in the pan. The quicksilver became dirty and floured, causing very heavy losses, and I had a costly process before me to separate the iron from the gold ; and although others followed me, and put up various devices for treating the tailings, it was only the advent of the cyanide process which solved this

Ic

INTRODUCTION. xiu

most difficult problem, as pan amalgamation would have been too expensive.

So far, the process has not achieved similar suc- cess in other countries, which to my mind proves that the ores on these fields contain the gold in a free condition ; and to further corroborate this view, I can state that at the Simmer and Jack mine the pyritic concentrates which are daily collected on the blankets yield by pan amalgamation 65 per cent. Of course these blanketings contain a considerable amount of free gold which escapes from the copper plates.

Within the last few months, Messrs. Siemens and Halske have successfully introduced on the same fields their patented process, which consists in precipi- tating the gold by electricity on sheets of lead ; and, owing to certain economical advantages, it may be anticipated that their process will prove a formidable rival to its predecessor. In the succeeding chapters will be found an account of the working details ol these processes, including some of the observations and researches of those gentlemen who have been most prominent in bringing them to their present state of perfection.

I must not conclude these remarks without express- ing my deep obligations to the several gentlemen —

Xiv INTRODUCTION.

whose names will be found duly recorded in the following pages — to whom I am indebted for much of the material and information embodied in the volume, and for the opportunities so freely afforded me, during my stay in the Witwatersrand gold fields, of making myself practically acquainted with the working of the cyanide process. It is not myself only but all who are interested — whether as metallurgists or investors — in the gold-mining industry, who have thus been laid imder obligation ; and as I have taken no small pains to put to good use the material and opportuni- ties which have been so freely placed at my disposal, I trust that the outcome of my efforts in the present volume will prove thoroughly acceptable to that numerous body.

M. Eissler.

37, Belsize Park, South Hampstead, London, N.W. December, 1894.

Postscript. — In 1897 the total output of the mines in the Witwatersrand district was 2,565,164 ozs., of which, on the basis of the returns for December, 1897, the amount recovered by the Cyanide and the Siemens processes would be equal to 32 per cent, of the whole.

(xv)

Contents.

Chapter I.

Pagk

ERECTION OF A CYANIDE PLANT i

Chapter 11. Extraction By Cyanide . 34

Chapter Iii. The Siemens-Halske Process 67

Chapter Iv. Particulars Of Operations At Various Works 84

Chapter V. The Chemistry Of The Cyanide Process . . 105

Appendix. Cyanide Plant At Mercur, Utah, U.S.A. . . .127

Index 139

(xvi)

List Of Illustrations.

Vig. Page

Frontispiece — Cyanide Plant of May Consolidated Gold Mining Company.

1. (Plate I.) Section of Princkss Works . . facing 4

2, 3. (Plate II.) Messrs. Butters and Mein's Automatic

Distributor . . . . . . . facing 8

4, 5. Do. Do. (Size A) . . .9

6. (Plate III.) Tailing Wheel, Vanner Room, and Cyanide

Vats AT THE Jumpers Mine facing 14

7. Staves cut to Circle . , 15

8. Construction of Filter Vats 16

9. (Plate IV.) Stone Foundation for Filter Vats 16

10. Solution Pipes 17

11,12. Butters' Discharge Lid 19

13, Zinc Precipitation Box 23

14. (Plate V.) Deeble*s United Cyanide and Pyrites

Works facing 26

15—24. Deeble's Patent Cyanide Pan 30—33

25. (Plate VI.) Zinc Precipitation Boxes at the Princess

Works facing 40

26. (Plate VII.) Electric Precipitation Boxes at the

Worcester Works facing 44

27. Rough Sketch of Apparatus for the Treatment of

Slimes at the Treasury Mine 49

28. 29. (Plates VIII. & IX.) Worcester Cyanide Plant facing 68 30,31,32. (Plate X.) Depositing Box ... ,,70

33. (Plate XI.) General View of the Simmer and Jack

Cyanide Plant facing 88

34. (Plate XII.) Simmer and Jack Filter Vats . „ 90

35. 36. (Plate XIII.) Simmer and Jack Extractor House „ 92

37. (Plate XIV.) Central Works of the Rand Central

Ore Reduction Company facing 94

38. (Plate XV.) Diagrammatic Sketch of Electric Pre-

cipitation Plant at Gibraltar Mine, N.S.W. facing 98 39—43. Cyanide Plant at Mercur, Utah, U.S.A. 129—16

The

Cyanide Process

For Thb

Extraction Of Gold.

Chapter I. Erection Of A Cyanide Plant.

Flanning the Works. — In the erection of a cyanide plant, before planning the same, some essential points have to be considered. Are the works to be erected to treat an old accu- mulated stock of tailings ? Or have they to be laid out to treat the tailings as they come from the battery? In many cases both these points have to be combined.

As tailings reservoirs are generally situated on the lowest point below the battery site, no provision exists below such dams for the erection of works which would permit of further dumping-ground and handling of the stufif by gravitation. At all events, the topographical conditions of Witwatersrand would allow such an arrangement only in rare instances. In localities where the fall of the ground below the reservoirs permits of the erection of the works, I would recommend this to be done, as it permits of the charging of tanks and their discharging by gravi- tation. In most cases the opposite course had to be resorted to on these fields ; the tailmgs from the old pits or reservoirs had to be hauled up-hill to the cyanide works, steam power being mostly used. The arrangement is simple enough, as the

B

2 Erection Of A Cyanide Plant.

dumping-cars are pulled up on an inclined trestle-work above the leaching tanks, and after discharging their contents they run back by gravitation, and are held back by the break of the haul- ing drum. In large works five to six trucks, holding 20 cubic feet each, are hauled up at a time. At every mine the mechanical arrangement for the filling of the tanks is different, depending on local conditions. Messrs. Fraser and Chalmers have intro- duced a system of mechanical haulage by means of endless wire ropes, which works very well.

To work old tailings by the cyanide process offers no difficulty to percolation, as they come to the works in the proper con- dition. They have been cleaned of the slimes by the natural system of concentration, which takes place in the collecting reservoirs. It is very interesting to stand at the discharge end of the launder carrying the tailings to the reservoir, and to see how the tailings arrange themselves according to the natural laws of gravitation, and are prepared here for subsequent treat- ment At the head of the tailing pit the coarser tailings accu- mulate, and near the dam the finest, and also slimes. The overflow from the first reservoir is collected in a second reser- voir, where the slimy, clayey residue accumulates, which, strange to say, is as rich, and even richer, than the tailings in the first reservoir.

To lay out plans for an accumulated stock of tailings offers no great difficulty, providing there is near by a sloping ground permitting of the discharging from the leaching tanks and their dumping by gravitation. If the country is flat, the re-worked tailings will have to be hauled up an incline again, and then dumped. On a flat site it will be necessary to place the leach- ing vats on masonry sufficiently high to give room for discharge, and gradient for the flow of the leaching solutions to the preci- pitation boxes.

When works are erected to treat tailings which are discharged from the battery, important appliances have to be resorted to, to prepare them for the cyanide treatment, and before they are collected in the leaching vats, owing to the physical condition of the powdered ore.

Slimes. 3

The discharge launder which carries the tailings from the battery to the cyanide works should have a grade of at least 3 ft. 6 in. in the loo ft to ensure a good flow. In a flat country where no grade exists, the tailings should be elevated by means of bucket wheels to the proper height. From the experience gained on these fields, tailing pumps have not given satisfaction ; it may be that they were not properly constructed, as I am told that in Australia they are in various places in successful opera- tion. There are on several mines here large tailing wheels in use, and I should consider them the best way of elevating tailings, as they require very little attention and repairs when properly constructed and set.

Supposing that we desire to erect a cyanide plant directly behind a battery, the following grade would be required for doing the whole work by gravitation. Supposing the plant to be located loo ft. from the battery —

Feet. Inches. The grade for the discharge launders will require . 3 6

Masonry for settling tanks 6 6

Settling tanks 10 o

Masonry for leaching tanks 6 6

Settling tanks 10 o

Precipitating boxes and grade for outflow pipes ..60

Total grade . . 42 6

To this could be added from 6 to 10 ft. of grade for the storage tanks holding the cyanide solutions, wash and alkaline waters, but these are differently placed, and a lack of further grade would present no difficulties, as will be explained later on.

Fig. I (Plate I.), showing a section of the cyanide plant of the Princess works, illustrates such a mode of arrangement as is here described.

Slimes. —The conglomerates on these fields, after stamping contain a very large percentage of slimes. By " slimes " are understood the very fine particles of talcose and clayey material mixed widi the very fine grains of quartz, iron oxides, and sulphides. If the whole of this fine material be allowed to collect

4 Erection Of A Cyanide Plant.

with the coarser grains, the percolation of fluids through the mass Ijecomes impossible, and, therefore, mechanical means have to be adopted, aiming at a separation of the slimes from the coarser material.

Two methods have been introduced, aiming at the elimina- tion of the slimes — the one by direct filling being the system introduced by Mr. Henning Jennings, the well-known mining engineer ; and the other the intermediate filling adopted by Mr. Charles Butters and Captain Mein, the manager of the Robinn mine.

I will take occasion to remark here that the appliances for the plant were materially changed by Mr. Charles Butters, who has done a great deal for the advancement of the metallurgical treatment of the ores on these fields, as he has introduced a great many practical details, all tending to lessen the cost of the process, and I consider it a very pleasant duty on my part to express my appreciation of his labours and of the good he has done by his work.

The elimination of the slimes has an economic bearing on the gold-mining industry of these fields, when it is considered that at least 30 per cent, of the Witwatersrand ores, after crushing, pass away into slime pits : therefore, with a produc- tion of 250,000 tons of ore per month (which was the scale of production at the time these pages were first written), more than 75,000 tons would go month by month into the same pits. Assuming for the Witwatersrand a production of 300,000 ounces of gold (or one million sterling) monthly, if we take the average value of the slimes all around at only 5 . per ton, this would represent nearly 20,000 ounces of gold going into the pits monthly.*

Up to the present no cheap method has been devised to deal with the slimes, so as to win the gold from them at a profit. The question of treating the slimes successfully is simply a mechanical one, as there is no chemical difficulty : on the con- trary, the solution of the gold can be easily effected, owing to the fine state of division in which the gold exists in the slimes. See Postscript, p. xiv, ante.

Effect Of Stamping On The Ore. 5

My impression is that the eventual solution of the problem will be a system of agitation in vats fitted with mechanical stirrers, and after agitation and settling the solution will be decanted and a weaker solution added again, again stirred and decanted, till the last wash waters will only show traces of gold. This method will become applicable owing to the introduction of the Siemens and Halske process, which effects the precipitation of the gold from the extremely weak solutions which will have to be used. The successful carrying out of this procedure will add largely to the gold production of the Rand.

The Eflfect of Stamping on the Ore. — When ores are stamped in a battery, the resulting product is very uneven, and this is one of the recognised disadvantages of the batteries in use in all gold countries.

The majority of the mines on these fields crush fine, using mostly screens of 900 mesh per square inch, and about 50 per cent, of the ore is converted into slimes. To illustrate the fineness to which stamping reduces the ore I may mention that Mr. Williams, the metallurgist of the Crown Reef mine, in- formed me that the slimes which flow into their pits do not settle in 24 hours, and that if the overflow is taken and allowed to settle, these second settlings will assay 6 . per ton, being richer, therefore, than the first settlings. He further stated that after the water leaves the second slime pit it still holds in suspension 2 per cent, of the total gold contained in the ore as it comes from the mine. This is the result from careful experi- ments and trials. In regard to the difficulty of settling the fine slimes which remain in suspension, it was ascertained at the Paarl Central works, that the overflow from a large tailing dam carried 8 per cent, of the rock crushed in suspension. As the water on these fields does not run to waste, but is used over and over again in the batteries, there is really no loss incurred, as the suspended slimes eventually settle, the water only retaining a certain constant portion.

To give a further illustration of the effects of stamping on the Witwatersrand conglomerates, I will furnish some figures

6 Erection Of A Cyanide Plant.

communicated to me by Mr. Bettel, a gentleman whose name has also been closely connected with the cyanide process.

40 lbs. of tailings were caught at a battery in a tub, and at least 30 per cent passed away as slimes, the ore coming through a 900 mesh screen. After dr3dng, it was sifted through a screen of 1,600 mesh per square inch, and there remained on the sieve i'85 per cent, (i)

It passed afterwards through three sieves as follows :

3,600 mesh and there remained on the sieve 27*93 Pr cent. (2)

14,400 „ „ „ „ T10 „ (4)

The sands passing the last sieve were panned, and

A. Remaining in the dish represented . . 1 1 -80 per cent. (5)

B. Finest sand panned away . . .22*34 „ (6)

C. Slimes collected from the panning water . 7*64 „ (7)

Total . . 100-00 Each of these grades was assayed, with the following results —

(I) (2) (3) (4) (5) (6) (7)

3*03 .

per ton

4*00 „

4*40 ,,

6-30 „

2-85 „

2-85 „

These figures are very instructive, as they show how very fine the ore is crushed in a battery, and that the material, after passing through a 120-mesh sieve per linear inch, can be concentrated, and will yield a concentrate of over double the value of the sands washed away. No doubt the practical part of these figures will be very soon appreciated on these fields when concentration will receive closer attention than it does at present.

After this notice of the important part which the slimes play in the metallurgy of the Rand gold fields, I will revert again to the planning of the works.

The Plant. 7

The Plant. — The main features of a cyanide plant are the settling vats, the filter vats, the solution storage tanks, and the predpitation-boxes.

The filter vats are either made of timber, or are brick vats lined with cement. At the Langlaagte £state and Gold Mining Company, circular excavations were made in the rocky ground, lined with bricks and cemented, forming tanks 40 ft. in diameter and 10 ft. deep, each holding 400 tons of tailings. Filter vats made of timber will last for years, as contact with cyanide solutions does not seem to destroy it.

I have not been able to obtain the cost of a plant con- structed in masonry, but I should consider it more expensive than the timber plant. Where wooden tanks are in use, they are placed in such a position that free access can be had to the bottoms in case of leakage.

The number of filtering vats required for a plant depends on the capacity of the battery, and the time it takes to treat a charge of ore. If we want to treat 1 00 tons of tailings daily, and it takes, say, 3 days to fill, leach, and discharge a vat, it will require 4 leaching vats of 100 tons capacity each (dimen- sions 22 ft. in diameter, and 5 ft. high), but for safety one extra tank is added. The tendency on these fields is to construct a few large vats for the plant, instead of a large number of small ones. As long as shallow vats were employed, there was no difficulty in shovelling out the tailings over the sides; but with deep vats the bottom discharge was introduced by Mr. Butters.

The time allowed for the zinc process is generally between 60 and 84 hours, vats of 300 tons or 400 tons capacity being usually discharged and filled in one working day of 10 to 12 hours, giving four days as the total time required per vat with the zinc process. In the Siemens and Halske electrical process the minimum time occupied is 5 days. Where 2 days are occupied in charging and discharging, six vats should be jsrected.

Before the tailings go to the leaching tanks they have to be

8 Erection Of A Cyanide Plant.

freed of their slimes, and I shall now describe the two methods adopted for this purpose.

Intermediate Filling. — Messrs. Butters and Mein's Distributors. — The first attempts at intermediate filling were made by running the battery tailings to the centre of a circular vat, and allowing the overflow to take place at one point. This did not prove successful, because the sand piled up in a central conical heap, and the slimes settled in the deeper water around the sides of the tank. The next plan was to run the pulp into the vat through a series of stationary launders, delivering at several fixed points. This method improved the distribution, but the result was still unsatisfactory. Then, in order to give an uniform overflow at every point of the periphery of the vat, a circular trough was fixed round the top to collect the overflow and deliver it to a launder.

Each of these alterations was a step in the right direction, but the system of settling could not be considered successful until after the introduction of an automatic revolving distri- butor. This appliance consists of a central casting, with a vertical spindle a revolving in a footstep bearing b, which casting carried a conical hopper £ and a number of radial pipes c with bent ends, as shown in section in Figs. 2, 3 (Plate II.), and 4 and 5.

The distributor is fixed on an iron column in the centre of the vat. The bends at the end of the pipes cause the appa- ratus to revolve by the reaction of the pulp as it leaves the pipes. Each pipe has a different length, in order to distribute over a number of concentric circles. This also had its faults, as it was found that the slimes collected in narrow rings be- tween the outlets of each pipe, giving rings of clean sand alter- nately with rings of slime. The difficulty was overcome by attaching flattened nozzles to the ends of the pipes, causing the pulp to spread over a wider area, and also by increasing the number of pipes.

As is noticed by this plan, the arrangement is a hemi- spherical bowl from which radiate 8, 12, to 16 pieces of pipes

10 Erection Of A Cyanide Plant.

of various length, which is set in motion by the centrifugal action of the discharging water, something similar to a garden sprinkler, only that it revolves slowly. The bowl is covered with a coarse screen so as to prevent chips or leaves to enter and choke the pipes. The diameter of the discharge pipes is

It is necessary to fill the vat with clean water before admitting the pulp. If this is not done, the water is prac- tically stationary, and a constant settlement of slimes takes place until the vat is full and the overflow begins, in which case the tailings in the lower part of the vat will always be more slimy than those in the upper part. For the same reason it is essential that the overflow be continuous until the vat is full of sand ; for if any stoppage takes place slime settlement in excess occurs, and a complete layer of slime is formed across the vat which prevents the overlying sand from draining dry. Therefore, when the battery is stopped, an equal quantity of water should be supplied to the vat. When the pulp is admitted into the tank previously filled with water, the light lime remains in suspension and over- flows into the annular ring which surrounds the tank at the top, and fi-om the discharge opening is carried by a launder to the slime pit.

When the vat is filled with tailings, the outlet pipe below the filter is opened and the water allowed to drain ofl*, the draining taking about fifteen to twenty-four hours. When holes are dug down to the discharge doors, water again commences to flow from the outlet, consequently it is advantageous to dig these holes about six hours before the discharging.

One very important matter is the proper size of vat to be used for a given tonnage crushed in the battery. It is, of course, desirable to catch as large a quantity of slimes with the sands in the tailings as is possible without rendering the product unleachable. When the vats are too small they cany away too much fine sand with the slime ; and if they are too large they catch too much slime, which settles in excess. The great difficulty to overcome yet with these intermediate vats is

to get the last foot or two near the bottom properly drained, and if discharged and transferred to the leaching tanks in this wet condition, the excess of moisture dilutes the cyanide solution.

To facilitate and hasten the leaching various devices have been adopted. At the Princess works, where the ground is steep, the drainage pipe has been extended down to the reser- voir, thereby causing a natural suction. At the Simmer and Jack works the drainage pipe is connected with a steam exhaust acting like an ejector, so as to cause a vacuum below the filter, and thereby the rate of leaching is increased. At the Worcester works the vats catch from the crushed ore from 75 to 80 per cent, of good leachable tailings, containing 12 per cent, moisture after draining eighteen to twenty-four hours.

The following are the sizes of the intermediate vats erected at some of the works : —

Meyer and Charlton Gold Mining Company, treating 120 tons per day, has 4 vats, each 20 ft. in diameter and 8 foot staves.

Pioneer Gold Mining Company, treating 70 tons daily, has 2 vats, each 20 ft. in diameter, and 14 foot staves.

Worcester Gold Mining Company, treating 70 tons daily, has 2 vats, each 20 ft. in diameter, and 8 foot staves.

Princess Gold Mining Company, treating 85 tons daily, has 2 vats, each 20 ft. in diameter, and 7 foot staves.

The Robinson Gold Mining Company, treating 330 tons per day, has 6 vats, each 24 ft. in diameter, and 11 foot staves.

When all the pulp is running into one vat, only about 66 per cent of the crushed ore is caught, but the whole of this is clean sand, and drains sufficiently. If, however, the total pulp from the battery was run into 2 vats, about 80 per cent of the crushed ore, instead of 66 per cent., would be obtained from the distributing tank. After the water has been leached out, the ore is discharged through bottom discharges into trucks and taken to the leaching tanks. In some localities the distributing

12 Erection Of A Cyanide Plant.

tanks are on a higher level than the leaching tanks, and the trucks are then run by gravitation to the leaching tanks. At some works the distributing tanks are at a lower level than the leaching tanks, and then the trucks have to be hauled up by steam power.

The framework of the tram lines on which the trucks are hauled up to the leaching tanks rests inside the tanks and on the masonry foundation, and at large works there is generally a double line of rails on top of the tanks. The vats and storage tanks are in the open, and not covered by a building.

The following are the advantages of intermediate filling, as introduced by Mr. Charles Butters : —

1. It is claimed that, by means of Mr. Butters* distributor, from 75 to 80 per cent, of sands, both coarse and fine, with some slimes, are collected in the intermediate tanks, the bulk of the slimes escaping with the effluent water, which is practically free from sands.

2. The water is drained oflf as near as possible, and when the intermediate vat is discharged through the bottom discharges, the sands during the operation get thoroughly mixed up, thus being in the best condition for treatment by cyanide.

3. Oxidation of pyrites is very slight, so that very little cyanide will be consumed.

To an impartial observer it would appear that the system of intermediate filling would commend itself as the one which is more practical, as the tailings undergo, so to say, a special preparation for the subsequent lixiviation. The expense of transferring the tailings from the intermediate tank to the leaching tank is so slight that it cannot be considered as an important item.

The cost of charging tailings and discharging the residues has been brought down at the Robinson Mine to lod. per ton of 2,000 lbs., and generally stands in the accounts of other works at about is.

Messrs. Butters and Mein's distributors are constructed in three sizes according to the following particulars : —

Size A distributor has 8 distributing pipes, all of ij in.

Direct Filling. 1 3

diameter, and is the size of distributor used in batteries up to 30 stamps.

Size B has 12 distributing pipes — 6 of 2 in. diameter, 2 of I J in. diameter, and 4 of ij in. diameter — and is the size of distributor used on batteries of from 30 to 70 stamps.

Size C has 16 distributing pipes — 2 of 2 J in. diameter, and 14 of 2 in. diameter — and is the size of distributor used on batteries of from 60 to 120 stamps.

The above sizes of distributors have been calculated on the average crushings per stamp for the Rand.

Direct Filling. — This method, introduced at the works of the Heriot, City and Suburban, Crovm Reef, Paarl Central, and Geldenhuis Estate companies, consists in passing the pulp as it leaves the plates into a hydraulic separator a kind of crude spitzlutte. The pulp is here divided into two streams, one overflowing, carrying slimes with very fine sands ; the other consisting of coarse sands, some fine sands and slimes, which are conveyed by means of an india-rubber hose to the leaching tanks, in which one or more Kaffirs are employed to effect the even distribution of the pulp, by moving the hose about to different parts of the vat. The water passes off by adjustable gates fitted inside the vats, carrying with it fine sands, slimes, and some coarse sands. The advantages of the process are : —

1 . This method treats pyritic tailings with the minimum of oxidation, as they are not exposed to the action of the air from the time they leave the battery.

2. A second handling of the tailings before treatment is avoided.

3. A preliminary rough concentration, or rather classifica- tion of the coarser particles of the tailings is effected.

Great controversy has arisen regarding the advantages of direct filling as against intermediate filling, and according to Mr. Bettel the disadvantages of the process are : —

" I. The tailings pack tightly in the vat, and consequently do not drain completely, and a diffusion of the first cyanide

14 Erection Of A Cyanide Plant.

solution which is applied takes place at the commencement of leaching, causing loss of cyanide and gold. At the Crown Reef works I noticed that the distribution seemed to be pretty regular, and drainage can be assisted by means of exhaust pumps.

" 2. The distribution of the sands and slimes is not so even, and some sands escape treatment, being protected by imper- vious layers of slime, the cyanide naturally escaping by the paths of least resistance. In leaching tanks where an uneven distribution of slimes and sands takes place, the slimy portion will not drain off ; and on discharging such a tank, it is easily noticed that the streaks of slime are saturated with moisture and are still gold-bearing, whereas the sandy portion has the solution drained off. The importance of an even distribution and mixture of the pulp can hardly be estimated.

" 3. At most of the works where direct filling is introduced, square cement tanks are employed, and the discharging of these is not so practical as the wooden ones fitted with bottom discharges."

In Fig. 6 (Plate III.), the tailing wheel, vanner room, and cyanide vats at the Jumpers mine are shown in section.

The Filter or Leaching Tanks. — These are, in mosc instances, made circular, that form being the strongest. They are from 20 to 42 ft. in diameter, and from 8 to 14 ft. in height, and should be constructed of well-seasoned lumber, with staves 3 to 4 in. thick, haAng their inner and outer faces cut to correspond to the arc of circle of the tank, and their edges radial to this circle (Fig. 7). The staves are not tongued or grooved, the pressure of the hoops being sufficient, if the tank is well made, to make them perfectly tight. The staves should be at least i ft. longer than the inside depth of the tank, and gained from i J in. into the bottom timbers, with a chime of several inches.

The bottoms are made of 3 by 9 in. deals, tongued and grooved (Fig. 8), and put together with white lead, or litharge and glycerine. The hoops should be made by wrought-iron

The Filter Tanks. 1 5

rods from i to in. in diameter, according to the size of the tank, with threaded ends passing through wrought-iron lugs and tightened by hexagonal nuts. When the tanks are of large diameter these hoops are made in sections. The outside of the tanks can be painted in lead paint.

The bottoms of the vats rest on wooden beams 6 by 9 in., placed 18 in. apart. These beams are placed across the stone foundation, and rest in their turn on planks by 11 in. The planks are put between the stone foundation and the beams to merely ensure a per- fectly level surface.

The construction of these vats should not be entrusted to the hands of any other workmen than experienced coopers.

It is obvious that tanks

Wlliiilii;;///

used the result has been that

u ,j A f a' 'WliiiUi;;////

tions, and in every case where 1 1 1 1 1 ; / // /

the tanks settled, got out of plumb, and leakages occurred.

The filters are constructed

of wooden slats, li by 4 in., .'

set 12 in. apart, fastened to V

the bottom by wooden pins. fio. 7.-Staves Cux to . Grooves in. deep and 3 in.

wide are cut in a number of places in the bottom of these slats to allow a free passage of the solution along the bottom. On top of these slats are laid strips of wood i by i in., only i in. apart from each other, making openings i in. square. Between the ends of this wooden grating and the inside of the tank an annular space of about in. wide is left, which is partly filled by a strip of wood i in. thick, bent to the circle of the tank. Over this and the slats is placed cocoa-nut matting and burlap.

Erection Of A Cyanide Plant.

and held by a rope i in. in diameter, which is driven into the space remaining between the strips of wood and the staves of the tank. On top of the matting are laid again slats of wood I by 3 in., parallel one to the other, about 6 in. apart, their object being to protect the matting from being injured when shovelling the tailings through the man-holes into the trucks below.

The stone foundations are usually 6 ft. 6 in. high above the

Fig. 8.— Construction of Filter VAts.

level of the rails, and are composed of a series of walls closed at their ends, leaving one or two passages underneath for the trucks (Fig. 9, Plate IV.).

Each leaching vat has a separate drain pipe, i to 2 in. in diameter and these pipes are so arranged in the extractor house, as to lead the strong solution to the strong extractor box and the weak solution to the weak extractor box. In some works there is one main collecting drain pipe for strong solution and

Plate IV.

E

Fig. 0. — Stone Foundation for Filter Vats.

Ic

t/i.rCKt

The Solution

one for weak solution and the connections are shown in Fig. 10. By shutting the valve, a, leading to strong collecting pipe, and opening valve, b, leading to weak collecting pipe, the flow is regulated. Filtration is best assisted by causing a vacuum under the filter bed; by connecting the drain pipe with a steam pipe and passing a jet of steam through the same, a vacuum is created under the filter bed.

I should also mention that the best and cheapest method of discharging the tailings firom the leaching vats is to sluice them out from a side door, but for this purpose a stream of running water is required, which on these fields is not available.

Mr. Feldtmann* describes a system of discharging tailings

we AH SOLUrtOAf p/f>£

Fig. 10.— Solution Pipes.

from the leaching vats through a bottom discharge door into a launder, whence a stream of water carries the residues into the creek below. The discharge doors can also be made on the side of the vat when the residues are to be sluiced out.

The round wooden filter vats on these fields are discharged

♦ "Notes on Gold Extraction." Argus Printing and Publishing Com- pany, Johannesburg, 1894.

1 8 Erection Of A Cyanide Plant.

by bottom discharge doors, which are closed by means of Butters' discharge lids. According to the size of the vats, there are two, four, six or eight of these discharge openings to each vat. Figs. II, 12 show the arrangement. On the bottom side of the tank a cast-iron ring, a, is bolted to the cast-iron cylin- der, B, inside the tank. Inside the cylinder is the projecting lug, c, upon which rests the hanger, d, forming part of the screw, E ; the cast-iron cover, when placed in position, is simply fastened by the nut g, and, screwing the same firmly, the whole arrangement becomes water-tight. The faces of the ring and the cover should be planed, so as to make a good joint. There are other methods of closing the discharge openings. When a tank is to be filled, a clay luting is given inside the iron cylinder, and then the same is rammed full with tailings. When filling the tank with tailings, especially into deep vats a length of 3 to 4 ft. pipe is put over the discharge holes, and then the tailings are dumped in. It will be easily understood that in discharging a deep tank it facilitates the running of the tailings into an outlet if the same is within a few feet of the surface, instead of having to push them down 13 or 14 ft. by means of long poles. The cocks and valves should be of iron.

At some works, the vats are arranged in two tiers, one directly above the other. The upper vats are the collecting vats, fitted usually with Butters' distributors, and the lower ones are the leaching vats. Such an arrangement obviates the necessity of tramming the tailings from the collecting to the leaching vats. Whenever the collecting vats are drained, the discharge doors are opened, and the tailings fall by gravitation into the vats below. The discharge doors are opened from below, access being given by a light platform over the lower vats.

Great attention has to be paid to the foundations and framing for these two-tier plants. They are carried by stout timbers resting on masonry piers running the length of the planty and forming the necessary passages below the leaching vats for the running of the trucks ; and this framework has, of necessity, to be carried into the lower vats, where numerous

Pumps.

posts of 12 in. by 12 in., or 14 in. by 14 in., timber are fitted to cany the longitudinal and transverse beams on which the upper vats are placed. The vertical distance between the rims of the lower vats and the staves of the upper ones is about six feet, so as to give head room and access' to all parts of the plant.

Figs, ii, 12.— Butters* Discharge Lid (scale x in. s z ft.).

Pumps. — Several varieties of pumps are in use in the- Witwatersrand to raise the solution from the sumps to the lixiviation tanks, and to provide circulation if needed, centri- fiigal pumps being mostly used. Manchester donkey-pumps, and other steam and centrifugal pumps have also been used.

20 Erection Of A Cyanide Plant.

" In every case (say Messrs. Butters and Smart*), it is abso- lutely necessary that all parts in contact with the liquor should be of iron or steel, not of brass or gun-metal, as these latter are quickly attacked and corroded by the solution. The Man- chester and Cameron pumps have worked satisfactorily in small plants, owing principally to the fact that repaurs are few and simple, and that no skill is required in their use. Some direct- acting steam-pumps have given much trouble, as the steam- cylinders were too large in proportion to the water-cylinder for lifts of between lo and 15 ft. For large works centrifugal- pumps are to be recommended, especially when the solution has to be pumped on to the tailings, as this requires to be done as quickly as possible."

Pipes. — In the paper already quoted, Messrs. Butters and Smart remark that piping arrangements in early practice were of defective design and bad construction, but have now (they say) been reduced to great simplicity : — " One large pipe, connecting, with separate cocks, all the solution-tanks to all the leaching-vats (through a powerful pump when the former are below the latter), suffices to throw on all the solutions in succession. Separate drain-pipes from each leaching-vat are led to a convenient point where cocks are placed, and from which a short hose will reach to any of the zinc-boxes. These separate pipes are necessary because filtration from the vats is continuous for many hours, and the different solutions cannot be mixed. Separate pipes from the precipitation-boxes to the solution-tanks, with or without pumps, according to their relative levels, are provided. As an instance of simplifica- tion in more recent works it may be mentioned that in a 1 5,000- ton plant built ten months ago twenty cocks were sufficient, while in a 10,0 00- ton plant built two years ago there are over 250 cocks. Iron plug-cocks have been largely used, but are unsuitable because the caustic-potash, resulting from

" Plant for the Extraction of Gold by the Cyanide Process," by Charles Butters and Edgar Smart, Assoc.M.Inst.C.E., Proceedings of Inst.CE,., vol. cxxi. (Session 1894—95), Part III,

Pipes.

2!

the decomposition of potassium cyanide, dissolves the film of oil in which the plug turns, with consequent leakage. There- fore the only valve to be used with safety is that of the packed cased type, such as the well-known Peet valve."

The following (according to the same authorities) are suit- able sizes for the charging- and leaching-pipes :-—

Diameter of Vats.

Diameter of Charging-Pipes.

Diameter ot Leaching-Pipes.

Feet. 20—24

32—40

Inches.

Inches.

It

Stock Solution Tanks. — There are generally three solution tanks at each plant, built very much the same as the leaching tanks, with the exception that they have no filters, man-holes, &c. They are of different capacity, according to the size of the works, and are required to be of sufficient dimensions to store enough solution to keep the works going, without having to run any to waste.

Inside the tanks are gauges indicating the volume of solution. The stock solution tanks are usually 20 ft in diameter, and from 7 to 14 ft. in height. One is for strong, one for weak solution, and one for alkaline wash. Every foot in height in a 20 ft. tank represents 10 tons of solution of 2,000 lbs. per ton.

According to Messrs. Butters and Smart, '' it is found that a tankful of tailings, containing between 10 per cent, and 15 per cent, of moisture, requires about one-third of its volume of water to fill it, and solution-tanks of between one-third and two-thirds the capacity of each leaching-vat have been used, but they have usually been all the same size at each works. There is, however, room for improvement in this respect. The amount of weak solution in use is generally largely in excess of

22 Erection Of A Cyanide Plant. .

the strong solution or alkaline wash, and by a test-treatment of the tailings in each case, the size of the tanks for each solution can be approximately determined. For ordinary clean tailings of moderate acidity, and a plant with five vats, each of 600 tons capacity — suitable sizes, giving ample margin for the alkaline-wash, strong-solution and weak-solution tanks, are of 300 tons, 300 tons and 450 tons capacity respectively. For convenience 26 cubic feet of water is reckoned to the ton."

To calculate the cubic contents of a circular tank, the follow- ing formula is employed. Multiply the square of the radius (10*) with 3'i4, and the product by the height of tank (6 ft).

io X 3*14 X 6 1884 cubic feet.

1 cubic foot of water weighs 62*3 lbs. Therefore,

1884 X 62-3 1 17573 : 2000 5878 tons of water. If we desire to prepare a 0*3 per cent, stock solution, 35271 lbs. of cyanide will have to

be dissolved in it, making allowance for any impurity in the cyanide. After treatment, we find that the solution analyses only o*i6 per cent. Consequently, by multiplying

H7S73'2 X o'i6

1 88' 1 1 lbs. is left in the solution ; and

100 - '

to make the same up to 0*3 per cent., another 164*60 lbs. of cyanide will have to be added.

Zinc Precipitation Boxes.— These are made of i to

2 in. boards, and are oblong boxes of various dimensions in proportion to the quantity of solution passing through them. In large works, the boxes are 20 ft. or more in length, 3 ft. high, and 3 to 4 ft. wide. There are separate boxes for the strong and for the weak solutions. At most works there are four of these boxes, placed in what

Precipitation Box.

ft :

ft ? g:

is called the extractor-house, which also contains the machinery, pumps, furnaces, &c.

The precipitation-box is divided into several compartments by parti- tions and baffle boards, in such a way that the solution is forced to flow upward through the zinc shavings, which are held in trays several inches above the bottom of the troughs. Fig. 13 shows the construction of the troughs.

The first division has not got any zinc shavings in the same, as here the solution enters, and any sediment or fine slime which may have passed through the filter settles here. If any intermediary settling tanks are used, as at the Worcester works, this first compartment can be utilisedjilso hold zinc shavings. From the first compartment the solution flows over the- partition, and then down the space, and upward through the tray holding the zinc shavings. The baffle- board is held in place in the position shown in the drawing by being nailed fast to the sides, and reaches a few inches above the level of the solution. From this explanation it becomes clear that the solution has to pass its downward and upward course till the last partition is reached, and from here passes through a pipe to the col- lecting sump or tank.

The zinc box compartments are fitted with removable trays, made of

24 Erection Of A Cyanide Plant.

wooden frames supporting wire screen of J-inch mesh. The gold in the solution settles on the zinc as a brown coating, and which soon, as it accumulates in a finely powdered state, tails through the screens to the bottom of the trough. In the last partition of each box there is no zinc, biit the tray here is utilised to hold cyanide of potassium in lumps to make up its standard strength before pumping the same into the storage tanks.

Over the zinc there is placed a light wooden grating, and the whole trough can be covered by a strong wire netting to secure against theft, as the same can be kept under lock and key.

At some works a wooden launder, covered with a lid and also under lock and key, is attached longitudinally to the box, and from each compartment in the trough when a clean-up is made, a plug is withdrawn and the slimes which have accumu- lated in the bottom are washed through the launder on to a filter and collected. In most works the clean-up is made diflferently, as described later on. The zinc trays rest on cleats, several inches above the bottom, and have handles on the sides so that they can be easily lifted out when cleaning up. After passing through the precipitation-boxes the solution is pumped back to the storage tanks, and is used continuously and not run to waste. The dissolved zinc does not accumulate in the stock solution to a great degree, and is probably precipitated in the lixiviation tanks with the charges of firesh ore.

Messrs. Butters and Smart* give the subjoined figures for use in determining the sizes of zinc-boxes, in conjunction with the details of the works to be described. " The total amount of solution to be precipitated per ton of sand varies between f ton in large works treating 600 tons per day, and i ton in small works treating 100 tons per day. Eflfectiv precipitation is obtained with an average rate of flow of about 26 ft per hour through boxes of between 18 ft. and 24 ft. long, 2 ft 3 ins. deep, and 3 ft. wide. The capacity of the boxes may be deter- mined according to the following rule which has given good For reference to their paper, see footnote to p. 20, (tnte.

Precipitation On Charcoal. 25

results : for each ton (26 cubic feet) of solution to be precipi- tated per hour, 21 cubic feet are to be allowed in the zinc-box." Each compartment is so designed as to hold i cubic inch of zinc, for every ton per month treated.

Precipitation on Charcoal at Deeble's Cyanide Works, Bendigo. — When upon a visit to Bendigo (Victoria), in 1897, I was much interested in the operations carried out by Mr. Deeble in treating the residues from the various chlorination works — containing several . per ton — and such tailings as he was able to collect at various mines containing sufficient gold to leave a profit.

He treats the material by agitation in round vats constructed of brick and cement. The appended specification for a patent taken out by him gives a full description of his mechanical appliances ; and the photograph reproduced in Fig. 14 (Plate V.) explains the arrangement, which, it will be seen, requires very little fall. There seemed to be no difficulty in decanting the clear solution after settling.

I elicited from Mr. Deeble the following information : —

(i) The strength of solution is 0*06 to o-o8.

(2) The volume of solution per ton of ore is 78 gallons — that is to say, 39 gallons of solution per ton of ore constitute the first solution, which dissolves the gold ; and 39 gallons of weak solution per ton of ore is necessary for a second wash of the ore, in order to draw off the whole of the gold-bearing solution.

(3) Time of treatment. — Twenty-five tons of ore require forty-eight hours. The first twelve hours after being started are necessary for the solution to dissolve the gold. After the first solution is drawn oflf a weak solution is put on; then re-agitate, draw oflf second solution, and empty vats.

(4) Number of filters employed. — The solution, after being drawn from the vats, is run into a receiver, and firom this receiver a series of twelve pipes convey the solution over the filters, which are in tiers of three, so that each jet of solution passes through three separate filters. In cleaning up these

oogle

26 Erection Of A Cyanide Plant.

filters each month the top tier only is taken for burning, and the next tier takes its place ; the third tier is filled from a large box filled with charcoal, into which the solution runs after leaving the third tier pf filters.

is) Weight of charcoal employed. — About i ton a month, with the plant putting through 450 tons a week.

(6) The quantity of gold recovered from the quantity of charcoal burnt. — This, of course, varies according to circum- stances, and depends, further, on the value of the tailings, the practical result reported by Mr. Deeble being that, from 22,737 tons treated, he had obtained a yield of 6,317 ozs. 4 . of gold, or an average of about 300 ozs. per month. This would represent approximately the quantity of gold recovered from a ton of charcoal.

(7) Tests regarding charcoal precipitation, which are as follows: — The solution, before coming in contact with the charcoal filters, contained 1-692 grains of gold per gallon of solution, and after leaving the first filter a gallon of solution contained 0254 grains of gold. From this it will be observed that the first filter abstracted 85 per cent, of the gold contents of the solution, and, after passing the third filter, the solution contained cos grains of gold per gallon. This would repre- sent a total extraction, after passing three filters, of 97 per cent, j but the remaining 3 per cent, is not altogether lost, because, after leaving the third filter, the solution runs into the large box already mentioned, and from there into a sump, and is pumped back again and used as a solution, with the required addition of cyanide.

Deeble's Patent Cyanide Pan. — Many endeavours have been made to invent a more efficient and rapid method of leaching than by the use of stationary vats, the object being to get the ore quickly and thoroughly mixed up with the leach- ing solution. The specification taken out by Mr. Deeble, for " an improved machine for use in the extraction of gold from auriferous material by the aid of chemical solvents,'* will in- dicate the general lines upon which such machines are con-

DEEBLE'S PATENT CYANIDE PAN. 2g

structed, and will illustrate a machine which is not unworthy of being brought to the notice of my readers. It reads as follows : —

This invention has been devised in order to provide a machine for use in extraction of gold from auriferous material by the aid of chemical solvents in order to insure the particles of auriferous material being brought into intimate contact with the cyanide or other solvent solutions.

" It consists of a vat or pan to receive the auriferous material to be treated, having at or about its centre a vertical shaft or spindle with one or more agitators or stirrers attached to its lower end.

"Motion is imparted to this shaft or spindle by bevel gearing or other convenient mechanical contrivances, and means are provided for reversing the rotation and controlling the speed of the agitators, as well as for raising or lowering the agitator shaft or spindle.

" These means may consist of a screw-threaded lifting rod fitted with a correspondingly threaded bevel wheel in gear, with a bevel pinion fitted with a crank handle whereby it may be rotated in the required direction, or if preferred a rack and pinion may be used for the purpose.

" The inner side of the wall of this vat or pan is provided with a series of projections which produce eddies or swirls in the material under treatment as it is carried round the vat or pan. " In order to drain or draw oflf the gold-bearing solvent firom said vat it is provided with a vertically sliding valve, or, if preferred, a pipe or pipes might be used for this purpose.

"A waste discharge valve may also be provided in the lower part of the vat or pan for the purpose of enabling the waste material to be sluiced therefrom after the gold has been dissolved and the gold-bearing solution has been drawn off through the valve above refened to."

The accompanying illustrations of the construction and arrangement of Mr. Deeble's machine are also reproduced from his specification.

Fig. 1 5 represents a vertical central section of the machine ;

Erection Of A Cyanide Plant.

Fig. 1 6 a plan of part of the driving gear ; Fig. 17a side elevation of the machine; Fig. 18 a sectional elevation ; and Fig. 19 a plan of a vat drawn to a smaller scale. Fig. 20 is a side eleva- tion of the vertically sliding valve above referred to ; Fig. 21 is

Fig. 15.

a plan thereof; Fig. 22 is a horizontal section on line 7-7 in Fig. 20 ; Fig. 23 is a vertical central section on line 8-8, Fig. 20 ; and Fig. 24 is a side elevation of the said valve. In all the illustrations the same letters of reference indicate the same or correspond- ing parts.

The specification proceeds as follows : — "a represents a vat to contain the auriferous material to be treated, whilst b represents the vertical spindle which is provided in or about the centre of said vat, and which carries the agitators or stirrers, c, on its lower end. These latter consist by preference of projecting bars or arms, as shown, and they may be attached at their inner ends to a casting, r, on the lower end of a

Fio. x6.

Deeble'S Patent Cyanide Pan.

spindle, b, which is preferably made square, or otherwise sided in cross section, and is passed through a corresponding hole in a bevel wheel, d, which is supported in a casting, resting upon bearers or beams, extending across from side to side of the vat, a,

'' This construction admits of rotary motion being imparted

Fig. 17.

to said spindle, b, and yet allows of its being raised or lowered as required. For this latter purpose a screw-threaded rod, e, is connected to end of said spindle by a shackle, and is fitted with a corresponding threaded pinion, in gear with a second

Fig. 18.

pinion, having a crank handle, Steps, /, are provided, leading up to a platform, f, in a convenient position over the centre of the vat to enable the crank handle, to be operated. Rotary motion is imparted to the stirrers by the bevel gearing.

Erection Of A Cyantoe Plant.

,and a clutch, (see Fig. i6), to connect or disconnect said gearing as required, hh represent the projections which are provided upon the inside of the wall of the vat, a, for the pur- pose of imparting a series of eddies or swirls to the material in the vat, so as to more thoroughly bring the whole of such material in contact with the solvent solution. These pro- jections may be made integral with said vat, or may be attached thereto, as will be readily understood, i represents the ver-

FiG. 19.

tically sliding valve, which is used for drawing oflf the solvent solution from the vat, a. It is arranged to be slid up and down in a frame, /', by means of a screw-thread f, so that it can be opened more or less according to requirements.

"j (Kg 17) represents a waste discharge valve, which is fitted in the lower part of the vat, and enables the dSis therein to be discharged after it has been treated and the gold- bearing solution has been drawn ofif. If preferred, a series of discharge pipes might be used for this purpose, instead of the

DEEBLE'S PATENT CYANmE PAN.

Fig. 21.

Fig. aa.

valve stoppers or plugs being withdrawn from said pipes as required.

" Having now particularly described and ascertained the nature of my said invention, and in what manner the same is to be performed, I declare that what I can claim is —

"(i) The combination and arrange- ment of parts constituting the herein- described machine for use in the extraction of gold from auriferous material by the aid of chemical sol- vents, the whole being constructed and arranged substantially as and for the purpose specified and as illus- trated in the accompanying drawings.

(2) In a machine for use in the extraction of gold from auriferous material by the aid of chemical sol- vents, agitators, and stirrers, project- ing from a shaft or spindle (such as b) capable of being adjusted vertically and being rotated by suitable gearing substantially as and for the purposes specified and as illustrated in the accompanying drawings.

" (3) In a machine for use in the extraction of gold from auriferous material by the aid of chemical sol- vents, a vat having projections (such as h) around the inner side of the walls substantially as and for the purposes specified and as illustrated in the accompanying drawings.

" (4) In a machine for use in the extraction of gold from auriferous material by the aid of chemical solvents, a valve (such as I) together with means whereby it may be adjusted vertically in the side of vat substantially as and for the purpose specified and as illustrated in the accompanying drawing."

Fig. 23.

Fio. 24.

Chapter Ii. Extraction By Cyanide.

Synopsis of the Process. — On the Witwatersrand gold fields, the cyanide process has been mainly adapted to the re-treatment of the tailings. As the largest proportion — and amongst it the coarser particles of gold — have been extracted by previous plate amalgamation, the precious metal in the tailings is in a very fine state of division, and therefore amenable to cyanide treatment. It must be here remarked that the coarser the gold the longer it takes to dissolve it ; and it is recom- mendable, therefore, that all ores should be first submitted to plate amalgamation before submitting them to the cyanide treatment.

When the first cyanide works were erected on these fields the old accumulated stocks of tailings had to be dealt with. Owing to their long exposure to the atmosphere, changes had taken place in their chemical composition, which caused at the onset some difficulties, but these were soon overcome on the application of the proper remedies, which will be found described further on towards the close of this chapter on page 52. The ores which came from the upper levels, or the oxidized zone, always carry a small proportion of iron pinite, which, on exposure, becomes oxidized. It is only when fi-ee milling ore tailings are taken directly from the battery to the cyanide works that they do not contain any decomposition products and are in a proper condition for cyanide treatment.

The course of the treatment in its successive stages may be thus summarised : —

First Stage. — Passing an Alkaline Solution or caustic wash through the ore to the point of saturation, as the tailings always contain a certain amount of organic matter, add salts, &c. By running on this alkaline solution, a considerable saving is efifected in the consumption of cyanide of potassium, and the tailings get the advantage afterwards of the full strength of the strong solution. Strength of this first alkaline solution may be o'i5 per cent. KCy, and may contain 4 ozs. of caustic soda per ton of solution. Caustic soda dissolves out organic matter. Excess of lime destroys cyanide.

Second Stage. — Tlie Strong Cyanide Solution, — This solution varies in strength from 0-3 to 0*5 per cent, in KCy to suit the richness and nature of the tailings under treatment, and the pro- portion of solution to be run on should not be less than one- third the weight of tailings in the vats. When this solution is run on, though the first solution has drained out, the tailings still contain a considerable quantity of the former solution. This should be displaced by allowing the second solution, or strong solution, to drain down immediately the vats are filled with the strong solution, for say about two hours (according to the capacity of vats and nature of tailings), and when this has been displaced, sufficient strong solution should be run on to make up the required proportional amount.

The vat is now full of strong solution, which, in some cases, may be leached out immediately, though generally it is advisable to leave it in contact with the mass for a short period —say, three hours — to give the solution time to penetrate any lumps.

When this solution is leached out, or only a very small stream is coming away, it should be allowed to continue draining about four hours longer. During these last four hours air is taking the place of the solution, and the gold is in con- tact with a strong solution of KCy, in presence of oxygen, which produces a more rapid and effectual dissolving of the gold.

For proof : Take two watch glasses, fill both with cyanide

30 Extraction By Cyanide.

solution of the same strength. In one, place a piece of gold leaf on the surface of the solution, while in the other immerse the gold under the solution. The gold leaf on the surface in presence with air disappears rapidly and is completely destroyed, whilst the other dissolves very slowly.

Third Stae,— The Weak Solution.— Mx.tr the strong solution has been run on, we may safely assume that most of the gold is now in solution, and the object of the third operation is to wash out the dissolved gold. Therefore, after the strong solu- tion has been drained out, sufficient weak solution containing 0*15 per cent. KCy is run on till the total quantity of solution and washes represent 75 to 80 per cent, of the weight of the ore.

Fourth Stage.— The Water Wash.—Aittx the weak solution water wash is applied, and the quantity so applied is not to be less than 7 per cent, of the weight of the ore, and, indeed, more is necessary.

The Precipitation. — The cyanide solutions containing gold, as they flow from the leaching vats, are passed through one or more precipitation boxes. On the zinc shavings the gold is precipitated, and the quantity of solution flowing through each box must be properly regulated. The only satisfactory method of knowing if proper precipitation is taking place is by having the solution regularly assayed. This is an important point, as good results cannot be obtained if the solutions leaving the precipitation boxes are rich in gold.

It is essential for the man on the shift to know the strength of the solution at the different stages of leaching : first, that he may know into which one of the precipitation boxes the solu- tion is to be passed ; and then he must ascertain the strength of the solution leaving the precipitation boxes, in order to let them flow into their respective storage tanks.

The gold in solution increases and decreases in quantity much in the same way and same time as the cyanide in solution

The Precipitation. 37

increases or decreases. In the precipitation of the gold in the boxes it is important to keep a sufficient stock of zinc shavings in each compartment, and so to regulate the flow of solutions as not to incur danger of fine gold precipitate being carried away. The zinc shavings are prepared by turning down zinc discs on a lathe. The discs for this purpose may be cut out of No. 15 gauge metal, and may measure from 6 to 12 in. in dia- meter, a hole being punched in the centre for the mandriL It is usual to put bundles of twenty such discs on a mandril. The speed of the lathe may be anywhere from 150 to 350 revolutions per minute, and the shavings are turned off by hand with an ordinary carpenter's mortice chisel.

Precipitation of the gold varies somewhat with different classes of ore treated. The completeness of the precipitation appears to depend in a measure on a slight excess of cyanide of potassium being present in the solutions. Roughly speaking, it may be said that if solutions leaving the zinc boxes assay more than 2 . per ton, the precipitation is not as it should be. This may be owing to the paucity of zinc in the boxes, which should be instantly rectified ; or to too great a speed in the flow of the solution; or, in very exceptional cases, to insuffi- cient cyanide in the solutions.

In the case of some Lydenburg cupriferous ores it was found that, by making the solutions up to working strength before passing them through the zinc boxes, the result was that solutions which had before assayed several ounces were reduced to a few . per ton.*

The zinc shavings in the boxes may require replenishing every day to replace the amount consumed, or they may run a week at a time without requiring replenishing.

Having, by passage of the solutions through the zinc shavings, reconverted the gold into the metallic form, the process of collecting the metal and putting it into ingots will be another stage of operations, and this will be described presently under the dean-up.

The exposure of the solution to the zinc after the gold is. According to Mr. Feldtmann.

38 Extraction By. Cyanide.

precipitated can only result in a loss of zinc, and — what is of much greater importance — unnecessary decomposition and loss of cyanide.

The precipitation boxes are set at a slight grade, as shown in the plan. Mr. Philip Argall, M.R.I.A., says that no iron or metal, other than zinc, should be exposed to the solutions in boxes ; even the iron wire screens, used to support the filiform zinc, have been shown to cause an unnecessary consumption of zinc and cyanide, through the electro-chemical action induced between the metals.

The method of Electric Precipitation is given in Chapter III. p. 67, and Precipitation on Charcoal in Chapter I., p. 25.

Conditions which Influence Precipitation. — Mr. W. Bettel states that in the precipitation of various metals from cyanide solutions by zinc, die precipitation is influenced by the following conditions : —

(a) Different strengths of solvents.

{d) Impurities in solutions (which may be derived from ores) afifecting precipitation.

(c) Influence of rate of flow of solution through extractors, — past a given area of zinc — in gallons per minute.

(ii) Amount of gold in solution to be precipitated, and maximum precipitation from solutions of various richness.

(e) Effect of alkalinity, acidity, or neutrality of liquors on rate and percentage of precipitation of metals from cyanide solutions.

(/) Influence on the rate of precipitation of gold of such inert substances as carbonate of lime, clay, oxide of zinc, and coating of zinc, &c.

Other physical conditions of zinc, affecting precipitation of metals from cyanide solution.

Time taken by Treatment on Clean Tailings. — The method of treatment described below has been adopted at the Rand Central Ore Reduction Company's works, the figures having been kindly supplied to me by the working chemist, Mr. Blomfield :—

The Clean-Up. 39

1st. Alkaline Wash.

Filling with 0'i6 per cent. KCy. + 4 oz.

Caustic soda per ton solution . . . time 2 hours. I

2n<l. Strong Solution, I

Filling with 0*35 KCy. solution and contact . „ 5 „ I

Leaching „ 8 „

3rd. Weak Solution.

4 washes at one hour each . . . n 4

4 leachings at 4 hours each . . . . „ 16 „

4th. Water Wash,

Filling „ I „

Leaching , 7

This treatment was applied to . tailings, residues being

from 15 grs. to I dwt. per ton; J lb. of cyanide being the consumption per ton of ore.

On these fields it has been found that 12 to 24 hours' contact with the strong solution is sufficient to effect a solu- tion of the gold in the tailings. The subsequent treatment with weak solutions has the object to displace the gold solution and to wash out the already dissolved gold. The displacing of the weak solution by a final water wash tends to reduce the loss of dissolved gold to a minimum, as it displaces the weak gold solution, and the gold which is left in the tailings is gene- rally that still retained in the coarser quartz particles, from which it has not been liberated by stamping, and whatever gold has remained undissolved in the iron pyrites.

The time of treatment of each vat varies, and takes from 50 to 140 hours, according to circumstances, and the size of the tanks employed.

The Clean-up. — In making the clean-up the tray holding the zinc shavings is lifted out from the last compartment, and pulsated up and down in the solution, so that the fine particles of slime and zinc fall through the sieve and settle at the bottom

40 Extraction By Cyanide.

of the box.* The zinc shavings are taken out of the tray, which is placed on a rack above the box, so as to allow solution to drain back into the box. They are also rubbed in the water, to remove, as much as possible, all gold adhering to them.

The tray is turned over and brushed down so as to remove any gold adhering to it The zinc shavings get very hot on account of oxidation, and it will be noticed that steam arises from them. They should be exposed, therefore, as little as possible to the air. The solution of water in the zinc boxes is pumped into settling tanks, where it is allowed to stand for two weeks, so as to give the extremely fine particles of gold which are held in suspension time to settle. In pumping out the pre- cipitation boxes great care should be taken not to disturb the gold zinc slimes in the bottom. With this aim the water is pumped out to within 2 in. of the slimes, and the india- rubber suction hose moved into the second compartment, and so on.

The slimes are then pushed back with a scoop to one comer, and the supernatant liquor allowed to stand for a while and pumped away again into the settling tanks. The slimes are now scooped into enamelled iron buckets and discharged on to a fine sieve — say 900 mesh — and washed and rubbed into the gold clean-up tank.

After the water is settled in the clean-up tank it is syphoned or pumped off, and the precipitate (called gold slimes) is drawn off through the plug holes on to a calico or linen filter, or into a filter press.

The zinc shavings, some of which will be found to have quite a brown coating of gold, which cannot be removed even by rubbing, are again returned to the precipitation boxes, and fresh zinc put on top. The gold which sticks to the zinc will be recovered in the next clean-up.

Before starting this operation it is advisable to run into the box a sufficient amount of clean water to remove the cyanide solution, as the latter is generally found to be injurious to the workmen. The Kaffirs, however, put their arms and hands into it without being affected.

bO

o

.?

o d

Oq Qq O

o

d

o

o M o

n

d o

o d

The Clean-Up. 43

Illustrations of the precipitation rooms at the Princess and Worcester Works (Transvaal), respectively, are given in the appended Figs. 25 and 26 (Plates VI. and VII.).

After the gold slimes are sufficiently dry to be handled with a scoop, they are dried on an iron plate or on iron pots, and are then ready for roasting and smelting. The object of the roasting is to oxidise the greater portion of the zinc which, in the form of small chips and shavings, has fallen through the zinc box trays, so as to cause it to combine in the subsequent smelting with the fluxes and leave the bullion fairly fine. Oxidation by the aid of atmospheric air is sufficient, but a certain amount of the zinc oxide subsequently becomes reduced by the carbon of the plumbago-melting pots, and re-enters the bullion.

A good method of roasting has been found to be the addition of a little nitre, say about 3 to 10 per cent, to the precipitate. Mr. Feldtmann suggests that it is best applied as a strong solution before drying the precipitate, so that it gets equally mixed with the whole mass. In the subsequent roasting, the nitre not only assists by yielding up oxygen to the zinc, but to some extent also appears to flux the zinc oxide, forming zincate of potash, which is not so readily reduced as zinc oxide. When the precipitate is very sandy — owing to tailings coming through the filters — nitre roasting is not so successful, as it tends to cake. By the addition of nitre the tendency of the precipitate to dust on stirring up in the roasting furnace is minimised, the amount of flux required in smelting is reduced, and the result in bullion is better.

In roasting the precipitate care should be taken not to raise the temperature much above a dull red heat (to avoid partially fusing it to a pasty mass), and not to stir too violently, especially just at the commencement of the roast, or dusting and con- sequent loss is the result.

Mr. Butters, at the Rand Central Works, has a muffle-roast- ing furnace in which to dry and roast the slimes. The bottom of the furnace consists of a cast-iron pan, and the wet slimes are charged on this pan, and, when dry, a damper is closed.

Extraction By Cyanide.

which turns the flame through an opening in the fire-bridge, under the iron pan, and the slimes are carefully stirred to avoid dusting, and as much of the zinc as possible is driven off during the roasting.

The clean-up is generally made once a month.

Smelting of the Slimei. — The dried precipitate is now ready for the smelting process, and graphite crucibles are employed for the operation. The fluxes commonly used are bi-carbonate of soda, borax, and sand.

Examples of various fusing mixtures are given below, but it should be well understood. that any one of the fluxes may have to be increased or decreased according to the amount of impurities present : —

Precipitate Bi-carbonate soda

z. Clean

Precipitate.

30 lbs.

IS n

2. Veryzincy 3olbs.

3. Very sandy*

Precipitate.

30lbs.

20 „

Borax

8 ,,

12 ,,

10 „

Sand Fluor spar

5 „

S

2 „

Precipitate and fluxes are well mixed and charged into the plumbago crucibles. The smelting furnaces, which may be constructed to take two or three pots at a time, should be good ones, as the heat required for this first fusion is rather in excess of the ordinary gold-melting temperature. After the charges in the pots are run down, more of the mixture may be added from time to time — the whole of a charge, as given above, will go into two No. 35 crucibles — and everything being fused until perfectly liquid, the contents of the pots are poured into moulds. Conical-shaped moulds are the best suited for this work. The metal settles to the bottom, and, after cooling, may be turned out and freed from the slag by breaking off the latter with a hammer. The slags, which contain a large per- centage of silicate of zinc and soda, corrode the pots, and during smelting there is a heavy evolution of fumes of zinc

According to Mr. Feldtmann*s formula.

Treatment Of Gold Slimes. 47

oxide, causing, most likely, losses of gold. At the Robinson Mine condensing flues connect the furnaces with the chimney.

The several pieces of bullion thus obtained at one clean-up are, subsequently, remelted with borax and run together into one ingot This remelting should be done at as low a tem- perature as possible, so that the metal may solidify almost as soon as it is in the mould, otherwise liquidation results and it becomes exceedingly difficult to obtain anything like a repre- sentative sample of the bullion for assay. The slags, which generally contain a considerable amount of gold in beads, are crushed up and panned, or cradled, to obtain the metal.

The slags from this operation are difficult to re-smelt, which may be owing to the presence of carbon contained in the zinc At the Rand Central Ore Reduction Works Mr. Butters is erecting a small matting furnace, where he intends to smelt the slags with copper ore concentrates and collect the gold in a matte.

The fineness of the bullion resulting from the cyanide process ranges from 600 to 800.

Besides gold, silver, and zinc there is some lead in the ingots, as the zinc employed contains a certain per cent, of this metal, and also carbon, which also is found in the ingots.

The zinc supplied by the Vieille Montagne Company is the best and purest for the purposes of the process.

Treatment of Gk>ld Slimes at the Treasury Mine.* — The course of treatment here followed is illustrated in the accompanying diagram (Fig. 27).

The slime sediment from the zinc extractor boxes is placed in the wooden tank, t, where also the zinc shavings are washed through a coarse screen. The cyanide solution brought with the slimes and shavings from the boxes is then pumped out by means of the india-rubber hose pipes (shown at side of tank) attached to the pump motion, and is forced through the

These additional notes on the treatment of zinc box precipitate (slimes) by the cyanide process, from the pen of Mr. T. H. Leggett, are taken from the Transactions of the Institution of, Mining and Metallurgy,

48 Extraction By Cyanide.

press, leaving the slimes in the bottom of the tank in a thick muddy condition.

Abucketfiilofconcentratedsulphuricacidisthen poured upon the slimes, dissolving the zinc To prevent the fumes getting into the house the hood, h, is lowered on to the ta;nk, leading the fumes through the roof. When the acid is spent the hood is raised and another bucketful poured in, and so on until the zinc is dissolved, when the contents of the tank should be about neutral.

The boiler, b, is then filled with cold water, and the water heated to boiling by a steam pipe which enters from below. This boiling water is then run into the tank on to the slimes (by means of the other rubber hose pipe shown in sketch) until the tank is fiUed, the slimes meanwhile being agitated with a wooden stirrer.

The pumps then draw off this hot slimy solution and force it through the filter press, the solution being stirred all the time and more hot water being added as required, until the tub is empty. More hot water is then pumped through the filter to further wash the slime cakes, and to remove all acidity and zinc sulphate.

Air may now be pumped through to dry the filtrates, after which the press is unscrewed and the slime cakes removed. These are dried in a closed muffle and without stirring, so that absolutely no dust is made. The dried slimes are then melted into bars in the usual way.

A bi-monthly clean-up at the Treasury Mine furnishes the following data : —

Weight of dried slimes, io8 lbs. avoirdupois i,575 ozs. troy. Weight of sulphuric acid consumed, 234 lbs. avoirdupois.

! Borax, 5 parts. Soda, 3 parts. Precipitate, 10 parts. Slag produced . 48 lbs. or 31*1 lbs. per 1,000 ozs. of precipitate. Grold „ . 446*81 ozs. fine gold.

Silver „ . 55*00 „ silver.

Bullion produced

750-3 fine gold : 92-9 fine silver.

Treatment Of Acid Tailings. 5 1

Or 843*2 total fineness in gold and silver, while the solutions escaping from the filter press assayed 4 dwt. per ton in gold.

Four hundred gallons of hot water were used as wash water per 1,000 ozs. of precipitate.

It will be noticed that lbs. less of slag per 1,000 ozs. of precipitate were formed than at the Standard Mine, while the bullion is 100 points less fine, due chiefly to the amount of copper pyrites in the Rand ores.

An important point in the economy of this method is the fact that the cakes from the filter press are merely dried with- out stirring, whereas in the roasting with nitre method in vogue on the Rand, the loss due to stirring and the escaping fames is something the Rand metallurgists do not like to talk about.

The only new feature in this method of treatment of cyanide slimes is the writer's adaptation of the hot water boiler, but this has rendered successful a treatment that without this feature has given but indiflferent results. By means of the hot water washes pumped through the slimes at a pressure of 60 to 80 lbs. to the square inch, all the disagreeable and loss-pro- voking sulphate of zinc (in the subsequent melting) is thoroughly eliminated.

The total losses by this method are less than one-tenth of I per cent., while a much less quantity and a far cleaner slag is produced than by the nitre method in general use, together with a higher grade of bullion.

The Treatment of Acid Ores or Tailings. — Under acid tailings" are understood those tailings which contain the decomposition products of the iron pyrites. These products consist chiefly of fi'ee sulphuric acid and soluble metallic salts, such as proto-sulphate or per-sulphate of iron, or insoluble basic iron salts. All these substances are destructive to cyan- ide, forming with it compounds useless in the extraction of gold. The reactions which take place, when these salts are in the tailings, will be found explained in a subsequent chapter (Chapter V.) on the Chemistry of the Cyanide Process.

52 Extraction By Cyanide.

The ores on these fields at a comparatively shallow depth become very pyritic, but outside of iron pyrites the main reef series carries no other sulphur combinations, or, at least, in such small proportions as to have no practical importance.

The treatment of acid tailings also offers more difficulty than in the case of the " free ore tailings." The oxidation products of the pyrites have to be neutralized by the addition of alkalies or alkaline earths — either caustic soda or lime— with or without a preliminary water washing, to remove such soluble salts or "cyanicides " as may be present.

Caustic lime, in a powdered form, is generally added now on these fields. With very acid tailmgs — namely, pyritic tailings — which have been exposed for some time to the oxi- dizing influence of the atmosphere, as much as 2i lbs. per ton is added. With fresh tailings i lb. of lime per ton of tailings is sufficient. In some works the requisite quantity of lime is added to each car load of tailings as the same is hauled up to the leaching tanks. In other works 6 or lo tons of tailings are dumped into the leaching tank, and the same levelled ofif and the lime sprinkled over it. The practice of putting all the lime on top of the tailings after the vat is filled cannot be recommended : the lime forms a pasty mass, and the alka- linity does not penetrate through the total height of the ore, so that in deep tanks it will be found that the top layer, say for one-third of the height of the ore, will be neutralized, whereas two-thirds near the bottom will remain acid. The length of treatment varies, and later on in this volume (Chapter IV.) I am giving the modus operandi of the process at some of the principal works on the Witwatersrand.

Solution of Gk>ld in Accumulated and Other Slimes.* — When slimes are settled in dams or pits, certain reactions take place, which result in the formation of decomposition products corresponding with conditions obtaining. The for- mation of ferrous sulphide proceeds very rapidly after the

Extracts from a paper read by Mr. W. A. Caldecott before the Chemical and Metallurgical Society of South Africa, July 31, 1897.

Solution Of Gold In Slimes. 53

deposition of slimes in the dams, where they settle, forming a compact mass practically impervious to air and water. After a few days the more pyritic layers of slime contain as much as 0*25 per cent, of ferrous sulphide, soluble ferrous salts and free acid being almost nil. Certain gold ores naturally contain a large amount of ferrous sulphide in the form of magnetic iron pjnites (6 FeS, FeS2), copper pyrites (CuS, FeS), or arsenical pyrites (FeAsA).

During the treatment of slimes by cyanide the presence of finely divided ferrous sulphide causes abstraction of oxygen from the solutions, whereby the solution of the gold is pre- vented. Other ferrous compounds, such as ferrous hydrate (FeOgHj), react in the same way, as does the decomposing organic matter always present in a more or less degree in slime dams.

The obvious remedy for these difficulties is to supply oxygen artificially, either in the form of air delivered from a perforated pipe fixed near the bottom of the agitation vat con- taining slime pulp, or in the form of an oxidizing agent. The cheapest and most suitable of the latter has been found to be potassic permanganate.

The present slime plants of the Rand Central Ore Reduc- tion Company have employed aeration of all pulp in their dis- solving vats as part of their regular treatment since the latter part of last year, accelerated when this operation took too long or when much organic matter was present with lb. to i lb. of KMnO* per ton of dry slimes.

All plants now under construction for the treatment of accumulated slimes will be equipped with large air compressors delivering up to 1800 cubic feet of air per minute.

Old accumulated slimes are considerably more acid than accumulated tailings, and although in the stirring process all possible efficiency is got out of the lime employed to neutralize this acidity, yet consumptions of from 8 to 20 lbs. of lime per ton of dry slimes are by no means uncommon. The ferrous hydrate produced by the action of lime both on ferrous sulphate in the slimes and on ferrous sulphate formed by oxidation of

54 Extraction By Cyanide.

ferrous sulphide through aeration has a tendency to abstract oxygen from the solutions, and also to consume cyanide by converting it into ferro-cyanide. By a further aeration this ferrous hydrate is converted into innocuous and inert ferric hydrate. Hence in the treatment of accumulated slimes pre- limmary aeration, or its more costly equivalent in the form of an oxidizing agent, has lowered the cyanide consumption by 50 per cent. — saving which more than pays for the cost of aeration ; even with the acid slimes above mentioned the actual cyanide consumption is now only i lb. per ton of dry slimes, or less than the cost of the lime employed. Of this i lb. only about one-half is consumed by the slimes themselves, the remainder being discharged as dilute solution with the residues, or lost by exposure to the air, and by decomposition in the pre- cipitating boxes. In the treatment of most accumulated slimes the gold does not dissolve without aeration, so that, instead of merely accelerating the reaction, it becomes as necessary for their treatment as cyanide itself. It is probable that before long aeration, now regularly used for the first time on a working scale, will serve also as a means of agitation in place of mechanically-driven paddles at present employed; but this application of its use, as well as various other adaptations now under investigation, is still in the experimental stage.

At the present time the method of dealing with accumulated slimes differs only in the earlier stages from the treatment of fresh battery slimes, the subsequent operations of settling, decantation, precipitation, &c., being much the same. These earlier operations, which vary in detail according to the par- ticular slimes under treatment, are generally as follows : —

The truck loads of slimes are dumped into a small constant- discharge pulping vat, containing fast-running paddles. A pipe delivers a continual stream of weak solution at the bottom of the vat near one side, and at the top, on the other side, is an overflow launder, through which the slime pulp continually discharges, as the lumps become disintegrated, into the dis- solving vat. Here it is aerated whilst being stirred by air delivered from an air compressor through a perforated pipe

Treatment Of Battery Slimes. 55

fixed near the bottom of the vat, until sulphides are no longer apparent in the pulp, on applying the acid and lead acetate test This aeration may take from two to twenty hours, and varies with every charge. After the sulphides are oxidized an hour or two longer aeration is given to oxidize any ferrous hydrate remaining, and the cyanide is then added. This brings the solution up to from 0*005 per cent. KCy to 0*008 per cent. KCy, which, though containing only a couple of ounces of potassic cyanide to a ton, is found amply strong enough to dissolve the gold.

During the charging and subsequent aeration lime is also added to bring the alkali strength of the solution up to 0*006 per cent, to 0*010 per cent., as any higher percentage of lime is found to be unnecessary for expediting the subsequent settlement of the slimes. The agitation and aeration are con- tinued for a period of from two to five hours after the addition of cyanide, as experience shows, and the slime pulp containing all the gold in a state of solution is then transferred to another vat and diluted preliminary to settling. Any permanganate to be added is put in as soon as the dissolving vat is full, as it is found that but little of the ferrous iron present is oxidized, whilst much organic matter is present.

The Treatment of Battery Slimes. — Mr. T. R.

Williams has given an account* of the method he introduced at the Crown Reef mine, where the slimes, after they leave the tailing plant, are dealt with as follows : —

To the slimy water flowing into a launder sufficient lime in the form of milk of lime is added to precipitate the slimes, which it does in a flocculent form. Regularity in the feeding of this lime plays a very important part ; an excess is quite as bad as too little. After being mixed with the lime the slimy water passes through two settling pits, each of the following dimensions: 20 by 20 ft. and 10 ft. deep. Here, roughly speaking, 80 to 90 per cent, of the slimes are deposited ; the

♦ Transactions of the Chemical and Metallurgical Society of South Africa.

56 Extraction By Cyanide.

overflow from these pits passing to the larger pit, which is 40 ft by 40 ft. and 10 ft. deep, from which the water runs away prac- tically clear. These pits come down to a point forming an inverted pyramid, on the same principle as a spitzkasten. They are all separately connected with the suction of a centrifugal pump, which pumps the settled slimes into eight treatment tanks.

These pits get rid of fully 90 per cent, of the total water used in the mill, so there is still to pump into these tanks about 10 per cent, of the water with the slimes. The treatment tanks are each 32 ft. in diameter and 10 ft. deep, having a conical bottom. These tanks are divided into four series ; thus tanks A A are used for receiving the sludge or slimes, where they are allowed to settle, the effluent water being drawn off by slats on the side of the tanks, and then conducted by a pipe into the larger settling pit, so that, in the event of it being turbid, it will get a chance of further settlement ; whilst one tank is being filled, the other will have been completely settled, and all the efiluent water drawn off. The sludge from tanks a is then pumped into tanks b b, using a hose with a jet, through which cyanide solution is pumped to sluice it into the suction of a centrifugal pump, sufficient solution at the same time being added to fill these tanks; the strength of the solution is simultaneously increased so as to bring the whole up to o'oi per cent, of cyanide of potassium.

Practically 80 per cent, of the total gold in the slimes is dissolved in its passage through the pump. However, further agitation is given by suction from the conical bottom, and charging over the top of the tank, and also near the bottom through three arms, two having the same radius as the tank, whilst the other throws across the diameter of the tank. This gives very good agitation, and is kept up for a period of from one to two hours, when the slimes are allowed to settle and the clear solution drawn off through a series of side cocks. The solution could be also drawn off by syphon pipes. Having drawn off all the solution, the residue slimes are pumped into tanks cc, where they are further agitated with a very weak solution of cyanide and allowed to settle.

Treatment Of Battery Slimes. 57

These solutions do not pass through the precipitation boxes, but are used for the first treatment in transferring from tanks A to B, sufficient cyanide at the same time being added to bring it up to the strength required. The slimes are then pumped into tanks d d, with very weak solutions, settled in the usual way, this solution being used for transferring from tank b to e. By circulating the liquors in this way the solutions become more concentrated, and reduces the quantity of solution passing through the boxes to two tons per hour. The first 50 per cent. of the gold in solution can be deposited in a very small fraction of the time required for the last 10 per cent.

Tests made by Mr. Williams show that in the first three hours 53 per cent, of the gold was deposited, whereas it took 21 hours to render this solution gold free. Another ad- vantage of this treble treatment is that pumping over from tank to tank ensures a most thorough agitation, which is shown by the fact that it has increased the theoretical extraction firom about 70 per cent., when using single treatment, to over 83 per cent., the extra cost being only a firaction of a penny per ton. The clear solutions from the tanks b b are run into two clarify- ing tanks, which are 15 ft. in diameter by 5 ft. deep. Here the solution is allowed sufficient time to settle any slimes that may have come with it ; thus, whilst one tank is being filled the other is settled and passing through the precipitation boxes.

There are four precipitation boxes of the following dimen- sions : — 30 'ft. long, 6 ft wide, and 4 ft, 9 in. deep. The iron anodes are placed at 6-in. centres, between which is placed a double wire frame, carrying on each wire three lead sheets cut into strips. The solution always passes upwards through the compartments in the same manner as the well-known zinc process. From these boxes the solution flows into a sump 40 ft. in diameter and 1 2 ft. deep, from where it is used over and over again.

The cost of treatment is 3s. 9*o5d. per ton. The actual extraction is 60 per cent.*

♦ The treatment of slimes is also referred to at p. 60.

58 Extraction By Cyanide.

Concentration and Treatment of Concentrates. — The question whether it pays to concentrate the tailings before sending them to the cyanide works has not yet been definitely settled on the Rand gold fields. Considering that during the short period of treatment which the tailings undergo, the gold cannot all be extracted from the pyritic particles, it seems rational that a separation of the pyritic matter should be effected, and the concentrates treated separately.

The only mines where concentration previous to cyaniding has been carried out for any length of time are the Crown Reef and the Langlaagte Estate and Langlaagte Block B Mines, where the concentrates are treated separately with cyanide. At these mines I am informed that it pays to follow this method.

Mr. Williams, of the Crown Reef, collects his concentrates by a crude system of classification, which costs him lod. per ton of concentrates, whereas at the Langlaagte Estate Frue vanners are employed. The manager there informs me that the cost of cyaniding the concentrates amounts to only 17s. per ton.

I believe that when the question has been thoroughly studied it will be found that a system of proper concentration before cyaniding will pay. Up to the present (December ist, 1894) there is not a single mine on the Rand equipped with a proper concentrating plant, as such ores as exist there have to be classified before an attempt can be made to concentrate them. At the Langlaagte Estate, where 22,000 tons of ore are crushed monthly, they get on an average 350 tons of concentrates, or about li per cent. ; this quantity could be easily doubled by the application of a proper plant. Most mills have Frue vanners, but this excellent machine cannot do the work alone on these ores. I may mention that Mr. Rademacher, engineer of the Humboldt works at Cologne, is now engaged at the Langlaagte Block B upon experiments with the view of con- centrating the pyrites out of the slimy particles of ore. The results there will throw considerable light on the subject, and prove if I am right.

High extraction by cyanide can be obtained from concen-

Treatment Of Concentrates. 59

trates, only the process does not seem to act alike everywhere on the concentrates, and possibly the physical nature of the pyrites is not alike in every mine. In some cases it may become necessary to grind the pyrites, so as to liberate the gold, and make it more amenable to cyanide. It is stated that gold contained in pyrites is extracted more easily when the concentrates are mixed with sands. Clean concentrates, therefore, are more difficult to treat, as owing to their high specific gravities they " pack " and resist percolation.

At the Crown Reef mine they collect by means of 3 spitz- lutten, about 600 tons of concentrates a month, out of 17,000 tons of ore. A li-in. pipe leads the heavy sands and pyrites to settling tanks 30 ft. in diameter and 6 ft. high, which assay 23 . per ton. The main object of this rough concentration is to eliminate the coarse sands so as to submit them to a prolonged treatment with cyanide. The cost of this system of concentration is only j£2 per month — namely, the pumping of the water which is needed for the hydraulic classifier — as they are worked by an ascending stream of water. The con- centrates, or rather the classified material, by being collected in wooden vats, is constantly under water and does not get oxidized. The material collected is not very coarse, as in the battery they use 900-mesh screens.

From the storage tanks the material is taken to the leaching tanks, and submitted to the action of the solutions for 16 to 18 days. The strength of solutions employed is 0-25, o'l, and 0*05 per cent, respectively ; about i lb. of lime is added to each ton of concentrates. Strong solution is run on for 4 days or more, till its strength remains constant.

The residues from these concentrates assay i j- ., giving an extraction of 94 per cent. The consumption of cyanide with these concentrates is about 1 j- lbs. per ton. The weaker solutions are constantly circulated through the tanks till a high point of extraction is reached. The Witwatersrand ores carry from 3 to 4 per cent, of pyrites; and I believe the time of treatment could be lessened by employing a somewhat stronger solution.

6b EXTRACTION BY CYANIDE.

At the Robinson mine 3 per cent, of the total weight of the ore is caught on Frae vanners, as concentrates assaying 4 to 5 oz. Extensive experiments have been made with cyanide, but the results were not satisfactory, and conse- quently the concentrates are treated at this mine by chlorina- tion, which costs £$ per ton.

The tailings at the Robinson are elevated by means of a bucket-wheel, and passed through intermediate tanks, fitted with Butters and Mein's distributors; there are six intermediate tanks, and 30 per cent, of the ore goes into the slime pits.

The tailings, which go to the cyanide works, assay 6 j- . The slimes assay $i . ; and Captain Mein tells me that the residues, after cyanide treatment, assay only i dwt. This would mean a higher extraction than on any mine on the Rand. He finds a discrepancy of i to 2 per cent between the actual gold produced and that determined by assays.

The Treatment of Battery Slimes.* — Since the dis- covery of the cyanide process the aim of every metallurgist has been the recovery of the total amount of gold from the battery slimes. The chief difficulties have been mechanical ones, for while it was comparatively easy to leach out the gold from the coarse tailings, its total extraction from the fine clayey slimes still presented many difficulties. Many processes have been tried, but most of them have been failures ; some on account of their cost, and others on account of their incomplete extrac- tion of the gold.

One of the processes tried was the mixing of the slimes, with or without previous drying, with coarse sand. This method was a failure, both because the more or less clayey mixture was insufficiently exposed to the action of the atmo- spheric oxygen, and the gold was therefore not totally dissolved, and also because it was found impossible to completely drain away the gold solution from the tailings. An endeavour was

For the remarks under this head I am indebted to Mr. Ad. Goerz, the chairman of the Rand Central Ore Reduction Company. See also Mr. Williams's account, ante, page 55. — M. E. (1898).

Treatment Of Battery Slimes. 6 1

made to overcome these difficulties by the introduction of stir- ring vats and suction filters, but this so added to the cost as to render the method useless except for very rich ores.

One mine in the Lydenburg district mixes its slimes with sand and leaches the mixture. In this case the satisfactory nature of the results is probably due to the fact that the slimes are free-milling, and so do not require much oxygen.

Amongst other proposed methods are the use of a filter press for squeezing out the gold solution from the residue, the roasting of the slimes to render them leachable, and the wetting of them with a soap solution : but none of these have apparently given satisfaction.

The constant attention and study devoted to this question since 1895 has led to the devising of a process which fulfils the particular requirements of the ores of the Rand, and which is based on the following facts : —

1 . An infinitely dilute solution of cyanide of potassium readily dissolves gold if the slimes are agitated.

2. Hydrate of lime causes a rapid settlement of slimes or any solid matter that is suspended in water.

3. The Siemens electric process will precipitate gold equally well from weak or from strong solutions.

In carrying out the process referred to, the battery pulp, freed from all coarse sands, is passed over a large spitzkasten ; a stream of lime water is introduced in order to settle the slimes; and the condensed pulp, consisting of one part of slimes to four parts of water, is continually discharged through small holes in the bottom of the spitzkasten. The settlement is almost complete if the spitzkasten is large enough, and the overflowing water can be used again for battery purposes.

The condensed slimes are allowed to settle in a tank, and the clear water drained away. The pulp, which now contains one part of slimes to one and a-half of water, is ready for treat- ment, and is mixed in the same tank with a dilute solution of cyanide ('oi per cent, or -005 per cent.KCy) ; the mixture is then agitated for a few hours by means of four wooden arms attached to a vertical shaft revolving in the centre of the tank.

62 Extraction By Cyanide.

The size of the latter will vary according to the capacity of the plant, from 15 to 50 feet in diameter, and 5 to 15 feet deep.

When working old slimes, containing half-decomposed pyrites or any organic substance, the solvent power of the solution is largely increased by the use of an oxidizing agent. The cheapest agent is compressed air, which is just as efficient as ferricyanide, permanganate of potash, &c.

After a few hours' agitation the reaction is finished and the pulp is allowed to settle. The clear gold solution is decanted off by means of a flexible pipe attached to the bottom of the tank, where the settled slimes remain. A second solution may be added, and the process repeated so as to recover any gold possibly left with the first settlement. It is evident that the percentage of extraction depends mainly on the proportion of solid slimes to liquids used. On an average, 8 or 10 tons of solution are used to every ton of slimes.

The gold solution is treated by the well-known Siemens process, and the gold recovered fi'om the lead plates by cupel- lation or by the Parkes process.

The treatment of slimes represents not only the recovery of gold from this battery product, but also influences and improves all operations connected with it. The necessity of saving as much of the battery pulp as possible for the leach- ing process is avoided, consequently the sand can be treated in the purest condition, yielding the highest extraction. An accurate estimate of working expenses can be determined only if the plant is worked independently of the other metalluigical operations. This is not often done, on account of the great advantages which are ofifered by combination : the sand and slimes plant being erected close together require only one set of men and one manager. Again, the cyanide solution used for treatment of slimes, being very dilute, is equally suitable for the sand treatment. Consequently both cyanide and labour do not add to the total expenditure. Since slimes treatment has been introduced a large number of companies have adopted the process.

An increase of about 7 per cent, on the total output of gold

Adverse Conditions Of Cyanide Treatment. 63

on the Witwatersrand, due to the treatment of battery slimes, can be fairly anticipated.

Advarse Conditions whioh affect the Cyanide Treat- ment.* — I. The action of impurities and base metals in ores prevents solution of the gold and effects the decomposition of cyanide. This will require a complete study by analysis of the component parts of the ore.

It is necessary to examine the condition of gold in residual tailings or concentrates from the cyanide treatment.

{a) In this connection it has to be ascertained if a certain percentage of dissolved gold (the auro-potassic cyanide) has or has not been completely washed out.

{b) Also if a certain per cent, of gold is still present in coarse particles capable of being amalgamated.

{c) Also the percentage of gold encased in quartz, or pyrites, to which the solution has not had access.

{d) Also the occurrence of gold in lumps of slime after treatment by cyanide; the gold supposed to be precipitated by the action of iron salts producing a ferrous ferricyanide from auro-potassic cyanide and free cyanide, thus explain- ing some low extractions from weathered ore and concen- trates.

The presence of copper combinations is very detrimental to the success of the MacArthur-Forrest process. The ores of the Black Reef series, near Johannesburg, carry a small quan- tity of copper pyrites, and consequently the extraction of gold by cyanide does not give as good results as on the Main Reef series.

In this connection it is important to mention here a point which the inventors of the cyanide process have strongly emphasised — namely, the selective action of weak solutions of cyanide for gold in preference to other metals. It is a strange fact that, while gold is one of the most difficult of metals to dissolve in acids, it is extremely susceptible to the weakest solution of cyanide, when copper and other metals might

According to Mr. Bettel.

64 Extraction By Cyanide.

remain unafifected in it. The successful use of very weak solutions may have much wider and more important con- sequences.

It is well known that the MacArthur-Forrest process has failed in some mining districts in the treatment of complex ores ; and it has seemed to me not unlikely that when experi- menters with the cyanide process have met with difficulties such as this, they have increased the strength of their solutions, using more cyanide in the hope of overcoming the disturbing presence of the baser metals.

I shall presently give an account of the process introduced by Messrs. Siemens and Halske, who have adopted the diametrically opposite course ; they have taken weaker solutions. If they had confined themselves to the chemical precipitation of the gold upon zinc or some other metal, as in the MacArthur- Forrest patent, they would have failed, because a point arrives at which chemical precipitation, in exceedingly dilute solutions, becomes imperfect. But the introduction by Messrs. Siemens and Halske of the agency of electricity solved the problem at once. Now it is quite possible that, when the complex ores of other countries are treated with very weak solutions and by the electrolytic method, the weak solution will attack the gold in the ore, but may leave the other metals unattacked. There is a wide field open here for investigation and research.

The Practical Besults. — Before we begin the treatment of the tailings, we meet with an acknowledged loss of 25 to 30 per cent, of the gold which goes into the slimes. It is true this gold is not lost, for it is still there. It is still an asset which accumulates monthly by thousands of ounces in the slime-pits ; but in the present state of our science and knowledge it is unavailable, and, therefore, for present practical purposes, represents a loss which has to be reckoned with. It may be possible, and quite probable, that in other mining regions this loss may not occur, but I put some emphasis on the point, as investigators in other countries will do well to closely study the nature of the ore they are dealing with before erecting

Cost Of Treatment. 65

their plant, and thereby probably save unnecessary expense on that head, to say nothing of costly experiments.

To illustrate the question of extraction by an example, I will suppose that we are dealing with a Rand mine whose ores contain an average assay value of 16 . per ton. In this case, I shall be safe in saying that the average extraction in the battery by plate amalgamation ranges from 60 to 70 per cent, especially from the pjrritic ores ; and, assuming that the battery recovers 10 . per ton, or about 63 per cent., this would leave 6 . in the tailings. Assuming now that 1,000 tons tailings at 6 . go to the cyanide works, containing 6,000 ., 30 per cent, of these, or 300 tons at 6 ., containing 1,800 ., go to the slime-pit, leaving 700 tons of tailings at 6 ., containing 4,200 ., which go to the leaching-vats. Of these 4,200 ., 70 per cent, is recovered by cyanide, equal to 2,940 ., so that the amount recovered from the original 1,000 tons of tailings represents 50 per cent, in round figures, or 3 . out of the 6 . Add to this the 10 . won on the plates, and we have 13 ., or a total recovery of 82 per cent, of the gold contained in the ore, which, when the slime treatment shall have been solved, will be raised to about 90 per cent., which percentage can be considered highly satisfactory.

Exact figures as to the actual extraction of the gold from the various establishments are not easily obtainable, but from the information I have gathered I should consider that 85 per cent, would be the average.

Cost of Treatment. — This necessarily varies with every plant, and mainly depends (i) on the size of the plant; (2) on the facilities for handling the tailings in charging and dis- charging ; while (3) the principal item is the consumption of cyanide.

The consumption of cyanide, I should judge, varies from i to I J lbs. per ton of tailings treated. The commercial article, as sold in Johannesburg, costs 2s. per lb.

The consumption of zinc is about 0*4 to j- lb. per ton of

66 Extraction By Cyanide.

ore, and costs 4. per lb. in Johannesburg. At the Kleinfon- tein works the consumption of cyanide is only i lb. per ton. The consumption depends greatly on the washing, and the alkaline wash destroying organic matter. The cost, in large works treating above 10,000 tons and over, would be 4s. per ton ; in works treating 5 to 7,000 tons, 5s. to 6s. ; in works treating 3 to 4,000 tons, 6s. to 7s. ; while in smaller establishments the cost may even be higher.

The cost of zinc precipitation can be estimated at from i. to 3d. per ton.

The Robinson cyanide works treated, in 1893, 55,200 tons of tailings; in which was contained fine gold, 20180*06 ozs., and from which was extracted fine gold, 13872*66 ozs. ; ex- traction, 687 per cent; bullion returned, 17921*20 ozs.

The cost of cyaniding was as follows : —

Total.

£ s. Wages (whites and natives, including food) . 3,406 1 1

General stores, assay material, &c. . . 1,408 9

Fuel 1,204 19

Cyanide — 64,41 1 lbs. equal to i 16 lb. per ton 5,563 o

Zinc — 12,521 lbs. equal to 0*23 lb. per ton . 260 4

Contractor (filling and discharging vats) . 4,325 12

Royalty 3,768 i

d.

Cost per ton. £ s. d

Total . . . ;fi9,936 18 2 £07 2-68

Cost of Plant. — A plant to treat 3,000 tons monthly would cost ;£5,ooo; a plant to treat 5,000 tons monthly, ;;7,5oo ; a plant to treat 7,000 tons monthly, ;£io,ooo; and a plant to treat 16,000 tons monthly, ;i 8,000 : the figures being, of course, in each case approximate only.

Chapter Iii.

The Siemens' Halske Process.

The Electric Freoipitation of Gold from Cyanide Solutions. — This process had been in practical operation at the Worcester mine for several months, when I had the oppor- tunity of personally observing and testing the operations there. But I should mention that for much of the information contained in this chapter I am indebted to a lecture by Mr. A. Von Gemet, delivered before the Chemical and Metallurgical Society of South Africa. Great credit is to be given to Mr. Von Gernet for the ability which he has displayed in developing this new process on the gold fields.

Upon reference to the plans (Plates VIII. and IX.), it will be noticed that the works are similar to those in use for the Mac Arthur- Forrest process, the only alteration being in the extractor-house.

How the Process was Discovered. — In 1887 Dr. Siemens, the eminent electrician of Berlin, found that the gold anodes, used in electro-plating at his works, lost weight when standing in the cyanide liquor without any electric current passing through the bath. This fact, of course, corroborated the statements made by chemical authorities that gold is soluble in cyanides, and induced him to try the use of cyanide solutions for the extraction of gold from ores.

He found that the zinc method, as introduced by MacArthur- Forrest, only gave good results with strong solutions, while the electric precipitation was equally effective with either strong or

Ic

68 The Siemens-Halske Process.

weak solutions, and its efficiency was not lessened by the presence of caustic soda.

He subsequently introduced the process in Europe, Asia, and America, and during the year 1894 erected a plant capable of dealing with 3,000 tons of tailings per month at the Wor- cester mine, near Johannesburg.

The Action of the Electric Current on the Gold Solution. — The electric current decomposes a solution of a metallic salt, the metal being deposited on the negative pole, while the metalloid is liberated at the positive pole of the electrolytic cell. In a fixed time a given electric current will deposit a certain quantity of metal, which quantity varies for different metsUs in direct proportion to their electro-chemical equivalents. This law holds good only for solutions strong in metal, but with very dilute solutions, as in use in the cyanide process, the current does not find sufficient of the metaUic compound present at the electrodes, and consequently de- composition of water also takes place. For this reason, to make the efficiency of the precipitation as great as possible, constant diffusion of the solution is requisite.

In order, therefore, to create an artificial diffusion, a me- chanical movement of the solution is important, and the most economical and convenient way of effecting this is to allow a slow but steady flow through the precipitation-boxes. But it is still more important to give a very large surface to the electrodes. In fact, a better effect is obtained by doubling the number of plates than by increasing the current tenfold.

Why Mercury cannot be uaed as a Cathode. — Mer- cury cathodes are not practical, for the reason that such an enormous quantity of this costly metal would have to be employed that the recovery of the gold would become difficult

To precipitate 100 tons of cyanide solution containing 5 . of gold per ton of solution in 24 hours, about 24,000 square feet of mercury surface is required. If the bottoms of the predpitating-boxes were covered with mercury, it would be necessary to have it at least a quarter of an inch deep, to make

Conditions Metal Cathode Must Fulfil. 69

up for differences of level, and thus ensure that the whole sur- face should be covered. This requires over 200 cubic feet of mercury, weighing 80 tons. At the end of a month there would be 750 ozs. of gold in this enormous quantity of mercury ; and even straining it most carefully, the gold would be so finely diffused that I doubt whether much of the gold would remain in the straining filter ; besides, the practicability of such an operation is questionable. I do not consider the initial outlay of the mercury, and the large loss sure to result when handling such large masses of this metal.

The vertical position in which metallic plates can be placed has the great advantage of keeping the surface of the cathodes clean, as any solid matter entering the boxes in suspension sinks to the bottom of the latter, no obstruction being offered to its downward course. Sheets of solid metal (as copper), coated with mercury, have also been tried, but have been unsuccessful because the mercury, owing to the action of the current, will penetrate the copper and form a dry amalgam which does not adhere to the plates.

Conditions which the Metal Cathode must Fulfil —

1. The precipitated gold must adhere to it.

2. It must be capable of being rolled out into very thin sheets to avoid unnecessary expense.

3. It must be easy to recover the gold from it.

4. It must not be more electro-positive than the anode, in order to prevent return currents being generated when the depositing current is stopped.

The most suitable metal for the purpose is lead, rolled out in very thin sheets, and this is accordingly used in the Siemens- Halske process, meeting all the requirements of the case. The lead sheets are fastened in light wooden frames. There are three sheets of lead, 2 by 3 ft., in each frame, giving each frame a surface of 18 ft., and 87 frames, which are in each precipitating-box, will expose a surface of 1,566 square feet. Each frame holding three lead sheets at i lb. weight each, makes 261 lbs. of lead in each box. See Plate X. (p. 70).

70 The Siemens -Halske Process.

The Anode. — Not less important is the question of anodes. By the action of the current a metalloid is liberated at the positive electrode, and the latter, when a metal, begins to oxidize. Carbon could be used as an anode, but it will not withstand the action of the current, and soon crumbles into a fine powder, which decomposes cyanide. This finely-divided carbon is in suspension, and cannot be removed from the solu- tion by filtration. Zinc used as an anode forms a white pre- cipitate of ferro-cyanide of zinc by the reaction of zinc oxide upon ferro-cyanide, formed during the leaching. Similarly, iron anodes form Prussian blue by the reaction of oxid of iron and ferro-cyanide. In consequence of this reaction the amount of ferro-cyanide in the cyanide solution does not increase.

From the Prussian blue the cyanide can be recovered by dissolving it in caustic soda, then evaporating the solution, and finally smelting with potassium carbonate.

This last process has been carried out only on a small scale, about 50 lbs. at a time, but a nice clean cyanide of potassium is obtained. In the treatment of tailings this regeneration of cyanide is not of great importance ; but with concentrates, which decompose the solution with formation of ferro-cyanide, it will effect a considerable economy.

Electric Current required for Precipitation. — In

order to precipitate the gold from cyanide solutions only, a very weak current is required, that is to say, a density of about o*o6 ampere per square foot. With cathodes about ij in. apart, 7 volt is sufficient to produce this current strength. The advantages gained by using such a weak current are : —

1. The gold is deposited hard on the plates.

2. The iron anodes are preserved for a long time, as their waste is in proportion to the current strength. In a plant treating 3,000 tons per month, 1,080 lbs. of iron are destroyed in that period.

3. Little power is required. 746 watts equal i horse- power. A 3,000-ton plant requires 2,400 watts, equal, theo-

Advantages Of Electrical Precipitation. 71

retically, to horse-power, and actually requiring about 5 indicated horse-power.

Advantages of Eleotrioal Precipitation. — The most important feature of electrical precipitation is, that it operates on the solution quite independently of the amount of cyanide or caustic soda it contains. Precipitation by means of a chemical reaction is invariably more complete with a solution strong in cyanide than with a weak one, but with electricity is absolutely of no importance whatever. Therefore, in the treat- ment of tailings, very dilute solutions can be used, the only liibit being a sufficient amount of cyanide to dissolve the gold satisfactorily. Moreover, however acid the solution may be when entering the boxes, the precipitation takes place equally as well, the same amount of gold being recovered as from a neutral or alkaline solution. There are none of the (x>mplica- tions arising from the formation of lime and alumina and hy- drate of iron, which, under similar circumstances, sometimes occasion so much trouble in the zinc process.

A solution, containing 0*03 per cent, of cyanide, will dis- solve gold just as efifectively as a solution containing 3 per cent., provided a longer time is allowed for treatment. In the first case, the decomposition of cyanide in the tailings is much less than in the second, and a corresponding economy is effected. The moisture in the original tailings being usually aboot the- same as that contained in the residues, there is, as a rule, no> chance to give a large water wash after the cyanide treatment is finished. The residues discharged contain 10 to 15 per, cent, moisture, carrying about i to 0*05 per cent, cyanide solution when the zinc process is employed. This is equiva- lent to i lb. of cyanide per ton of tailings. This last may be reduced to tV lb. by using as weak wash a solution containing only 0*01 per cent, of cyanide, which strength will bcj perfectly suitable for electrical precipitation, though difficult to deal with by the chemical method.

Practical experience has taught us that, if we treat an ore containing copper with a strong, say 2 to 3 per cent., solution of

F

72 The Siemens-Halske Process.

cyanide, the potassic cyanide may be all decomposed, but if we apply a much weaker Bolution, say per cent., although the cyanide will be also decomposed, the same extraction of gold will be effected. So that, though in the former case the treat- ment may appear commercially impracticable, the proper carry- ing out of the process, by leaching with weak solutions, will prove to be both effectual and economical.

The presence of copper in the ores also may affect, to some extent, the second operation in the process, i.e., the precipita- tion of the gold.

Mr. Feldtmann mentions in this connection the result of some experiments made by him on some cupriferous ores at the Transvaal Company's plant near Lydenburg, as illustrating the selective affinity which weak cyanogen compounds possess for gold, that although there was sufficient copper mineral present to decompose a solution of potassic cyanide of over i per cent strength, still good results—/.., 70 per cent, extraction on 18 . ore — were obtained with J per cent, solutions. In this instance t'lere was reason to believe that the gold was dissolved as auric cyanide, instead of as auro-potassic cyanide, as usually formed.

As I have already intimated, it is my belief that the appli- cation of tlie extremely weak solutions, such as are utilised in this process, will lead to important results in the treatment of the so-called rebellious ores. But further investigation of the subject is required.

The vats hold 135 tons, and it takes 5 hours to fill them. The tailings first get an alkaline wash of 10 tons, after which wash 70 tons of strong solution of 0*05 to 0*08 per cent, is pumped on ; and afterwards 20 tons of weak solution, of coi per cent, strength is pumped on. The total quantity of solu- tion used is 100 tons, and it takes days to leach, filter, and discharge each vat.

Practical Working Results.— At the Worcester works, which are in charge of Mr. Oswell, there are in use 5 leaching- vats of 20 ft. diameter with 10 ft staves, each holding 3,100

Practical Working Results. 73

cubic ft., JSS toJs. One tank is discharged and filled every day. The strong solution used contains from 0*05 to 0*08 per cent, cyanide, and the weak washes o'oi per cent. The actual extraction of fine gold has averaged 70 per cent., while the con- sumption of cyanide has been i lb. per ton of tailings treated. The precipitation plant consists of four boxes, 20 x 8 ft wide and 4 ft. deep. Copper wires are fixed along the top of the sides of the boxes, and convey the current firom the dynamo to the electrodes. The anodes are iron plates, 7 ft long, 3 ft wide, J in. thick. They stand on wooden strips, placed on the bottom of the box, and are kept in vertical position by wooden strips fixed to its sides. In order to effect the circulation of the solution in passing through the box, some of the iron sheets rest right down on the bottom, while others are raised about I in. above the level of the solution, thus forming a series of compartments similar to those of a zinc precipitating-box, the difference being that the solution passes alternately up and down through successive compartments.

The iron sheets are covered with canvas to prevent short

circuit. The lead sheets are stretched between two iron wires,

fixed in a light wooden frame which is then suspended between

the iron plates. The boxes are kept locked, being opened once a

month for the purpose of a " clean-up," which is carried out in

the following manner : — The firames carrying lead cathodes are

taken out one at a time. The lead is removed and replaced

by a fresh sheet and the frame returned to the box, the whole

operation taking but a few minutes for each frame. By this

means the ordinary working is not interrupted at all, and the

cleaning out of the boxes, which is necessary in the zinc boxes,

is only required at long intervals. The lead, which contains

from 2 to 12 per cent of gold, is then smelted into bars and

cupelled. The gold is deposited on the lead sheets as a thin,

bright yellow film, which adheres firmly to the lead. The

consumption of lead at these works is 750 lbs. per month,

equal to id. per ton of tailings, and the working expenses for

treating 3,000 tons per month are as follows : —

The Siemens-Halske Process.

£ Filling and discharging leaching-vats 125 monthly

Cyanide

Lime

Is

Caustic soda

Lead

Iron . . .

White labour .

Native wages and food

Coal

Stores and general charges

0*10 per ton.

1*2

0*5

3 shillings

per ton.

At a great many works, where correct sampling and assaying is carried out, the results — that is, the gold won — agree pretty closely with the assay results. At the Worcester the following were the results for the month of August, 1894 : —

Ozs. . grs. If 350 tons tailings taken from settling vats containing . 443 i o gold 1*750 If M ty upper tailing dam containing 498 15 12 ,,

3,100 tons tailings which according to assay contain 3*10 „ residues „ „ „ „

Balance, showing the theoretical output or 74*6 per cent. 702 90,, Actual gold won, or 74*1 per cent 697 15 15 „

The tailings which were treated assayed from 6 to 8 . The tailings, or residues, after treatment, assayed from i to 2 . The solutions which leave the precipitation-boxes still contain some gold, and the analysis in this case showed that the strong solution contained 4 . 8 grs. per ton of solu- tion. The weak solution contained only 10 grs. p'i- f-'h)l'-

On an average, in these works, strong solutions carry from 4 to 5 ., and the weak from o to i dwt.

In taking samples of the residue, it should be borne in mind that the portion of the residues in which the largest proportion of unextracted gold is contained is near the bottom of the filter-vat, say the last 12 to 15 inches.

The best way to sample is to take a sampling iron and probe

Practical Working Results.

every carload as it goes to the works, and to do the like with every car as it leaves the works with the worked residues.

In the subjoined pages (77 to 83) will be found the de- tails of a working scheme for one week, as carried out at the Worcester works ; and to follow the method of treatment, I have chosen vat No. 4 as an example.

On the 2oth August, 1894, this vat, which holds 135 tons, has been filled with tailings. It takes about five hours to fill such a tank. One ton of tailings is equal to 27 cubic feet.

From 3.15 p.m. till 6.20 p.m. 10 tons of the alka- line wash are pumped into the vat .

At 8.10 p.m. strong solution is pumped on .

„ 3.30 a.m. „ „ „

2 1ST August, 1894. At 9 a.m. strong solution is pumped on

tt 3.40 a.m. „ „ „

22ND August, 1894. At 9.15 a.m. strong solution is pumped on „ 1.30 p.m. „ „ „

tt 'SQ tt It tt tt

„ 4 a.m. „ „ „

23RD August, 1894. At 7.30 a.m. strong solution is pumped on .

Total strong solution

At 1.50 p.m. weak solution is pumped on .

„ 1.25 a.m. „ „ „

10 tons

tons

tt

tt

5

tons

))

5

tt

tt

It

tons

It

5

tt

tt

tt

5 tons

70 tons

7 tons

tt

tt

Total weak solution

21 tons

76 The Siemens-Halske Process.

24TH August, 1894.

At 7.30 a.in* water wash is pumped on . . 6 tons

Total water wash . . . 11 tons

The tailings are then leached dry and discharged in the morning of the 25th August.

It takes, therefore, five and a-half days to leach and filter one tank, and the solutions consist of

1. Alkaline wash 10 tons

2. Strong solution . . . 70 „

3. Weak solution 21 „

Total quantity of solution loi tons

This is by working the Siemens and Halske process.

Scheme Of Working.

On

fi

ri

eq

a

d

S

s.

o 2 cq

d p -ti

,3 "S ft

'£ 9

Is

aw

ll ll l.l

N4 M

M

S

O

1i iS

o o o

o

Pl

o

O O O O

O ITi

'-

s -

a . .

-i

g5

5 " ""d

o "

2 "

S 3

fO LO

s s s

sa

bS

ddetf

dd

p @

§

H

t-i OvfOr

to

i r°

s ..

"*a

s - -

Q Si t

J

m

ar

a s s

aaa

SgSaa

"S

ft

Po2

p

crMTici

ONVO tii

Oo'vO i-i

m

N

ro

d

Js

Js

!z;

The Stemens-Halskk Process.

9

I"

a 'I

IS o

Is.

o

a-o

00*

c/3*jc/3*ic/3*:c/3*; .5.

I?

W "

O N

o

., I -§8 8 s;.a

S8

T3 'd'O I

S d s 8 xR 8 X

On

g

H Co

<5£

O O d

§

aaas

c4 diCvO

aaa

a a a

too O

o

aa

O Q

ba

a

d

8,

5 "

a 5 a a

dg (in

%-Bs.s,

O

iz;

Scheme Of Working.

C/3 w

A A

o

o

a . 8

-2?

iP

a

S .

c/3 O

S a . o

O

O C/3 W

o

O ' c/3

. o o 5p

Ii

.

d o d

9

tuB- —

S Q s a dart

o 8-

roOO W

t

§..

aas

a

S.i;?ft

O "

O roro

iM* OS

O

ONvd N

o

a s s

o

a a

o -

aaa

d dici

8o

The Siemens-Halske Process.

1§

Solution 'nt.K.Cy.

Solution nt.K.Cy.

I. Solu- per cent.

Solution cent. K.

I. Solu- per cent.

I. Solu- per cent.

I. Solu- per cent.

tion 0.054

5©

Tank No. 3. ed 0-044 per c€

Tank No. 3. ed 0-024 per cc

. Tank No. ntained 0*005

Tank No. i. ed 0008 per

. Tank No. ntained 0-016

ank No. led 0-024

ank No. led 003

.d

a

ctf d OS a

'A i§

#

too

M

g

to

ligs

ei

S-s§S

to

toio

|ti5=fS

Vo

!3

en

H

ss

Discharging and Filling . to 5.45 p.m. 10 ton

No. 3 tank, m. 5 tons. 9.25 p.m. m. 5 „

B

B Bb

r

to to

1"

t>.

lOtOiO

B B B

S cud

d drt*

cS

oo

a

- Is-

rotON

ONtON

OnW W

M

ir%

Scheme Of Working.

Os

tt o

§-2 2 2

r

bo

a

i

totOir>

sag

:r8a

t-l ONfO

i

a s s

c6 a.rt

s-sg,

I

00

o

a a

rt d

s 6 a

C3 d ctt

The Siehens-Halske Frocrss.

Sg

cap

sl al A

3 9 PS o'l (

C. OH Of5 Or™ O

H w

o

o

H to

M

Q

o o o o ,

a. a

H

a

s

C)

o

Bsa

adrt

N§ 1

iJ

M con;

s,

s a s

asBsf

d diPrjvO

o o

O

to to to

a a 6

Scheme Of Working.

Is

90}

c4 (3 cS (3 cS O

. b

p o "

bo I

bJ3

J

H

, a

a

o - -

aaa

o o

a a

u-)0 N On

a a

O O O T

O - -

a a a

cQ dd 2 ITS

So :r

isr

Ofi-ssa

ex o

O

Chapter Iv. Particulars Of Operations At Various Works.

Crown Beef Worka— At these works* the tailings undergo the following treatment by the direct process.

The 1 20 stamp battery crushes on an average 17,000 tons of ore monthly. Of this quantity 12,000 tons of tailings go to the cyanide works, which assay from 4 to . The slimes, amounting to 5,000 tons, assaying ., go to the reservoirs.

The tailings from the battery pass through 3 spitzlutten, one placed in front of the other, and here a classification of the tailings is eflfected. From the spitzlutten the stream is divided and passed to the separators or spitzkasten, of which there are four, placed in front of the leaching-tanks. At the bottom is a T-piece, with a nozzle screwed on either side, and a rubber hose, the discharge of which can be regulated by a clamp, so that the stream can be diverted into one or the other of the tanks. Mr. Williams claims that he eliminates more slimes by this system of direct filling, and that the tailings Which he retains are richer. He claims that 70 tons of his tailings con- tain more gold than 75 tons of tailings saved by the interme- diary tanks, and that is the reason why he discarded them. According to these figures he saves the treatment of. 855 tons of tailings every month.

From the spitzkasten the tailings discharge into six large cement filter-tanks, each 40 ft. from front to back, 34 ft. wide, and 10 ft. high, measured inside the tank. At the bottom the

For the information in regard to these works I am indebted to Mr. Williams, the chemist in charge. — M. E.

Crown Reef Works. 85

tanks are narrower, as the inside is built of sloping walls. The side walls are 4 ft. thick at the bottom and 18 in. at the top. Their capacity, when full, is about 500 tons of tailings.

The sides and walls are of brickwork set in hydraulic lime, and plastered with cement plaster. The bottom is of cement concrete at least 12 in. thick. The concrete is made up of three parts of treated tailings sifted, five parts broken stone down to 2 in. cube, and one part of Portland cement. It is mixed, put in place, rammed, and levelled up with a grout mixed two of sand and one of cement The bottom of the cement-tank slopes to a central gutter, having a 2-in. fall from the front and back to the centre, where a gutter, in. wide, starts at the division walls of the tank with a fall of 4 in. towards the centre, where is a small pit 15 by 15 in. and 1 2 in. deep, formed in the cement bottom, connected with a 4-in. pipe bedded in the concrete, and through which the a-in. solution pipe is threaded.

The filter bottom averages 5 to 6 in. in thickness, and is made by filling in with broken stone a 3 by 3 in. wooden combing, laid in cement mortar and bolted to the bottom of the tank all round. Over this is laid the filter cloth.

Each tank is provided with three iron gates, through which the residues are trammed to the dump. Each gate is suspended by a vertical pulley from a horizontal axle running on wheels, which roll to and fro on rails running on the top of the wall on either side of the door.

A cast-iron frame is built into the brickwork round the doorway, and the door made water-tight by wedging against rubber packing. Inside the filter-tanks are laid three i8-in. tramway tracks on longitudinal sleepers, which remain in the filter, and are only removed when the filter-cloth is to be raised. In the doorway is fitted a short piece of rail composed of 2-in. square iron, provided with pins projecting from its lower surface, which pins are dropped into holes in iron plates fitted on the sills inside and outside the doorway. This piece of rail is removed when it is necessary to open the door. To still further reduce the quantity of slimes treated, there are

86 Operations At Various Works.

slime gates in the tanks. These are between the doors of the tanks, and are simply a series of wooden slats about 4 in. wide, sliding in a vertical groove 2 in. from the wall of the tank. During the filling of the tank with tailings, any slimes which remain in suspension flow over into this sluice-gate, which is raised by additional slats as the tanks fill higher and higher, and pass away by a 4-in. drain-pipe to the reservoir.

Even distribution of the tailings of the tank is effected by constantly moving the nozzle of the charging hose.

Any leakage of cyanide from the doors of the tanks is provided for by a gutter 6 in. wide being formed in the con- crete, outside the wall, and about 12 in. from it. This, starting at one end of the row of tanks, falls about 12 in. in its total length, and terminates in a small brick water-tight tank, 2 ft. square by 2 ft. 6 in. deep, from which a pipe leads to the extractor-house.

The solution drainage pipes are in. in diameter, and, until outside the tanks, are threaded through a 4-in. pipe.

The tailings are under treatment for 102 hours. Each tank of 500 tons gets about 400 tons of solutions and washes.

The strong solution contains 0*3 per cent. KCy. The weaker solution contains 0*15 per cent. KCy. The weakest solution contains 0*05 per cent. KCy.

The final wash water amounts to 25 tons.

The treatment in the leaching- vats is as follows : — It takes 30 hours to fill one of these vats. No. 3 vat, for instance, holds 550 tons of tailings. Say at 12.30 on the 29th August, 50 tons of 0-05 per cent, of cyanide solution is run on to drive out the water. When the solution is run through 75 tons of strong solution, 0*3 per cent, is put on. On the 30th August 75 more tons of 0*03 per cent, is put on. On the 31st August 65 tons of 01 5 per cent, is put on. From the ist to the jrd September 300 tons of 0*05 per cent, solution, inclusive of 25 tons of wash water, are put on. Altogether the treatment takes 6 days, the total quantity of solution employed being 565 tons for this particular tank, in which the tailings assayed 5- ..

Crown Reef Works. 87

and the residues i.i . An extraction of 80 per cent, was obtained, which is rather above the average.

The quantity of tailings treated in each tank is tallied by counting the number of trucks which are discharged. From the leaching-vats the solution runs to the extractor-house, where there are three large stock solution- tanks excavated in the ground, brick-lined and cemented, and of the same dimen- sions as the lixiviation tanks.

Into No. I tank are run all solutions up to o*o8 per cent. KCy, which makes up an average of 0*05 per cent, with the weaker solutions.

No. 2 tank receives all solutions from o'o8 to 0*2 per cent. KCy, making up an average of 0-15 per cent. KCy.

No. 3 tank receives everything from 0*2 per cent, upward, and is made up to 0*03 per cent, by adding extra strong solu- tion from the dissolving tank, and the same is agitated by means of a circulating pump.

There is also a vacuum well, consisting of two cylindrical iron receivers connected to an air-pump, in order to create a vacuum in case the filter is stopped up, or when it is necessary to assist percolation, or when it is necessary to dry the ore before discharging.

From the stock solution-tanks centrifugal pumps take the solution either from one or the other, and pump it into the filter- tanks. Before the solution passes into the extractor- boxes, the same flows into three small wooden tanks, where any impurities settle.

At this mine 50 to 60 per cent, of the gold leaving the battery is obtained in the cyanide works. The consumption of cyanide is claimed to be J lb, per ton treated and ozs. zinc per ounce of gold won. The works produced 2,914 ozs. gold in August, 1894, The works are nicely planned and splendidly constructed, and the mechanical details have been, no doubt, under the supervision of a skilled man, and may be considered as one of the fine plants on the gold fields, but they gave me the impression that their cost of erection must be far in excess of those plants which use wooden tanks.

G

88 Operations At Various Works.

Simmer and Jack Works. — When these works were erected the Company had a large accumulation of tailings, the result of five or six years' working, crushing first with 50 and afterwards with 100 stamps.

A contract was entered into for the treatment of 100,000 tons with the Rand Central Ore Reduction Company, and under the direction of Mr. Butters, one of the finest and best plants on the gold fields was erected within three months. While in Johannesburg, I made this plant the special object of my study, and I am indebted to Messrs. Torrente and Smart, the chemists in charge, for some of the working details here given, as well as for the illustrations of the works given in Plates XL, XII., XIII.

The plant has a capacity to treat 600 tons tailings daily, and consists of five vats of 600 tons capacity each ; three stock solution storage-tanks of 300 tons capacity each ; building con- taining four precipitation-boxes ; three 2-in. centrifugal pumps, for returning the solutions from the precipitation-boxes to the stoiage-tanks i one 4-in. centrifugal pump, for pumping the solutions from the storage-vats to the leaching-vats ; two lathes for cutting the zinc shavings ; engine to work lathes and pumps and a 60 horse-power boiler to supply steam to engine, and two hauling gears.

There is stabling also for forty mules, besides a large com- pound to house 200 Kaffirs, a manager's house, and six rooms for the white employees.

The tailing vats are 42 ft. in diameter and 14 ft. high, firmly bound with iron hoops i in. thick. It takes 10 to 11 hours to fill each vat.

The zinc boxes, four in number, are placed in a building which also encloses the stores, office, engines, lathes, and boiler. They are 24 ft. long, 3 ft, 6 in. broad, 2 ft. 6 in. deep. They are divided in 13 compartments. In every compartment there is a tray with a sieve of 20 to 25 holes to the inch at the bottom, on which the zinc shavings are loosely placed till they reach within 6 in. of the top of the boxes. The trays reach

Working Of The Plant. 89

within 4 in, of the bottom of the boxes, this space being left to collect the slimes resulting from the precipitation of the gold on the ainc. The compartments are divided from each other by two boards, the one reaching to the bottom of the box, and the other within in. This ensures the greatest amount of contact of the solution with zinc, and therefore an almost perfect precipitation.

A 4-in. centrifugal pump is used to run the solutions from the storages into the tailing vats, and three small 2 -in. centri- fugals return the solutions from the zinc boxes into their respective storages. A donkey pump is used for cleaning the zinc boxes, and three small tanks, called the clean-up tanks, which are used to settle the zinc slimes.

Working of the Plant. — I shall here restrict myself to the actual manipulation of the solutions. The process as car- ried out at the Simmer and Jack works has to deal (i) with old tailings containing acid ; (2) with the tailings coming daily down from the mill, where they keep a certain number of stamps working on free-milling ore (oxidized), and the remainder pn pyritic metal (blue rock).

The quicklime is ground in a ball mill, conitructed by the Gruson works, in Magdeburg, and in every truck of tailings a certain quantity is put, in proportion to the amount of the free acid therein contained.

A good practical way of ascertaining this is to take a given quantity of tailings, say i kilo., and place them in an enamelled bucket or basin, adding two or three times their weight in water. Leave them there for two or three hours, occasionally stirring the same, and the blue litmus paper will show if any acid is present. Given the presence of acid : take a weighed quantity of quicklime, finely ground (say 50 grnis.), and with a spoon throw a small quantity in the vessel containing the tailings, stirring the same all the time, and keep on adding lime, little by little, until the red litmus paper turns slightly blue, when the acid will be neutralized. By weighing again the remaining

go OPERATIONS AT VARIOUS WORKS.

lime the quantity used will be known, and if, say, 2 . has been used, we shall have that —

If I Kg. takes 2 . of lime

1,000 Kgs. (I ton) will require 1,000 times more, or 2,000 grams, which is equal to 2 Kgs.

Now, if every truck carries, say, ton of tailings, it will require I kilo, of lime, or 2*12 English lbs. The advantage of sup- plying every truck with its proper quantity of lime is to ensure its even distribution through the mass in the vat.

When the vat is full it is carefully levelled, and then is leady to be treated.

In the case when lime has been used, the tank is filled with water and allowed to stand full for one hour. By this time the acid has been neutralized, and it can be started leach- ing. The water from the leaching carries, as a rule, a small excess of alkali, and it is pumped back into one of the storages, to be used again in the next tank. To this water the name of alkaline solution, or caustic wash, is given ; and, as a rule, the moisture in the tailings is enough to keep the amount in hand constant, so that very seldom is it necessary to add any more water to it. It is necessary to keep the solution in a storage tank, so as to avoid any loss of gold, as it always carries with it whatever gold may have been left in the filters, &c., from previous treatment. It will also be found that, after treating a few tanks, the solution will contain a small percentage of cyanide.

When the caustic wash is partly drained off, the top of the tailings are turned over with a shovel, so as to break the slimy film which forms ; and then follows the solution, whose quantity must be so regulated as to give about f lb. of cyanide (or less) per ton of tailings, and be of a strength of '25 to '3 per cent.

This solution is followed by another, which is called weak solution, and whose strength varies between '08 and '15 per cent, of cyanide, and is kept on until the tank is ready to leach dry.

Precipitation Of Gold From The Solution, 9 1

Should the solution in the storages get short, a water wash is given to terminate the operation : otherwise the weak solu- tion is used to the last.

The system followed in pumping up the solutions is to keep the tanks always well covered up from the moment the strong solution is first pumped up. As soon as the tailings show, more solution is added until the full amount has been pumped up. This is followed at once with a weak solution, which, in its turn, must keep the tailings always well covered till the tank is ready to water wash or leach dry.

The time employed in these diverse operations is about as follows : —

1st. Caustic wash (weUer or otherwise).

Pumping 3 hours

Contact .1 hour

Leaching 8 hours

2Dd. Strong solution.

Is pumped without stopping the leaching until all is on,

takes about . 8 hours

(Amount used about 160 tons.)

3rd. Weak solution.

Is kept up circulating as previously explained about . . 40 hours

4th. Water wash {if necessary).

Leach dry 24 hours

Time it took to load tank 12 „

Total from the time of starting filling till ready to discharge, 96 hours (Total amount of strong and weak solution between 500 and 600 tons.)

Precipitation of the Gk>ld from the Solution. — From the bottom of the tailing tanks, and under the filter, a pipe 2 in. in diameter carries the solution to the zinc boxes, four in num- ber, as already mentioned.

The object of these four zinc boxes is to return the solutions to their respective storages, for which purpose they are con- nected : —

2 with the strong solution storage, I with the weak solution storage, 1 with the caustic wash storage.

92 Operations At Various Works.

A simple assay will show to the man in charge when to lead the leachings through the one or through the other box. The boxes have a fall of 9 in. in their full length.

'Beoovery of the Gold from the Gold Slimes. — The gold slimes are settled in the slime tank by means of alum or sodium sulphide. After 12 hours the clear portion is syphoned down, and these slimes are then ready for further treatment. The test of the sodium sulphate is lead paper, which takes a brownish colour when the residues in ihe tank have been saturated.

To make sodium sulphide : —

gaJs. water,

62 lbs. caustic soda,

41 lbs. sulphur (broken small).

Boil the water with the soda till dissolved, and then add the sulphur slowly, as it is apt to boil over. Dilute to 40 gals. Smaller quantities in proportion.

The slimes are transferred to a special calcinating furnace, where they are dried and the zinc slightly oxidized. They are left to roast at a dull red heat for four hours, although this time (epends upon the quantity and nature of the slimes. After roasting, they are put into iron dishes and mixed with the smelting fluxes.

To 6 shovels of slimes add —

shovel flnor spar,

3 shovels borax,

I shovd carbonate of soda.

If the slimes are hard, and cake, they must be ground fine so as to mix them well with the fluxes.

If the fire gets too fierce while smelting the slimes, and the pots overflow, keep the cover off", and throw a little borax on the top. This overflowing is objectionable, as the slag cakes at the bottom of the grate and chokes the fires. When all the gold belonging to one lot of slimes has been secured it is then melted down into one bar.

Rand Central Ore Reduction Company. 93

Langlaagte Estate. — At these works the pulp on leaving the mill passes into three settling dams in rotation. The lower retaining walls are composed of sand-bags and two doors are provided in each dam, one for the passage of the full, and the other for the empty trucks.

The discharge of the slimes, when filling the dams, is regulated by strips of wood in the passages, the same system being employed at the Randfontein Estate. Lines of rails are placed inside these dams; and at right angles to these, immediately inside the lower walls, runs a line of rails sunk in the ground, on which is placed a travelling-carriage, on to which the tipping trucks are run and brought outside the dams, to be picked up by a mechanical haulage composed of an end- less wire rope, and carried to the cyanide works.

Here are ten vats sunk in the ground, five on each side, and when the tailings are treated they are emptied by means of a travelling crane, which lowers the body of 25 cubic feet capacity trucks into the vats, to be filled by Kaffirs, and are then lifted out again and placed on their respective carriage, and conveyed by mechanical haulage to the waste dumps. A vat can be emptied in six hours.

The mechanical haulage is operated by a 25 n.h.p. com- pound engine, with 24-in. stroke and 25 h.p. boiler, consuming about if tons of coal per day.

Band Central Ore Beduetion Company. — A promi- nent position in the treatment of tailings and concentrates is occupied by the above company, which was organised by Mr. Ad. Goerz, M.E., representing a powerful German syndicate. The Company, who own the Siemens and Halske patents for South Africa, have bought over 1,000,000 tons of tailings, of an average value of ., and 6,173 of concentrates, of an average assay and contents of about 4 ozs. The tech- nical work is directed by Mr. Charles Butters.

The establishments of the Company are six in number, and include (i) their Central works (shown in Plate XIV.) — con- sisting of a chlorination plant and cyanide works, with work-

94 Operations At Various Works.

shops and foundry, where everything is made for vats, imple- ments, etc, required for the construction of cyanide plants ; (2) works Maraisburg, which were built to treat tailings from the Main Reef, Aurora, and Aurora West Companies ; (3) the Simmer and Jack works, already described ; (4) the Worcester works, ah-eady described ; (5) the No. i works, which treat tailings from the Pioneer Company ; and (6) the No. 2 works, which treat tailings from the United Langlaagte.

During the year ending June 30th, 1894, the Company treated at their six works 311,561 tons of tailings, yielding 52,434*46 ozs., of a value of ;;209,737 i6s. 8d.

Dry Cruflhing. — As proper percolation of the solution can only be obtained by having the pulp coarse enough to allow of easy leaching, several establishments in the United States have adopted dry crushing.

Mr. Philip Argall, M.R.I.A. (see Mining Journal London, October 13th, 1894), says : best results hitherto obtained on rather friable oxidized siliceous ores have been produced by the following series of machines : —

{a) An ordinary Blake crusher, which reduces the ore to about li in. ;

" ip) A three-jaw multiple Blake crusher, which brings the ore down to J in, ; followed by

" (r) A five-jaw multiple, that reduces it to about 6 mesh ; and

" Rolls that finish the pulp to pass a 30 mesh screen.

" The ore is screened after each crushing, so that the portion reduced to the desired fineness passes direct to the finished ore- bin, instead of, as is too often the case, going to the next machine to be reduced to a further and unnecessary degree of fineness. This plant, crushing to 30 mesh, produces only from 5 to 10 per cent, fine enough to pass 200 mesh.

*With soft, clayey ore, from 5 to 15 per cent of the finest dust has to be removed before a leachable product is obtained."

The preparation of the ore, in dry as well as in wet crush-

Dry Crushing. 95

ing, is a very important step in the direct treatment by the cyanide process, and the latter has only been brought to its present state of perfection within the last six months. Owing to the scarcity of water in West Australia, I believe that many mines will have to adopt dry crushing plants, and that the system will reach a high degree of perfection.

Where wet crushing is resorted to, owing to the large quantity of ore crushed per stamp, a large amount of water is necessarily used, and consequently an appreciable quantity of finely-divided gold is carried away in the rush of water.

Mr. Crosse has shown (in a lecture before the Chemical and Metallurgical Society of Johannesburg) that what takes place in wet crushing also takes place in dry crushing. He says : " I received about three-quarters of a ton of ore from the Roodepoort Deep Level Gold Mining Company, which was crushed in a Grusonwerk ball mill ; the sieving used had 500 holes to the square inch. It assayed per short ton : Fine gold, 5 ozs. 12 . ; fine silver, 16 , ; and contained 3*31 per cent. ofpyriteSy determined as bisulphide of iron.

There remained on A -)- 1,600 mesh — 28*35 of he ore B + 3,600 „ - 2250 C + 6,400 „ - 14-15 „ passed through D — 6,400 „ — 35*o „ „

A contained 2*01 per cent, pyrites and assayed 4 ozs. i dwt. 16 grs. B „ 2-96 „ „ „ 4 „ 14 . 12 „

D „ 4*64 „ „ „ 6 „ 16 „ 12 ,,

Proportion

of gold to pyrites.

A contained 22*03 P-c Of the gold and 15*51 p.c of the pyrites 143

B „ 20-28 „ „ 18-21 „ „ 1-182

C „ 25-03 „ „ 17-14 „ „ 1-196

D „ 42-66 „ „ 49*07 „ „ 1-196

" This table shows that, though the crushed ore was passed through a coarse sieve, the gold-bearing portions were more finely crushed, and that 35 per cent, of the crushed ore passed

96 Operations At Various Works.

through a sieve having 6,400 holes per square inch, and this portion contained 42*66 per cent, of the gold, and 4907 per cent, of the pyrites. A milling assay gave 36*6 per cent, of free gold. The sample came from a depth of 700 ft. The ore leached easily.

"The results were as follows : — Total gold in the ore treated, 3-509 ozs. ; left in residues, 0*446 ozs. ; extraction, 887 per cent. A cubic foot of the dry crushed ore weighed 103 lbs. ; one ton would occupy 19*4 cubic feet"

Mr. Crosse suggests that the best method for dealing with banket would be, to arrange a series ot Gates crushers, one Kroni roll, and then a short treatment in a ball mill, with a 50 to 60 mesh screen to grind up the pyrites and cement, and so liberate the finely divided gold.

The G-ibraltar Cyanide Works, New South Wales. — These works, at which the Siemens and Halske process is followed, are so situated that the tailings, after they leave the Frue vanners, flow by gravitation into the collecting vats, which are fitted with Butters and Mein's distributors, and which col- lect a perfectly leachable product amounting to 90 per cent, of the ore crushed. The ore does not slime, and furnishes an even-grained product, and Mr. F. W. Alsop, the manager of the works, who has had a large South African experience, considers the Gibraltar tailings the most uniform-grained stuff that he has met with.

There are three collecting vats 18 ft. in diameter, with 8 ft. 6 in. stays, holding 60 tons apiece.

There is very little acidity in the ore, and the quantity of lime added is consequently very small. After draining the collecting vats are emptied through Butters' patent discharge, and trucked to the leaching vats. There are six leaching vats of the same dimensions as the collecting vats. The advantage of having a large leaching capacity is to admit of the use of very weak solutions and prolonged time of contact, and having six vats it gives uniformity in the daily operations.

Gibraltar Cyanide Works. 97

Thus if No. I vat is filled on Monday it receives the strong solution on Tuesday, is under weak solution Wednesday, Thurs- day, and Friday, water-washed Friday night, discharged on Saturday, ready for refilling on the following Monday.

In this manner, on every day of each week, every vat is in the same stage of the cycle of operations. By this uniformity errors are avoided.

From the leaching vats the solution flows into inter- mediate vats which regulate the supply and pressure to the precipitation boxes. There are three intermediate vats 15 ft. in diameter.

In the precipitation house there are four electric precipita- tion boxes, 18 ft. long by 5 ft. wide by 4 ft. deep.

Each box is capable of allowing a flow of 15 tons of solution per 24 hours.

There are lead cathodes and 4Qr iron anodes S ft. by 3 ft, in each box, giving a lead surface of 1,260 square feet, and the iron giving the same surface.

There are three solution vats — one for the strong solution, one for the weak, and one for the alkaline wash. The whole pumping is done by a 3-in. centrifugal pump.

The motive power consists of a lo-horse multitubular boiler, and 6 h. p. Tangycolonial engine. The dynamo, made by Siemens and Halske, is of 20 volts and 90 ampbres.

The capacity of the works is for 1,500 tons monthly, and they have been designed and erected by Mr. Edgar Smart, A.M.I.C.E., the engineer for the General Exploration Com- pany, who are working the Siemens and Halske patent in Australia. The total cost of the works was ;£'4,4oo.

In the annexed illustration (Fig. 38, Plate XV.) is given a detailed diagrammatic sketch of the electric precipitation plant at these works.

I am unable to give working results, as at the time of writing (September, 1897) the works were only just about to start operations, but experiments on 500 lb. lots have shown that the actual extraction will be 80 per cent, and over.

98 Operations At Various Works.

Cyanide Plant of the May Consolidated €k>ld Mininflj Company, Witwatersrand.* — These works were designed and erected by the Rand Central Ore Reduction Company, and were originally intended to treat the concentrates and sands from a 6o-stamp battery, but were afterwards enlarged to be able to treat the crushings from their present loo-stamp battery. A view of the plant, showing the upper row of vats, is given in the frontispiece to this volume.

The pulp after leaving the battery runs into a pit situated close to the battery house, from which it is raised by a vertical plunger pump and delivered into a launder carried overhead on wooden trestles, high enough to deliver into the collecting vats.

In this plant the collecting vats are arranged alongside the leaching vats, and high enough to allow of the easy and cheap transferring of the sands from the one vat to the other, the sands being shovelled through the discharge doors in the bottom of collecting vats into trucks which run by gravitation to the leaching vats on a tram-line placed over these vats. A vat of 200 tons may in this method be transferred from one vat to another by 14 Kaffirs in about three hours.

The collecting vats are 24 feet diameter by 13 feet deep, and are carried on independent wooden trestles of 9 in. by 9 in. timbers, and 20 feet high. They are each fitted with one of Butters and Meins* automatic distributors, for distributing the pulp evenly over the vat ; they are also fitted with wood filter frames and filter cloth, the cloth being protected by narrow wooden shovelling strips. They are also arranged with four of Butters' patent bottom discharge doors. These collecting vats were at first arranged with an annular ring round the top to receive the clear water overflow, but it not being convenient to fill the vat with water before running in the pulp, or when the battery pulp stopped before fresh water could be run on, the slimes which were in suspension in the vat gradually began to settle.

♦ For the description of this plant, I am indebted to the courtesy of Mr. Charles Butters.— M. E. (1898).

May Consolidated Works. 99

To get over these difficulties the slat-gate was introduced. This appliance consists of two upright deals, with grooves np the side in which are dropped slats about 4 inches deep ; as the sand rises, a screen is fastened behind the slats to prevent any chips of wood, &c., getting in to the spitzkasten. A slime oveiflow door is arranged in connection with it, so that the vats can be closed while the charge is under preliminary cyanide treatment. The width of these slat-gates varies accord- ing to the diameter of vats. For a vat 24 feet wide, a slat- gate is used I foot 8 inches wide.

With this arrangement, the pulp from the battery is run through the distributor into the empty vat, which fills up to the overflow of the first slat about 9 inches deep, when the slimes begin to overflow into the launder going to the slime dams ; as the sand rises in the vats, another slat is lowered, and so on until the vat is full. The water is then drained off the charge; and now that double treatment is generally adopted in modem plants, the first portion of this operation is carried on in these vats by running through a dilute cyanide solution, the charge then being transferred to the leaching vats.

Over the collecting vats a gangway is run from where the slat gates and other general work in connection with these vats can be attended to.

The leaching vats are 30 feet in diameter by 7 feet deep, and are carried about 6 feet above the ground on stone walls, to give sufficient head room underneath the vats for the dis- charging of sands. Two discharge lines are arranged under them, the vats being supported over these discharge lines on wooden joists. The framework of the tram-lines, on which the trucks run from the collecting to the leaching vats, rests on the masonry foundation, and also inside the vats, two lines of trams being introduced, enabling the charge to be transferred as quickly as possible. The top of this tram-line is made of narrow slats of wood placed about inches apart, thus forming a sort of grating, so that when a -load of sand is dis- charged on the top of it, the sand is fairly well disintegrated before falling into position in the vat This is claimed to be.

lOO OPERATIONS AT VARIOUS WORKS.

an advantage, as lumps of sand are broken up which might otherwise remain whole, thus exposing more surfaces to the oxygen, and hastening the solution of the gold.

This tram-line is guarded all round by hand-railing, and is so arranged that the discharging trucks can be run from any collecting to any leaching vat, so that one set of trucks will do all the transferring. These vats are fitted with a wood filter-bed and filter-cloth, the same as in the collecting vats. They are each arranged with six of Butters* patent discharge doors, fitted in two lines at 12 feet centres, and directly over the discharge tunnels.

The leaching pipes from the bottom of all collecting and leaching vats are ij inches in diameter, and are all led together to the head of the precipitation boxes inside the depositing shed. The pipes from the collecting vats are arranged so that all the water may be drained off the sands before the first cyanide solution is pumped on.

There are four storage vats, where the various solutions are stored after the gold has been precipitated firom them, and where the strong solution is made up ready for use on a fresh lot of tailings ; the storages hold strong, weak, and alkaline solutions and water-wash. A 4-inch centrifugal pump and pipe system is used for pumping these various solutions, and is arranged to draw from either solution storage and denver into any collecting or leaching vat.

In the depositing shed are arranged the depositing boxes, testing house, motor, precipitating machine and pumps. The motor is for 500 volts, 30 amps., and 700 revs., and drives on to a 2 -inch counter shaft, from where power is taken for the precipitation machine of 15 volts, 400 amps., and for the 4-mch centrifugal pump throwing on solutions.

The Siemens and Halske electrical precipitation process is used in this plant, and most excellent results are obtained by it. The box is constructed of 9-inch by 2-inch wrought- tongued and grooved pine boards, and is divided length- wise into compartments, each compartment being filled with iron plates placed parallel with the side of box, and between

CAUSE OF CYANIDE CONSUMPTION. lOl

which hang the lead frames. The anodes or iron plates are 2 feet by 2 feet 3 inches, by -ft- inch thick, and are sewn up in jute bags and held loosely in position at 4-inch centres, by narrow wooden strips fastened to sides of compart- ments; they are all supported about 6 inches from bottom of box on wooden strips, to keep them out of any mud that may collect in the box. The cathodes or lead frames are made of No. 8 gauge galvanized wire, the lead sheets being cut into strips about J-inch wide, one end about 2 inches wide, being left uncut to lap over the wire frame, and fastened to same with clips made of 4 lbs. per foot lead, each frame carry- ing four sheets of lead.

The cathodes or lead strips are cut from sheets weighing 2*62 oz. per square foot. On each side of the box run the main copper wires from which there go alternate branch wires across the box connecting with the anodes and cathodes.

Determination of the Cause of Cyanide Consump- tion.* — Should the consumption of cyanide be high, the cause of consumption may be determined by an analysis of the cyanide solution: — For every part of cyanide rendered in- operative, a corresponding proportion of metal enters solution. Thus one part by weight of iron consumes seven parts by weight of potassiuhi cyanide. The following equations repre- sent the reactions most frequently encountered :

FeSO* + 6KCN K4Fe(CN)6 + K2SO4 56(at. weight of Fe) : 390(mol. wt. 6KCN) 1:7

2nS04 + 4KCN =K22n(CN)4 + K2SO4 or 65(at. weight 2n) : 26o(mol. wt. 4KCN) i : 4

Salts of aluminium and magnesium act in a different manner with potassium cyanide, their hydrates being formed with the liberation of hydrocyanic acid, thus :

♦ Laboratory Tests in connection with the Extraction of Gold from Ors by the Cyanide Process, by H. Van F. Funnan. TransacUons of the American Institute 0/ Mining Engineers 1896.

)02 Operations At Various Works.

Al2(S04)8 + 6KCN + 6H2O - Al2(OH)6 + 3K2SO4 + 6HCN MgSO* + 2KCN + 2H2O - Mg(0H)2 + K2SO4 + 2HCN

A preliminary alkaline treatment overcomes this objection- able feature, their hydrates being precipitated, which are then inert towards potassium cyanide, thus :

MgSOi + Ca(OH)3 Mg(0H)2 + CaSO*

insoluble magnesium hydrate and insoluble calcium sulphate being formed.

Soluble sulphides, formed by the action of potassium cyanide on some metallic sulphides, again react to some extent on the cyanide, with the formation of sulpho-cyanide of potassium, thus :

2nS + 4KCN K22n(CN)4 + K2S

K2S + Kcn + H,0 + O 2Koh + Kcns

To determine the cause of the consumption of cyanide, place 100 . of the pulp in a wide-mouthed bottle, add 200 C.C. of the cyanide solution and agitate for 15 hours, filter, take 20 c.c. of the filtrate (equivalent to 10 . of ore), and evaporate almost to dryness in a porcelain dish. Add some strong sulphuric acid, evaporate almost to dryness, and cool. Dilute with water, add some hydrochloric acid, and heat to effect solution if necessary. The metal in solution may now be determined by the usual methods.

The strong sulphuric acid at a high temperature decomposes the metallic cyanides, thus :

2AgCN + 3H2SO4 + 2H2O AgaSOi + 2NH4HSO4 + 2C02

Strong nitric acid may be used in place of strong sulphuric acid ; but hydrochloric acid cannot be used, as it leaves the metal in the form of a double cyanide salt, which is soluble. The reactions with nitric and hydrochloric acids are :

AgCN + HNO + 2H2O AgNO, + COa + NHs + H K4Fe(CN)6 + 4HCI H4Fe(CN)6 + 4KCI

Determination of the Cause gf Non-Extraction.— This may be due to numerous causes ; such as a tosl de$truc

Cause Of Non-Extraction. Io3

tion of potassium cyanide by certain salts of the base metals present in a form readily attacked by the potassium cyanide.

The gold may be combined or alloyed with tellurium, antimony, bismuth, &c., &c, in which case the cyanide is in- operative until the combination is broken up ; the presence of soluble sulphides in solution ; the character of the gangue, such as kaolin or talc, which may be present in such quantities as to effectually prevent percolation.

To overcome these difficulties the following methods may be tried : —

In the case of an ore which consumes a large quantity of cyanide, if a preliminary wash with water, weak acid, or alkali is effective, the ore may be classed as one not adapted to the process.

The difficulty due to the presence of bismuth, antimony, &C., in combination or as an alloy with the gold, may some- times be overcome by fine grinding and long contact with the cyanide solution ; but the usual method is to treat the ore to a preliminary roast, which converts the gold into a condition in which it is readily attacked by cyanide.

The difficulty due to the presence of soluble sulphides can be overcome by the addition of a soluble lead salt, or the addition of an oxidizing agent.

Should the ore contain kaolin or talc, if coarse crushing is ineffectual, the ore has to be treated by the method now in vogue on the Rand for slime treatment, namely, the ore has to be ground fine and submitted to that process.

Ores containing considerable quantities of oxidized copper minerals are to be classed as not adapted to the process.

Tanitsky says that he experimented for four months with the ores from the Lyndhurst Gold Field Mines, New South Wales, which contain a large percentage of sulphate of alumina, and decomposition products of iron pyrites.

He roasted the ores so as to convert the protoxide of iron into peroxide, and most of the alumina sulphate into alumina oxide. The remaining sulphate of alumina must be washed out, as it would decompose large quantities of potassium

H

104 Operations At Various Works.

cyanide. The neutralized alumina existing as hydrated oxide in the roasted ore reacts upon potassium cyanide, forming aluminate of potassium, and setting free hydrocyanic acid, and a large quantity of potassium cyanide is consumed also in this way.

Messrs. Park and Skey say : " The presence of a small percentage of aniimonite in ores prevents the successful operation of the process described in the specification. Weak solutions of cyanide of potassium were used in treating a large accumu- lation of tailings containing antimonite at Boatman's Creek, near Reefton, New Zealand ; and all attempts to treat them have failed, owing chiefly to the large consumption of cyanide and the low rate of extraction."

Chapter V.

The Chemistry Of The Cyanide Process.*

Solution of the Gold. — The solubility of gold in a solu- tion of cyanide of potassium has long been known. The pre- sence of oxygen is necessary for the reaction, after the formula :

2Au + 4KCy + O + HoO 2KAuCy2 + 2KHO.

That is, a double cyanide of gold and potassium is formed. This was proved by the formation, upon evaporating the solu- tion, of octahedral crystals answering to the formula. From the solution the gold is precipitated by filiform zinc. Based on these two reactions, the MacArthur-Forrest Company have obtained the following patent : —

"The invention consists in subjecting the auriferous or argentiferous ores to the action of a solution containing a small quantity of cyanide, as hereinafter set forth, without any other chemically active agent, such quantity of cyanide being reckoned according to its cyanogen, and the cyanogen being proportioned to the quantity of gold or silver, estimated by assay or otherwise to be in the ores under treatment. By treating the ores with the dilute and simple solution of a cyanide, the gold or silver is, or the gold and silver are, obtained in solution, while any base metals in the ores are left undissolved, except to a practically inappreciable extent ;

The chemistry of the process has been described by Messrs. Butters and Clennell in the Engineering and Mining Journal of October 22nd and 29th, 1892, and from their articles (with their permission) this matter is extracted.

I06 The Chemistry Of The Cyanide Process.

whereas when a cyanide is used in combination with an electric current, or in conjunction with another active chemirial agent- such as carbonate of ammonium, or chloride of sodium, or phosphoric acid — or when the solution contains too much cyanide, not only is there a greater expenditure of chemicals in the first instance, but the base metals are disi>olved to a large extent along with the gold or silver, and for their sub- sequent separation involve extra expense which is saved by our process.

" In carrying out our invention practically, we take the ore in a powdered state and mix with it the solution of cyanide in a vessel made of, or lined with, any material not appreciably acted on by the solution. . . . We regulate the quantity of cyanide so that its cyanogen will be in proportion to the quan- tity of gold or silver in the charge of ore ; but in all cases we dissolve it in sufficient water to keep the solution extremely dilute, because it is when the solution is dilute that it has a selective action, such as to dissolve the gold or silver in prefe- rence to the baser metals.

" In dealing with ores containing 20 oz. or less of gold or silver, or gold and silver, per ton, we find it most advan- tageous to use a quantity of cyanide, the cyanogen of which is equal in weight to from i to 4 parts for every 1,000 parts of the ore, and we dissolve the cyanide in a quantity ot water of about half the weight of the ore. In the case of richer ores, while increasing the quantity of cyanide to suit the greater quantity of gold or silver, we also increase ihe quantity of water so as to keep the solution dilute. In other words, the cyanide solution should contain from 2 to 8 parts, by weight, of cyanogen to 1,000 parts of water, and the quantity of the solution used should be determined by the richness of the ore. After the solution has been decanted or separated from the undissolved residues, the gold and silver may be obtained from it in any convenient way, such as evaporating the .solution to dryness and fusing the resulting saline residue, or by treating the solution with sodium amalgam. , Having fully described our invention, what we desire to

Solution Of Gold In Cyanide. 107

claim and secure by letters patent is : The process of separating precious metals from ore containing base metal, which process consists in subjecting the powdered ore to the action of a cyanide solution containing cyanogen in the proportion not exceeding 8 parts of cyanogen to i,ooo parts of water.**

Some time after this patent was issued, patents covering the use of zinc, preferably filiform, or threadlike, for a precipitating agent, and the use of caustic alkalies, for neutralizing ores con- taining acids, or acid salts, were granted to Messrs. MacArthur and Forrest. It will be seen, therefore, that their patents cover substantially three points : namely, the use of dilute solutions of cyanide (not more than 8 parts of cyanogen to 1,000 parts of water) ; the use of zinc, preferably filiform, as a precipitant ; and the employment of caustic alkalies for neutralizing acid ores.

Mr. Louis Janin claims,* that the necessity of oxygen in the reaction which takes place in the solution of the gold in the cyanide has not been proved, and establishes the following formula : —

Au + 2KCy + H2O KAuCyi + KHO + H.

More recent investigation has proved that oxygenation of the gold in the process is essential to its success, and the idea has been suggested that an artificial oxygenation of the gold in rebellious ores may lead to the solution of the problem of treating them successfully by the cyanide process

Mr. R. Feldtmann, in his " Notes on Gold Extraction," gives the following additional formulae, throwing light on the reac- tions which take place in the solutions of gold in cyanide : —

When hydrocyanic acid dissolves gold from acid tailings (assuming that the cyanicide has destroyed all the cyanide of potassium),

2Au + 8HCy + 3O lAuHCyi + 3H2O,

an auric hydrocyanide is produced, and this compound is not, or is at any rate only imperfectly, precipitated by zinc. Even rendering the solution alkaline by addition of caustic soda or potash, does not appear in such a case to cause a good precipi- Mineral Industry 1892, Scientific Publishing Co. Nmt York.

108 The Chemistry Of The Cyanide Process.

tation. This may be owing simply to the auric compound being more stable than the aurous salt ordinarily obtained, or may be owing to the absence of free cyanide of potassium. The addition of alkali to such a solution of auri cyanic acid may be assumed to form auric potassic cyanide in this way : —

AuHCy4 + KOH AuKCyi + H2O.

The addition of an acid to working solutions containing free potassic cyanide, and a certain amount of auro potassic cyanide, would appear — probably indirectly — to convert at least a por- tion of the latter into auri cyanic acid. Probably the hydro- cyanic acid, liberated by the decomposition of the potassic cyanide, combines with the auro potassic cyanide to form auri potassic cyanide : —

AuKCy2 + 2HCy + O AuKCy* + HaO,

which, in its turn, acted on by the mineral acid, is converted mto auri cyanic acid : —

2AuKCy4 + H2SO4 2AuHCy4 + K2SO4.

Solubility of other Metals and Minerals. — According to Gmelin, zinc, iron, nickel, and copper are dissolved by potassium cyanide, with evolution of hydrogen ; cadmium and silver in the presence of oxygen ; and tin, mercury, and platinum not at all. Sulphide of silver is dissolved by strong solutions and a sufficient quantity of weak solution. Silver arsenate Aga As O4, and silver antimonate AgSbOs are readily dissolved by potassium cyanide, as are many of the argentiferous arsenical and antimonial minerals found in nature. Chloride of silver dissolves readily, forming chloride of the alkali and a double cyanide of silver and potassium. While metallic silver, when sufficiently fine, dissolves readily in the solution, that found native in ores is not attacked, unless exist- ing in thin laminae. The oxides and sulphides of copper are attacked by the solution and dissolved, as is metallic copper.

It is claimed that the presence of copper sulphide in a silver or gold ore prevents the precious metals from going into soli*-

SOLUBILITY OF METALS. lOQ

tion. Although experiments have shown that little or no silver or gold is dissolved in certain ores containing sulphide of cop- per, this question is by no means settled, as artificially prepared sulphide of silver is dissolved in actual contact with the copper compounds. Metallic iron is attacked, but very slowly. Ferric hydrate is not attacked by the solution, but ferrous hydrate, formed in the neutralization of the iron salts by alkali, is attacked by cyanide, according to the reaction : —

FesHO + 6KCy KiFeCy* + KOH.

Thus ferrocyanide of potassium and caustic potash is formed.

Louis Janin's Experiments on Silver Ores.* — Mr. Janin made a series of experiments on the applicability of the cyanide process to various types of ores, with the following results : —

Sample No, i. — Grand Central Mine, Arizona. Siliceous ores, containing considerable quantities of lime and manganese. The silver minerals are principally cerargyrite and argentite, and gave an extraction of 92*6 per cent.

Sample No. 2.— Christy Mine, Silver Reef, Utah. The sil- ver minerals are chlorides, sulphides, metallic silver, in a gangue of sandstone, somewhat discoloured by carbonate of copper. Extraction, 80 per cent.

Sample No. 3. — Horn Silver Ore, Utah. The silver prin- cipally as chloride. Extraction, 93 per cent

Sample No, 4. — Tybo, Nevada. The mineral is principally a sulphide and fahlore. Extraction, 71*8 per cent.

Sample No* 5. — Gangue siliceous with chloride of silver. Extraction, 97 per cent.

Sample No, 6. — Ramshorn, Idaho. Containing galena and carbonate of lead, pyrites, and zinc blende. Extraction, 80 per cent.

Sample No. 7.— Broken Hill, New South Wales. Chloro bro- mide of silver in kaolin, qujirtz, and garnet. Extraction, 99*7 per cent. From Mineral Industry. The Scientific Publishing Co., New York.

no THE CHEMISTRY OF THE CYANIDE PROCESS.

Sample No. 8. — Broken Hill, New South Wales. A siliceous iron ore containing 38 per cent. Fe O. Extraction, 84*6 per cent

Sample No, 9. — BuUionville, Nevada. Tailings, containing 10 per cent, carbonate of lead, galena, and iron in a siliceous gangue. Extraction, 32 per cent

Sample No, lo. — Bertrand and Geddes, Nevada. Contains antimoniate of lead, with which the silver was combined. Ex- traction, 1 1 '8 per cent

Sample No, n. — Argenta, Montana. Contains over 40 per cent lead. Extraction, 57 per cent.

Sample Nos, 12 and i. — Belmont, Nevada. Contains arseni- cal pyrites, pyrite, blende, and galena, with the silver as fahlore, and arsenical and antimonial ruby forms. Extraction, 35 and 47 'S per cent.

Sample No, 14. — Las Yedras, Mexico. Contains large quantities of carbonate of lime, with the silver in the form of ruby silver, and arsenical pyrites. Extraction, 41 5 per cent.

Sample Nos. 15 and 16. — Ontario and Daly ores, Utah. Con- tain principally fahlore, more or less decomposed with some lead and zinc. Extraction, 72*5 and 81*1 per cent.

The conclusions and deductions to be derived from a study of the foregoing are, that silver in oxidized surface ores, or where it occurs as a chloride, is readily attacked by cyanide of potassium, and that where no minerals are present which exert an unfavourable influence, this method may prove economical. It must be confessed, however, that even with these conditions it has but a limited range of usefulness. On the other hand, where lead, oxide of copper, or oxides of iron occur, the results are so poor as to preclude the use of the process.

The results obtained from different samples of silver ore from the same mine vary greatly, for a slight increase of an undesirable element, which would not affect amalgamation in the slightest degree, causes a great decrease in the percentage of extraction by cyanide.

In the case of a number of experiments on Daly ores, the

Treatment Of Pyritic Ores. Hi

results varied 16*2 per cent., while, with amalgamation there was a variation of but 2 per cent.

Besiilts of Experiments made with Potassium Cy- anide on Gold and Silver Ores. — It would seem proba- ble that, in ores containing both gold and silver, only the oxidized surface ores can be treated with success, both the silver and gold minerals from depth proving refractory. With the majority of these ores the consumption of cyanide would be large, as many minerals other than those of silver are con- tained in them, and would have a decomposing action upon the solution.

Experiments on Gregory's concentrates, consisting of pyrites, some arsenopyrite and chalcopyrite, with traces of galena and blende, gave an extraction of 90 per cent, gold and only 3 per cent, silver.

Delamar, Idaho, consisting of a siliceous and limestone gangue, impregnated with silver sulphides and chlorides, and iron pyrites. Results : gold, 90 per cent, ; silver, 83 per cent.

Revenue Mine, Montana — Oxidized surface ores, contain- ing a considerable quantity of iron. Gold, 94 per cent. ; silver, 5-2 per cent.

Southern Cross, Montana, contains 40 to 50 per cent, of limonite. Gold, 93 per cent. ; silver, 50 per cent.

I am not aware, so far, that the cyanide treatment has been applied on a large scale to the extraction of silver ores.

Treatment of Fyritio Ore Previous to Lixiviation with Cyanide. — Pyritous ore, if but slightly exposed to atmospheric action, always contains free sulphuric acid and soluble salts of iron. To prevent reactions on the solution, it is necessary to leach these ores with water previous to lixivia- tion with cyanide of potassium, and before washing with an alkali. If the alkali solution were to be added directly to the ore, the consumption of alkali would be extremely large, and the amount of solution necessary, if lime were to be used, would prove inconvenient to handle.

112 Tiie Chemistry Of The Cyanide Process.

The alkali solution, assuming that caustic soda is used, reacts on basic iron salts, insoluble in water, according to the following reactions : —

Fe208,SOs + 2NaH0 + H,0 Fea(H0)6 + Na2S04

and

Fc,03,2SOs + 4NaHO + H2O Fe2(HO)6 + 2Na2S04.

Thus ferric hydrate and sodium sulphate (or calcium sul- phate, if lime is used) are formed. Sodium sulphate is soluble and passes off with the wash water, but calcium sulphate remains. The hydrate of the sesquioxide of iron is insoluble in water, and, to all appearances, is unattacked by the cyanide solution ; but the hydrate of protoxide is dissolved with forma- lion of ferrocyanide of potassium.

Mr. C. W. Merrill precipitated ferrous hydrate by caustic potash from a cyanide solution. The solution contained but a small percentage of free cyanide, however, as it had already acted on ore and zinc in the precipitation of the dissolved gold, and it is unlikely that this reaction, regenerating the cyanide of potassium, which had been rendered inert by the solution of the iron, would occur in a comparatively strong solution.

Treatment of Fyritic Ores by Cyanide of Potassmm. — Owing to the rapid decomposition of the pyriiic ores when exposed to the atmosphere, precautionary measures have to be taken to treat them effectively. Owing to their successful treatment by the chlorination process, it is at present difficult to say if the cyanide process will eventually prove a serious competitor, as no doubt the process which is the cheapest will maintain itself.

Pyrite (Fe S2) is decomposed by the oxygen of the air and moisture into soluble ferrous sulphate and free monohydrated sulphuric acid, according to the reaction :

FeS2 + H2O + 70 FeSOi + H2SO4.

The ferrous sulphate is decomposed by the action of the air to insoluble basic sulphates. In addition, normal ferric sulphate (Fe 3 SO4) is produced, which gradually loses acid

Treatment Of Pyritic Ores. Ii3

and becomes a soluble basic sulphate, Fca Oj aSOj. There are many basic salts of somewhat complex and doubtful com- position formed likewise. Thus in an oxidized ore which has contained pyrite are found sulphuric acid, ferrous sulphate, basic ferrous sulphates, ferric sulphate, and basic ferric sul- phates, all of which react upon potassium cyanide.

Sulphuric acid reacts upon potassium cyanide with evolu- tion of hydrocyanic acid according to the reaction :

2KCy + H2SO4 K2SO4 + 2HCy.

Ferrous sulphate reacts upon cyanide with the formation of ferrous cyanide, a yellowish red flocculent precipitate :

FeSOi + 2KCy FeCya + K3SO4.

This ferrous cyanide is attacked by the excess of cyanide in the solution, and ferrocyanids of potassium is formed according to the reaction :

Fe Cy2 + 4KCy KUFeCye.

That is to say, one molecule of ferrous sulphate decomposes or renders inert six molecules of cyanide of potassium. Other things being equal, if i per cent, or 20 lbs. of ferrous cyanide existed in the ore, some 51 lbs. of cyanide would be rendered inert for the solution of gold, and, in fact, would be lost. This, at the average price of chemically pure cyanide, would cobt over jC to j£6 per ton of ore treated.

The ferrocyanide of potassium, formed according to the last reaction, is reacted upon, if sufficient acid be present, by an additional quantity of ferrous sulphate, with production of Prussian blue according to the reaction :

SKUFeCye + bFeSO* + 30 FejOa + 6K2SO4 + FcCyig.

This production of Prussian blue gives a blue colour to the surface of the tailings, or to the solution, and indicates at once that the washing and neutralizing operations have not been carried on properly, and that a great loss of cyanide is taking place.

Ferric salts, when present, unmixed with ferrous salts, de-

114 The Chemistry Of The Cyanide Process.

compose the cyanide solution with formation of hydrocyanic acid and precipitation of ferric hydrate, according to the re- action :

Fc,(S04)s + KCy Fe2Cy6 + jKaSO*.

With further decomposition :

FeaCye + 6H,0 Fe,(OH)e + 6HCy.

This means that, other things being equal, one molecule of ferric sulphate decomposes six molecules of cyanide. If i per cent., or 20 lbs., of ferric sulphate existed in the ore, very nearly the same weight of cyanide, costing £2 to would be destroyed.

If a mixture of ferric and ferrous sulphate, as is probable, exists in partially -oxidized ores, it causes the production, when ferrous sulphate is in excess, of ferrous cyanide, according to the reaction :

i2KCy + sFeSOi + 4Fe2(S04)8 — Fes(FeCy6)2 + 6K2SO4

when ferric sulphate is in excess, the production of ferric ferri- cyanide (Prussian blue), according to the reaction : —

iSKCy + sFeSOi + 2Fe2(S04)8 Fe4(FeCy6)3 + 9K2SO4.

These reactions show clearly that washing by water and neutralization by a caustic alkali must be employed to arrive at satisfactory and economical results. It is more than pro- bable that many of the failures already recorded are due to the lack of these precautions. In addition to these reactions, there are many with unknown compounds, the composition of which cannot be expressed, even where the greatest precautions are used, and the operations supervised with the greatest ability and knowledge.

Freoipitation of the Gold. — Zinc precipitates the dis- solved gold, as the cyanide has more affinity for it than for the gold. Theoretical reaction is :

2KAuCy2 -f Zn 2 Au + K2ZnCy4. But much more zinc goes into solution than this reaction calls

Precipitation Of The Gold. Ii5

for. According to the chemical reaction, the consumption should be i oz. of zinc to 6 oz. of gold. The excessive con- sumption of zinc must be ascribed to other action than the mere replacement of zinc for gold in the double cyanide of gold and potassium.

There is comparatively little exact knowledge of the reac- tions taking place in the zinc precipitation-boxes. One fact is known positively, and that is, that hydrogen is evolved. This does not occur, however, when zinc alone is exposed to a cyanide solution, but after gold is deposited on the zinc, or when zinc is placed in contact with iron. In other words, a galvanic couple is formed, the water is decomposed, and hydrate of zinc is formed, which is attacked by the cyanide forming a double cyanide of zinc and caustic potash. The probable reactions may be expressed as follows :

Zn + 2H20 2H + Zii(H0)2. Zn(H0)2 + 4KCy ZnK2Cy4 -f 2KHO.

The production of caustic alkali explains the increased alka- linity of the solution after passing the zinc precipitation-boxes. It may be considered advantageous to a certain extent, how- ever, as carbonic acid, which decomposes the solution, is absorbed by the caustic potash, with formation of a carbonate of the alkalies. Ammonia is fotmed also, as is indicated by the strong odour of the gas about the boxes.

The precipitate contains, besides the precious metals, many of the base metals, which may be dissolved by the solution. The principal of these are copper, arsenic, and antimony. When a weak cyanide solution contains copper, the copper may be precipitated in preference to the gold, whereas, by increasing the quantity of cyanide, the copper can be kept in solution until the precipitation of the gold is complete.

It is also asserted that, in ordinary cases, the accelerating influence on precipitation of excess ol potassic cyanide is pro- bably due to generation of nascent hydrogen.

4KCy -f Zn + 2H3O ZnK2Cy4 + K2H2O -f Hi.

Il6 THE CHEMISTRY OF THE CYANIDE PROCESS.

This nascent hydrogen steps into the place of the gold in the auro potassic cyanide :

2AuKCy2 + Ha 2KCy + K2H2O + Au.

The hydrocyanic acid thus formed recombining with any f-ee alkali present, there is no loss of such cyanogen as was combined with the gold ; from the former of the two equations it would appear that some proportion of the potassic cyanide must be consumed in the zinc boxes. As a matter of fact, there is a consumption in the case of strong cyanide solu- tions, which, however, in the case of ordinary working (it is claimed*), when solutions are coming off, 0*2 per cent, or so, is quite inappreciable. Indeed, it would appear as if a re- generation of the zinc potassic cyanide took place, the zinc possibly forming a hydrate, and remaining in solution as such, owing to the presence of the free alkali. Given favourable conditions, indeed, the zinc potassic cyanide is itself capable of dissolving gold from ores, and by addition of free alkali to this salt all the cyanogen in it may be determined in the ordi- nary manner by means of nitrate of silver solution.

Owing to the large amount of zinc which is dissolved in the precipitating-boxes, one would surmise that the working solu- tions would become in time very highly charged with zinc com- pounds. Mr. Feldtmann* remarks on this subject, that, as a matter of experience, it may be stated that they do not, to any great extent, and the probable reason for this is, that the small quantities of alkaline sulphides formed serve to precipitate, at least, a portion of the zinc as insoluble sulphide, a regeneration of potassic cyanide taking place simultaneously. ZnK2Cy4 -f K2S ZnS -f 4KCy.

The presence, or rather the formation, of alkaline sulphides in the solutions is explained by the action of potassic cyanide on the iron sulphide contained in partially-decomposed pyritous ores.

6KCy + FeS KiFeCye + K2S.

Notes on Gold Extraction by means of Cyanide of Potassium," by W. R. Feldtmann,

Decomposition Of The Cyanide. Ii7

Mr. T. S. MacArthur has even found that, in very excep- tional cases, sufficient alkaline sulphide may be formed to be of hindrance to the action of cyanide on the gold, and has dis- covered a remedy for this in the addition of metallic (parti- cularly lead) salts, capable of forming insoluble sulphides.

Deoomposition of the Cyanide. — The compound of cyanogen and potassium is extremely unstable. Not only is it decomposed by mineral acids and acid salts, but by the action at ordinary temperatures of atmospheric carbonic acid, accord- ing to the reaction : .

2KCy + COa + H2O K2CO3 + 2HCy.

Hydrocyanic acid is given off, a portion of which remains in solution and is available for the extraction of gold, but the greater part is dissipated into the air.

The cyanide is easily oxidized to cyanate :

KCy + KCyO.

The cyanate is further oxidized to carbonate according to the reaction :

2KCyO -I- 3O KjCOs + COa -f N,.

The nitrogen given off may cause a stiil further decomposition, for when a current of nitrogen is passed through a cold dilute solution of cyanide of potassium, hydrocyanic acid is evolved without the nitrogen entering into the reaction. This action, when the presence of a chemical causes a reaction between other chemicals in aqueous solution without entering into the reaction itself, is called hydrolysis, and further reaction must be attributed to this property of caustic alkalies, which are, and must be, always present in a working solution of potassium cyanide. If the solution is boiled with acids or alkalies, hydro- lysis of the cyanide occurs rapidly, ammonia and formic acid being formed thus :

KCN + 2OH2 NH3 -f HCO2K (+ 9*5)

According to T. K. Rose, B.Sc, "The Metallurgy of Gold." London, 1S94.

Il8 TIIE CHEMISTRY OF THE CYANIDE PROCESS.

If lime is in solution as cyanide of calcium, the following re- action may explain the formation of calcium carbonate :

Ca2(CN) + 30Ha + 2H CaCOa + 2H3N + CO2 + OH2.

Hydrogen is always evolved wnen the gold is precipitated ; without it the foregoing reaction could not be completed. This reaction accounts, moreover, for the formation of ammonia at the zinc boxes. The carbonic acid evolved attacks the cyanide, as mentioned before, thus causing still greater decom- position of the solvent.

It will be seen, therefore, that the decomposition of the cyanide solution, and loss of the solvent energy of the solution for gold, may be divided under the following heads —

I. Actual decomposition of the solution :

(a) By acids and acid salts present in the ore.

(d) By atmospheric carbonic acid.

{c) By oxidation.

(d) By reaction owing to hydrolysis.

a. The solution of other metals than gold which are not precipitated : —

(a) By metals or other compounds present in the ore, as oxides or carbonates of lead, which are first attacked by the caustic alkali ; by oxides or carbonates of the alkali ; by car* tain compounds of iron insoluble in water.

(d) By the replacement of gold in solution by zinc in the precipitation.

(c) By dissolving zinc hydrate, formed by electrolysis, in the precipitation.

It will thus be seen that these losses, under incompetent management, may be frequent, and in many cases may cause unprofitable results. .

Testing the Strength of Solutions. — The method em- ployed for estimating the amount of cyanide in a solution is based on the capacity of cyanide of potassium to form a double cyanide with silver — which is added to it in the shape of a titrated solution of nitrate of silver — and on the fact that any

Test For Strength Of Solution. Hq

silver solution which is added, beyond the exact quantity which is required, to convert all the potassic cyanide into argentic potassic cyanide, will cause a white precipitate.

2KCy + AgNOj AgKCyj + KNOj.*

As the combining weights of AgNOs and K.Cy. are 170 and 65*13 respectively, it follows that 170 parts by weight of argentic nitrate may be added to 2 by 65* 13 130*26 parts of potassic cyanide before a permanent precipitate ensues. If, therefore, we add from a burette a solution of argentic nitrate containing 17 . in a litre, or 1,000 cubic centimetres, to the solu- tion of potassic cyanide to be tested, until a faint precipitate appears, each cc. of silver solution added will correspond to

0*013 . of pure potassic cyanide. From the

amount of cyanide solution operated on, the percentage con- tents can be calculated. It is obvicns that the strength of the silver nitrate solution may be so adjusted as to save all calcu- lation.

If, for instance, it is made by dissolving 13*05 . of pure silver nitrate in a litre of water, and 10 cc. of the cyanide solution be taken for a test, then each cc. of silver solution added will correspond to 0*1 per cent, of pure K.Cy. in the sample tested. In testing very strong solutions — from the dissolving tank, for instance — one-tenth of the quantity of sample may be taken by measuring 10 cc, diluting with water to 100 cc, and then drawing ofif 10 cc. for test. Of course, in such a case, i cc. of standard silver solution will indicate i per cent, of K.Cy. in the original sample of cyanide solution. Addition of a few drops of potassic iodide to the solution to be tested will enhance the accuracy of the test, and will, moreover, annul the danger of our estimating the quantity of cyanide present, consequent on the strong alkalinity of the solution.

In estimating very dilute solutions, such as are employed, for instance, in the Siemens-Halske process, I would recom-

From Mr. Feldtmann, "Notes on Gold Extraction.*'

t20 THE CHEMISTRY OF THE CYANIDE PROCESS.

mend the employment of a standard decimal silver solution, at the same would ensure greater accuracy.

By taking lOo c.c. of the ordinary standard solution, diluting the same to i,ooo c.c, i cc. of this decimal solution would be equal to o'oi per cent, of K.Cy. The strength of the standard nitrate solution should be occasionally controlled by a standard cyanide solution.

In making the analysis, the solutions are filtered through a little quicklime. On addition of the silver nitrate a white curdy precipitate forms, which redissolves, and it is added till the solution shows a white precipitate.

An easier and more accurate method is by titration with a standard solution of iodine in potassium iodide until a blue coloration is apparent a starch solution having been added to the cyanide solution. The reaction is as follows : —

K.Cy + I2 KI + ICy.

The solution may be checked on chemically pure cyanide, or better yet, on sodium hyposulphite, and the cubic centimetres equivalent to i per cent, of potassium cyanide calculated.

If the percentage of zinc in the solution is required, the solution is evaporated to degrees and the residue treated by any of the well-known methods of analysis for zinc.

The cyanide employed is not chemically pure, and the commercial article contains about 80 to 90 per cent or over. It is customary in some works to dissolve in a small tank several hundredweight of the cyanide, and test the strength of this storage solution and make up from it the stock solutions. The lumps of cyanide are placed on a filter, made of a coarse sieve covered with jute. A pump is used for causing circulation of the solution.

The insoluble impurities, chiefly carbide of iron, contained in the commercial cyanide remain in the tray. The pumping of the solution into this tray should be kept going fast enough to keep the lumps of cyanide covered by solution, as it is found that alternate exposure of the carbide to the air and iiiimersion in solution of cyanide, causes a certain amount of

Test For Strength Of Solution. 121

decomposition of the latter. A water wash may be applied to remove the last trace of cyanide from the carbide before throw- ing the latter away. On no account should this carbide be put on the top of the sand in the filter vats, as is sometimes done* Having determined the strength and quantity of the stock solution, and the strength of the solution in the dissolving tank, the following is a simple formula for arriving at the quan- tity of the latter requisite to bring the former up to the desired strength : —

A being desired strength of stock solution in per cent ; B being present strength of stock solution in per cent, ; C the strength of dissolving tank solution in per cent. ; D the quantity in tons, lbs., gals., Htres, &c., &c., of stock;

3

y. . X D ssB quantity of dissolving tank solution to be added (in torj,

lbs., gals., litres, &c., &c.).

For example, supposing the stock solution to consist of 100,000 galls, of 0*4 per cent, strength, and it be desired to bring this up to 0.6 per cent, by adding some 10 per cent solution, then —

J X 100,000 2I27'65 gals, of the dissolving tank solution.

The following table (pp. 122, 123), which was kindly pre- pared for me by Mr. Blomfield, chemist of the works, will exemplify the strengths of the different solutions at the various stages while leaching from tailing-vat (this particular lot, which was treated at the Rand Central Ore Reduction works, being somewhat slimey, took a longer time to treat, i.e., about 60 hours) : —

See Feldtmann on " Gold Extraction.'*

122 The Chemistry Of The Cyanide Process.

TiMB or Lbachimo.

Original

Strbmoth or

Solution.

or

Solution after

Contact.

I hour 3 hours 5

7

1st Weak Solution o*i6 per cent.

0-05 per ceot. 005 „ 0075 „ o-i

There was consider- able moistuTe in the tailings previous to running on the ist weak solution.

After I

hour.

3

loun.

M 5

„ 7

9

ft

Ii

tt

13

Strong Solution. 0-35 per cent.

o*i per cent.

0*25

0*24 0*26

Draining dry.

1Est For Strength Of Solutiok.

Time ov Lbachino.

Original

Strength op

Solution.

Strength op

Solution apter

Contact.

After I hour. 3 hours.

1st Weak Wash. 0*16 percent.

0*26 per cent. 03 „ 033 "

Remarks.

I hour. 3 hours. 5 7

2nd Weak Wash. 0*16 per cent.

0-33 per cent. 03 „ 029 „ 022 „

I hour. 3 hours.

5

3rd Weak Wash. 0*16 per cent,

'o-2 per cent.

I hour. 3 hours. 5

4th Weak Wash. 0*16 per cent.

0-15 percent.

0-15 „

Final Water Wash.

I hour. 3 hours.

5

7 M

0*15 percent. 01 „ 013 „ 005 „

124 The Chemistry Of The Cyanide Process.

Determination of Gold in Cyanide Solutions.— Buchanan's method consists in precipitating a known quantity of solution with excess of argentine nitrate, decomposing the precipitate formed by means of a reducing agent, filtering, drying, and cupelling direct. In detail he found the best method of procedure to be as follows : — 195 c.c. of the cyanide solution are transferred to a flask of about 500 c.c. capacity* and mixed with a few drops of potassic chromate. Argentic nitrate solution of any convenient strength, say, 5 per cent., is tJien added until the characteristic reddish tinge of the chro- mate shows that the reaction is complete. Then take 10 to 20 . of zinc dust or shavings, mix them thoroughly with the precipitate and solution in the flask, and add 2 or 3 c.c oi 10 per cent sulphuric acid. Allow to stand for 10 minutes, add excess of sulphuric acid to dissolve the remainder of the zinc, filter, wash once, dry, and incinerate on a roasting dish in the muffle and cupel with a little lead. This method does away with the trouble of fluxing, and has the advantage of allowing of a great number of samples being done together. The results on the whole are slightly lower than those obtained by the ordinary precipitation method.

Crosses method consists in taking J or i litre of cyanide solution, containing gold, and adding excess of nitrate of silver. A precipitate of cyanide of silver is formed, and also argento- auro-cyanide, which is insoluble. This precipitate quickly settles, and is filtered on to a large filter. It is then put in a crucible covered with flux and, say, 500 grs. of litharge. In 10 minutes the fusion is complete, the button of lead obtained is cupelled, and the gold parted from the silver. This method allows of a larger quantity of liquid to be operated upon, is easily carried out, and does not require much watching. In the case of strong cyanide solutions most of the potassic cyanide is decomposed by addition of acid previous to adding the silver nitrate to avoid the otherwise heavy consumption of the latter.

Extraction Tests. — Mr. Feldtmann gives the following

feldtmann's extraction test. 125

instructions as to the best tests for determining if an ore is amenable to cyanide treatment.

It is assumed that the total sample is crushed fine enough to pass a 30-mesh sieve.

1. Assay a portion of the sample.

2. Determine the amount of cyanide it will consume by shaking test. For example : 200 . of ore are placed in a glass stoppered bottle with a 100 cc of solution of cyanide of o*5 per cent, strength, and shaken for twenty minutes or so.

A portion of the solution is then filtered off and tested. Supposing it to be reduced in strength to 0*4 per cent., show- ing a consumption of o'l per cent, on the solution, or half as much, 005 per cent, on the ore (or i lb. per ton), we might safely conclude that the ore will not require any preliminary treatment before leaching with cyanide. The largest con- sumption of cyanide takes place almost immediately after the olution comes in contact with the ore, and after twenty minutes* shaking it is generally safe to assume that there will not be much further consumption.

3. If it is found that the consumption of cyanide is exces- sive, a third portion of the ore is tested for cyanicide," by which is meant free acid, soluble and basic iron salts, and indeed any cyanide destroying substance which may be coun- teracted by alkali. A solution of caustic soda of known strength is run, little by little, from a burette into a weighed quantity of the ore mixed with water, the mixture being well stirred after each addition of alkali, until a drop, taken out on a glass rod, will just turn red litmus slightly blue. A con- venient quantity of ore to operate on is 200 ., and, using an alkali solution of 10 . commercial caustic soda to the litre of water, each cubic centimetre will correspond to iVth lb. of the same quality caustic soda required to the ton (of 2,000) of ore. If the consumption of soda be more than 3 lbs. per ton, it will generally be found advisable to water wash the ore before giving alkaline treatment. It is easy to determine' the amount of alkali which may be saved through a preliminary water wash, by first estimating total " cyanicide,"

126 The Chemistry Of The Cyanide Process.

then taking another sample, water washing first, and estimating remaining " cyanicide."

It is usual, when reporting the amount of **cyanicide" in an ore, to do so in terms of pounds and fractions of caustic soda required to neutralise a ton of the ore.

Should the consumption oi' cyanide in No. a test have proved larger than the amount of iron salts and acid present would account for, there is probably copper in the ore. The cyanide solution from test No. a may be conveniently exam- ined for copper by evaporating with nitric acid, taking up with a little more nitric acid, diluting and precipitating with ammonia, when copper will be indicated by the characteristic blue colouration of the liquid.

4. Several samples of ore are weighed out for extraction tests. A suitable vessel for testing extraction is a lamp glass, fitted with an india-rubber stopper, with a glass tube through it, which may be closed by means of a small piece of rubber tubing and burette clip. A filter is formed over the rubber stopper by means of a piece of sponge, some filter-paper, or some asbestos fibre.

Into such vessels the samples are placed — say, 200 . each— and they must then receive whatever preliminary treat- ment test No. 3 has shown to be needful in the way of water and alkali washes, 100 c.c. of a 0*5 per cent, solution of cyanide is then poured on. The various samples may be allowed to stand for different periods, say, one, two, and three days respectively, or more, if thought fit ; the cyanide solutions being then drawn off and tested, and the ore, after water washing to remove all dissolved gold, being assayed again.

Appendix.

THE CYANIDE PLANT OF THE GOLDEN GATE MILL OF THE DE LA MAR COMPANY AT MERCURy UTAH

Just as the present volume is being sent to press, there has come to hand the June 1898 number of Mining Machinery (Chicago, 111., U.S.A.), containing a description of the plans and methods pursued at the above-mentione d recently-erected mill, which appears to be of sufficient interest for insertion here. The illustrations are reduced from those appearing in Mining Machinery. The description reads as follows : —

There is now being completed at Mercur, Utah, one of the largest cyanide mills in the world. Its capacity, as now built, is 500 tons of ore in twenty-four hours, which will be increased in the future to 800 tons. This plant is notable from several stand- points. The application of the cyanide process to all of the various ores found in the Mercur district was a difficult problem to solve, and the development of a suitable process was all that was required to make this one of the most important gold camps on the Continent, because the immensity of the ore bodies has already been demonstrated. It is one of the largest reduction works in America. The power and light for the entire mine and mill are furnished by a high tension alternating current generated by water power in Provo Canon, and brought over the longest individual transmission in this country, which, as an achievement in electrical engineering, takes rank next after the Niagara Falls and Fresno, CaL, transmission plants.

A complete gas-producing plant furnishes the fuel for the

128 Appendix.

drying and the roasting furnaces, this being the first application on a large scale of producer gas to the metallurgy of the pre- cious metals.

It will be seen from the above that the completion of the Golden Gate Mill marks an important advance in the extrac- tion of gold from low-grade ores.

The Mercur Company was the pioneer in this district. Their mine is located at the apex of the limestone ledge, which dips in both directions at an angle of about 45 degrees. After many vicissitudes, this company began operations with an 80-ton cyanide mill. This was the first practical application of the cyanide process in the United States. In the Mercur mine the limestone is very soft, and porous, and full of seams. The gold is found in the crevices, having probably been deposited there in some former age, by the seepage of solutions in liquid or gaseous form.

Nature of the Ore. — The ore is coarsely crushed, and delivered to the leaching tanks without intermediate treat- ment. Being oxidised and very porous, this ore is an ideal one for simple cyaniding. The values are extracted cheaply, but the percentage of extraction is not as high as it would be if a roasting plant were added to the equipment In the Golden Gate mine the ledge is tapped at a depth of 1,500 feet greater than in the Mercur mine. Much of the ore contains either arsenic or sulphur, and is therefore not amenable to the simple cyanide process. For two years the De La Mar Company carried on an extensive series of experiments with the view of finding a modification of the process that could be successfully applied to their ores. In the meantime the company has been gradually adding to its holdings. The most recent addition is less than seven acres in extent, and more than a quarter of a million dollars was paid for it. The development work has gone on continuously, and enormous bodies of ore have been blocked out.

Like other ores found in such large bodies, the De La Mar ore is low grade, running from 7 dollars to 12 dollars per ton.

Appendix.

Process Employed. The treatment, as finally developed by the company's metallurgists and chemists, is the McArthur- Forrest cyanide process modified, and while the details have

not been made public, the general knowledge that the most important problem of the camp has been solved has stimulated development work to a remarkable degree.

130 Appendix.

The Steel Mill Building.— The mill building is a mon- strous structure of steel, with huge ore bins magnificently heavy, and large roof trusses gracefully light. It is admirably adapted to its purpose, being conveniently arranged and well lighted, and is a good example of the great superiority of steel over wood for plants of this character.

The mill is built on the side of a 30-degree hill, and an idea of its size and imposing appearance may be gained from the fact that it is 160 feet from the lowest to the highest floor level, and that the building is 380 feet long and 295 feet wide. The building, together with the other structural work in the plant, was designed and erected by the Gillett-Herzog Com- pany, of Minneapolis, Minn. The complete plant includes the main mill, which is divided into six parts, for coarse nish- ing, drying, line crushing, roasting, leaching and precipitation ; the gas plant, located on one side of the mill; the assay house and laboratory, located on the other side ; the hoisting plant, just behind the mill, and the transformer house behind that.

The ore is brought from the mme through an inclined shaft, and from the mouth of the shaft over a three-track inclined steel bridge to the top of the coarse crushing building, which is about 100 feet high. Two of the tracks are utilised for the hoisting of ore, while the third is used for the man hoist. The hoists were built by Webster, Camp, and Lane, and are driven by Westinghouse induction motors.

The Crusher House. — At the top of the crusher house the ore is dumped into one of three enormous steel bins having a combined capacity of 2,000 tons. The bins containing the sulphide and arsenical ores each discharge into the hopper of a No. 6 Gates crusher, which reduces the ore to about li inch mesh, and discharges it into another set of bins of the same capacity as those above. . . . Each No. 6 is capable of crushing 100 tons of Mercur ore per hour. The crushing head being in the shape of a truncated cone, the size of the product is readily varied by raising or lowering the shafts. These two crushers

l UNIVERSITY APPENDIxS Aui FORH

are also driven by a Westinghousc induction motor, which is the t]rpe employed for driving all the machinery in the mill.

jiifl ,

Driflfw Turiocu

"T

f-

L-

Lco.ekv'M' Dep't.

S.

.. JfftUL"

.P.

Fig. 40.— Ground Plan or Mill.

Dryers for Sulphide and Arsenioal Ores. — The sul- phide and arsenical ore from the second set of bins in the crusher building is discharged into two Brown straight drying furnaces. These dryers occupy the second portion of the mill.

132 Appendix.

and are located near the side walls, leaving room in the centre for an additional dryer when the capacity of the mill is increased to 800 tons per day. The dryers are of the reverberatory type, with hearths 60 feet long and 12 feet wide, and they are fired by gas. They have a nominal capacity of 175 tons each per day, but if necessary can handle more than 250 tons each. Only the arsenical and sulphide ores are run through the dryers, one furnace being reserved for each kind.

The Belt Conyeying System. — The silicious ore is carried from the storage bin in the crusher house to the rolls by a Robbins belt conveyor. This is a moving trough, the belt being bent up at the sides by rollers. When run at a moderate speed, a very small Robbins belt has a remarkably large conveying capacity. To insure uniform loading, the ore is fed to the belt conveyer by a Gates roll feeder. This feeder is described in connection with the rolls. The placing of a suitable feeder at every point where ore is transferred from a bin to a conveyor or from one conveying device to another is a good investment, as it increases the life of the conveyers by keeping them uniformly loaded at a capacity at which they operate most economically, and it saves a great deal of labour.

The Bolls and their Operation. — The third compart- ment of the mill contains the fine-crushing machinery, which is divided into three units, one for each kind of ore, but all of it receiving the same treatment. The ore is first sized in a Ber- thellet separator, which is an inclined covered screen, the cover being lined on the inside with rubber belting so arranged that the ore impinging on the screen strikes the rubber belt, and being thrown back on to the screen again, the continuation of this operation keeps the screen in a vibratory state. This permits a fine separation to be made, and as the ore particles strike the screen at an angle of 45 degrees a much coarser screen can be used for a given sized product than that indi- cated by the size of the screened ore. The oversize from the Berthellet separators passes through one of the three sets of

Appendix.

36 X 15 Gates rolls, and then into one or two of four sets

of 26 X 15 Gates rolls.

The fines from the

Berthellet separator

ore also passed

through this latter

set of rolls.

An important feature of these rolls lies in the construc- tion of the roll bear- ings, a feature that has no doubt con- tributed largely to their popularity. The swivel pillow blocks in which the roll shafts have their bearings are solid castings, and are slipped over the ends of the shafts. They are babbitted on one side only, the bearing being unusually long. Spaces are left on the opposite side of the pillow block, into which oil-soaked waste is inserted. This effectually lubri- . cates the bearings, /p 5[ and keeps them con- p stantly free from grit, etc. Two of these pillow blocks are placed in sliding pedestals with wide bearings.

134 Appendix.

It is a well-known fact that it is almost impossible to feed rolls in such a manner that a perfectly even strain is at all times sustained by the boxes, but this form of construction (which is a patented feature) will allow the rolls to separate more at one end than at the other, and no binding will result, the shaft remaining at all times parallel with its bearings. When the necessity arises of removing the shaft, this may be accomplished without disturbing the spring adjustment of the rolls, as the shaft together with the pillow blocks intact are lifted out with but little loss of time. A discharge hopper and spout are located beneath and form part of the machine, the spout being arranged in such a manner that the product of the rolls may be directed towards the front, back, or to either side of the machine. These rolls are equipped with Gates automatic roll feeders, which are attached directly to the housings

The Boasting Fnmaoes. — From the rolls the ore is elevated to revolving screens, from which the oversize is sent back to the rolls, and the ore that has been crushed to lo-mesh into one of three bins of i,ooo tons capacity. The ore from these last bins is carried on Robbins belt conveyors to the roasting furnaces. The roasting department of the mill is divided into three levels, each of which is 33 feet wide and 295 feet long. Two of these levels are occupied by the four furnaces now in use, the third one being reserved for future enlargement. The roasting furnaces are of the reverberatory type, designed by Horace F. Brown, in which the operating mechanism of the stirring carriages is carried in side compart- ments cut off from the main furnace hearth by a slotted Wall, which preserves the moving parts from heat and dirt. These furnaces are each 12 feet wide by 100 feet long, and, like the dryers, are fired by a mixture of producer and water gas, furnished by the Loomis gas plant. The products of com- bustion from the four roasting furnaces, together with the sulphur and arsenic fumes, pass into a dust chamber and then into a large flue running underground up the hill. The pro-

Appendix.

ducts of combustion from the two dryers also discharge into this flue, -which termi- nates further up the hill in a steel smoke- stack. The ore, after roasting on the main hearth of the furnace, is elevated to a cool- ing hearth built over thearchof the furnace and running itsentire length, is discharged into a trough and conveyed by a spiral conveyor laterally to the side of the fur- nace, to a spout, down which it travels to a storage bin cut out under the furnace floor in the centre of the mill.

The silicious ore is discharged into this bin directly from the steel storage bins without roasting.

The Gas Plant, and its Advantages over Solid Fuel. —

The Loomis gas plant is located out- side of the mill build- ing on the level of the roasting furnaces. It contains four generators and two coolers, one cooler being connected to two generators. In

K

Appendix.

the Loomis process, producer and water gas are made alter- nately, first one and then the other. While one of a pair of generators is making water gas the other is making producer gas. Air is first drawn down through the fire and the coal in the generator is brought to a state of incandescence ; then the air blast is shut off and the top doors are closed, and steam is ad- mitted into the bottom of the generator. This steam, in passing through the in- candescent bed of fuel, is converted into water gas, which is drawn off at the top and passed through the cooler, where it gives up its heat in generating steam. It is then sent cold to the holder.

At the end of about five minutes the process is reversed by shutting off the steam supply at the bottom and introducing it at the top, the door at the top is opened and air is drawn in, which with the steam passes through the fuel. This generates a fixed producer gas which is

entirely free from tar. It is passed through the cooler and

sent to the holder to mix with the water gas.

Appendix. 137

The use of gas for roasting and drying has effected a saving of at least 30 per cent, in fuel, in comparison with the direct burning of coal or wood in the furnace fireboxes. Where solid fuel is employed, it is almost impossible to maintain uniform conditions in a roasting furnace, except by firing lightly and frequently, which means' increased labour. In this plant, when the valves have once been properly set, they require no further attention until the quantity or character of ore passing through the furnaces is changed. The gas is carried from the holder to the furnaces in spiral riveted steel pipe, made by the Abendroth and Root Company. There being no radiation losses to prevent, the pipes are not lined.

The Leaching Tanks. — The next level in the mill con- tains ten 25 X 50 feet leaching tanks 5. feet deep. From the level of the top of these tanks a tunnel runs under the ore bin, in which the residue of the silicious ore is stored. The proper amount of each kind of ore is drawn from the storage bin and carried in cars through the tunnel to the tank being filled. Men with shovels spread the ore evenly after it is dumped from the cars. The cyanide solution is then turned on to the ore, and, after leaching out the gold, passes through the filter which covers the bottom of the tank, and is drawn off, to the precipitating tanks, which occupy the last level of the mill.

After the cyanide solution has been drawn off, and the ore washed with several applications of water to remove the re- mainder of the solution, the tailings are shovelled through two trap-doors in the bottom of the tanks into cars, which convey these tailings to the dump.

Index.

ACCUMULATED slimes, solu- tion of gold in, 52 Add ores and tailings, 51

adverse conditions affecting treatment, 63 Alkaline solution, 35 Anodes, 70

Antimonite in ores, 104 Argall, Mr. Philip, on precipita- tion, 38 on dry crushing, 94

BALL Mill Gruson works, 62 Battery slimes, treatment of, Bendigo, cyanide works at, 25 Betty, Mr. W., 6, 13, 38, 63 Black Reef, 63 Blomfield, Mr., of Rand Central

Ore Reduction Company, 85 Bottom of yats, 15 Butters, Mr. Charles, 12 patent discharge lid, 19 on plant at May Consolidated works, 98 Butters and Mein*s distributor, 8 Butters and Smart, Messrs., on pipes, 20 on pumps, 20

CALDECOTT, Mr. W. A., on solution of gold in slimes, 52 Cathodes, 68, 69 Caustic lime, 52

wash, 35 Charcoal used for predpitation, 25 Charging and discharging, 12 Clean-up, 39

Concentrates, treatment of, 58 Concentration, 58 Conditions, adverse, 63

influencmg precipitation, 38 Cost of plant, 66

of treatment, 65 Crown Reef mine, 5, 58, 84 Crushing dry, 94 Cyanidde, 125 Cyanide, consumption of, 39

determination of cause of con- sumption of, 10 1

decomposition of, 1 13

non-extraction, cause of, 102

plant, erection of, I

process, chemistry of, 105 synopsis of, 34

solutions, 35

known in middle ages, iz. Cyanide treatment, adverse con* ditions affecting, 63

Index.

DEEBLE'S cyanide works, 25 patent pan, 26 DetermiDation of gold in solutions,

Buchanan's method, 124 Cross's method, 120 of C3ranide consumption, 10 1 of non-extraction, 102 Direct filling, advantages and dis- advantages of, 13 Discharge lid. Butters' patent, 19

launders, grade of, 3 Discharging tailings 18 Distributor, Butters and Mein's, 8

sizes of, 12 Dry crushing, 94

ELECTRIC current, action of, on gold solutions, 68 required for precipitation, 70 precipitation, 67, 71 at the Gibraltar works, 97 Extraction tests, 124

FELDTMANN, Mr., on "Gold Extraction," 18, 37, 43, 116, Filter-vats or leaching tanks, 14

number required, 7 Fluxing of gold slimes, 44 Frue vanners, 58

GERMAN Government, report to, xii Gibraltar cyanide works, New South Wales, 96

Goertz, Mr. Ad., on gold slimes, 90

Grold slimes, 44

Gold, at the Treasury mine, treat- ment of, 47 Goertz on the treatment of, 60 Williams on the treatment of, 55 recovery of, from slimes, 92

Gregory's concentrates, 1 1 1

Ttaulage,2

Messrs. Fraser and Chalmers' SjTstem of, 2 Hydraulic separators, 13

INTERMEDIATE filling, 8 1 advantages of, 12

JANIN, Mr. Louis, experiments on silver ores, 109 Jennings, Mr. H., 4

LANGLAAGTE Block B, 34 Langlaagte Estate and Gold Mining Company, 7, 58, 93 Leaching tanks, 14 Lydenbiu-g cupriferous ores, 26

MACARTHUR- FORREST Company, public indebted- ness to, xi May Consolidated works cyanide

plant at, FrotUupiect Mechanical haulage, 2, 93 Mercur mine, Utah, cyanide plant

at, 127 Mercury and electric deposition, 68 Metal cathodes, 69

Index.

Metallurgy, methods of, vii Meyer and Charlton Gold Mining

Company, ii Muffle roasting furnace, 43

N

ON-EXTRACTIONofcyanide,

rVPERATIONS at various works, Oswell, Mr., of Worcester works, 72

PAN amalgamation, yield of, ziii Pan, Deeble's patent, 26 Paarl Central Works, 5 Park and Skey, Messrs., on anti-

monite in ores, 104 Pioneer Crold Mining Company, 11 Pipes, arrangement of, 20

Messrs. Butters and Smart on, Plan of works for cyanide plant, i Plant, 7

cost of, 66

at May Consolidated works, 98 at Mercur mine, Utah, 127 Potassium cyanide, experiments

with. III Practical results, 64 Precipitate, roasting the, 43 Precipitation of gold, 36, 91, 1 14 conditions influencing, 38 boxes, construction of, 22 on charcoal, 25

electric, 50, 67, 71, 72, 97, 100 Princess works, 3

and Worcester works, zinc process, 40

Process, zinc, time required, 7

synopsis of, 34 Pumps, 19

Messrs. Butters and Smart on,

Pyritic ores, treatment of, iii, 112

RADEMACHER, Mr., of Hum- boldt works, 58 Rand Cential works, 38, 93 Recovery of gold from slimes, 92 Results of processes, 64, 73 Roasting the precipitate, 43 Robinson mine, 60, 66 Rothschild, Messrs., report to, zi

SCHEME of working, 77 Schmeisser, Bergrath, report by, xii Siemens-Halske process, 67

results of, 72 Silver ores, experiments on, 105 results of, 72 scheme of working, 77 Simmer and Jack works, 88 working of the plant, 89 yield of concentrates, xiii Skey and Park, Messrs., on anti-

monite in ores, 104

Slimes, elimination of, 3

smelting of, 44

solution of gold in accumu- lated, 52 Mr. Ad. Groerz on, 60 Mr. T. R. Williams on, 5, 55 treatment of, at Treasury Mine, Johannesburg, 47

Index.

Smart and Torrente, Messrs., 88 Smith, Mr. Hamilton, report by, zi Solubility of metals and minerals

othw than gold, io8 Solution drainage pipes, 20

of the gold, 91, loi

of gold in accumulated and other slimes, 52 Solutions, testing strength of, 118 Solvents for gold, iz Stamping, effect of, on ores, 5 Stock solution tanks, 21

AILING dams jnd resenroin, 2 1 pumps, 3

wheels, 3 Tailings, treatment of, 38, 51 Testing strength of solutions, 118 Torrente and Smart, Messrs., SS Treasury mine, Johannesburg, treat- ment of slimes at, 47 Treatment, cost of, 65

of battery slimes, 55, 60

of gold slimes. Treasury mine,

Mr. T. R. Williams on, 55 Mr. Ad. Goertz on, 60

Treatment of tailings, time taken by,

VATS required, number of, 7 calculating contents of, 22 methods of filling, 7, 13 Von Gemet, Mr., and the Siemens- Halske process, 67

WATER wash, 36 AVUHams, Mr. T. R., of Crown Reef mine, on slimes, 5, 55 on the treatment of gold slimes,

Witwatersrand, production of gold

in, 14 Worcester Gold Mining Co., 11 works, working scheme at, 77 Working, scheme of, 77

ZINC for cyanide process, 47 precipitation boxes, 22 at the Princess and Worcester

works, 40 process, time required, 7 shavings, 37

University

PKIKTBD BT J. S. TI&TUX AND 00., LIMITSD, GXTT BOAD, TXn>OV.

, -Cjigitized by

/

University Op California Library

This Book Is Due Ok The Last Date Stamped Below

JAN Id liitO FE 20 1915

Dec 18 1993

A;V!i:'.CClc

ilOV291999

Jun 2 9 2002

g:c[)'33

30m-6,'14