circulate round and round the milling system, only a small portion being eliminated each time as finished product. The cost of reducing this residuum is disproportionately high and in many instances will not result in a final profit.

It sometimes happens too that this tough residuum contains a lower assay value than any other portion of the ore, making it still more unprofitable to carry the process to a conclusion.

The amount of oversize finally allowed to pass to treatment will determine the point as to whether it is worth while to separate a part to be leached or to treat the whole by agitation. As a rule when the amount of coarse material is not sufficient to interfere with the mechanical working of the agitators and yields its maximum extraction within the time available, it will pay better to treat the whole as if it were slime.

What has been said thus far deals with those ores on which allsliming has been indicated as the ideal treatment. Statements may often be read to the effect that while gold ores can be treated advantageously by a combination of leaching and agitation, for silver ores all-sliming is the only way to obtain a satisfactory extraction. Like most generalizations, these statements are to be received with caution. While it is probably true to say that most silver ores require finer grinding than the majority of gold ores, yet in the writer's experience the silver ores that unquestionably need "all-slime" treatment are in the minority. Very many of them show no appreciable gain in extraction when the -100 mesh mark has been passed, and with some the limit may be put at -80 or even coarser. In cases such as these it is difficult to see any point in trying to make an "all-slime" product. If the costs are increased (as they undoubtedly will be) without resulting in any increased recovery, the commercial result will be a loss.

Even when the finer comminution does show a substantial increase in extraction the coarser grinding may result in a greater net profit on account of lower costs and larger output from any given crushing unit.

If the degree of comminution decided on results in a pulp 50% of which will remain on a 200 sieve, separation of the sand for

leaching will almost always be advisable. Early in the development of the cyanide process agitation methods were tried for sand, but were abandoned on account of the expense and mechanical difficulties involved, and the same objections hold good today, though perhaps not quite to the same extent.

Coagulation of the Slime for Settlement.-Since the slime treatment is conducted at a much lower ratio of liquid to solid than that usually employed in the operations of milling and classification there must be a dewatering of the slime between the mill and the cyanide plant. Such dewatering is almost always effected by settlement and decantation but owing to the invariable presence in the pulp of amorphous or clayey matter, it is impossible to obtain a settlement accompanied by a clear effluent liquor without the presence of some substance which will cause a coagulation or flocculation of this clayey material. Such substances or the solutions resulting from them are known as electrolytes, in view of the electrical nature of the phenomena exhibited by finely divided matter in suspension.

Action of Electrolytes.-The substance most commonly used for this purpose is lime both on account of its cheapness and also because of its usefulness as an alkaline protector for the cyanide. Many other substances have a similar effect though varying in degree, such as alum, calcium carbonate, sulphuric acid, ferrous and ferric sulphate, and calcium sulphate. Lime usually forms a coarser grained and more flocculent curd with a more brilliantly clear supernatant liquor than any other electrolyte, but the precipitated mass does not settle as densely as that produced by some of the others. Caustic soda in some cases acts as a coagulant and while the liquor produced is not quite as clear as that resulting from lime, the curd has a finer grain and settles more slowly but much more densely. Sodium carbonate, on the other hand, tends to prevent coagulation and renders clarification almost impossible.

Addition of Lime.—When milling is done in cyanide solution the whole of the lime necessary is usually added to the ore before it enters the mill; by this means it gets pulverized with the ore

and performs its function of protective alkali. Milling in water is (1919) seldom employed except when it is desired to pass the pulp over amalgamated plates before cyanidation. Even in these circumstances the lime is often added in the mill, notably in the Rand practice. When the method was first adopted there it was strongly objected to by the mill men on the ground that it hardened the amalgam. As soon, however, as they got accustomed to the changed conditions, it was found that no detrimental results followed; in fact, in some cases a marked increase occurred in the amount of gold recovered on the plates due to an improved amalgamation of the fine gold in the slime. As a consequence certain slimes which it had been designed to treat by cyanide were now found to be low enough in gold content to be run to waste. In other parts of the world the plan of amalgamating in lime water has not commended itself after fair trial and the necessary quantity of lime is added (usually in the form of milk of lime) to the ground ore after amalgamation, or even later on, as at the Homestake, after the separation of sand and slime.

The amount of lime added varies of course with the ore, but for satisfactory settlement of the slime a strength in the solution of 0.02% CaO is usually sufficient. When milling in cyanide the quantity of lime maintained in solution is regulated rather with a view to protective alkalinity and is usually considerably in excess of the bare amount needed for slime settlement.

Thickening the Pulp. The slime pulp on leaving the sand classifier with its proper admixture of lime is first led to some form of thickening device if, as is usually the case, the ratio of liquid to solid is too large for economic handling in the cyanide treatment. The form of thickener almost universally used in the early days consisted of a large spitzkasten composed of a number of pointed boxes arranged as a unit in such a way as to give a continuous superficial area.

At the Minas Prietas plant of Chas. Butters and Co., capable of treating 9000 tons of mixed sand and slime tailings per month, the slime spitzkasten was 36 ft. by 42 ft. composed of 42 pointed boxes 6 ft. square and 4' 9" deep; at the apex of each box was a

centre line with small holes bored at intervals for the exit of the pulp, and the clear solution overflowed a weir at both sides, so

nipple and hose discharging into a common launder. The slime pulp entered by a launder stretching from end to end along the

that on each side of the feed launder there were three rows of seven boxes for the settlement of the slime. The overflow weirs were raised sufficiently to submerge the partition walls of the various compartments to a depth of nine inches, giving a continuous surface area of solution of 1512 sq. ft. and a depth of 5′ 6′′ from the apex of each compartment to the surface of the solution. The apex discharges of the two rows nearest the intake were allowed to flow continuously, the discharges of the second rows were closed off part of the time and the third rows were only opened for a few hours each day, the general regulation being so arranged as to give as thick a pulp as possible in the underflow while maintaining a clear overflow for return to the mill (or in this particular instance, to the pulping tanks).

This thickened pulp was at times further dewatered in a collecting tank, usually a large circular tank with peripheral overflow and fitted with mechanical stirring arms. The pulp enters at the centre and when the tank is full the slime settles and a clear liquor overflows into the ring-launder to be returned to the mill. The pulp continues to flow in until the tank contains sufficient slime to constitute a treatment charge, when the stream is diverted to another collecting tank.

A very efficient modification of this type of collecting tank is the Dorr thickener. The entry of pulp and overflow of clear liquor are the same but in the Dorr system the thickened product is continuously removed from the bottom by means of slowly revolving rabbles which convey it to the centre where it falls into a discharge pipe whose aperture can be adjusted to regulate the speed of outflow. The device works smoothly and continuously and has superseded all the older methods for thickening slime pulp.

Regarding the capacity of Dorr thickeners J. V. N. Dorr in a paper presented to the American Institute of Mining Engineers, August, 1914, says,

"The capacity of Dorr thickeners has been found to be primarily a function of area, although the depth of the tank has an influence depending on the dilution of the feed and the dilution of the underflow