PULP TO MAKE One Ton of DRY SLIME-(Continued) 2099 2107 2115 2122 2129 2136 1010 6.77 351.1 6.80 352.4 2.78 234.1 2.79 235.0 1050 103.9 1053 104.3 1057 104.7 1061 105.0 1065 105.5 1068 105.8 699.8 34.4 702.3 34.5 704.8 34.7707.3 34.8 709.7 34.9712.0 35.1 524.8 17.14 526.7 17.21 528.6 17.27 530.4 17.33 532.2 17.39 534.0 17.46 419.9 10.22 421.4 10.26 422.9 10.30 424.4 10.34 425.8 10.38 427.2 10.42 349.9 299.9 262.4 233.3 1020 1030 1040 1050 6.83 353.6 6.86 354.8 4.85 303.1 4.87 304.2 3.62 265.2 3.64 266.1 CHAPTER VIII DISCREPANCIES BETWEEN ACTUAL RECOVERY SECTION III SAMPLING OF MILL HEADS AND RESIDUES Sampling of Heads. This is the second of the more fertile causes of failure to check on bullion recovery. When milling in water the most reliable sample of ore heads is probably that taken from the pulp after it leaves the mortar boxes. This may be done by catching a portion in a cup or other receptacle at regular intervals, and if care is taken that the whole depth and width of the stream issuing from the screens is included, that the cup is not allowed to overflow, and that the intervals are regularly timed and sufficiently frequent, the resulting sample is usually reliable. A better way is to place an automatic sampler in the stream of pulp, thus eliminating the personal factor, but great care is necessary to see that the sampler used is correctly designed to cut a perfect section of the stream both in breadth and depth, because a stream of pulp flowing in a launder is not homogenous, and moreover the effect of a slight error in the product of a mechanical sampler is cumulative, that is, it is invariably in the same direction and is not subject to the correction which may automatically occur in other circumstances where errors in excess and defect may neutralize one another over a period and result in a representative average reading. An automatic sampler that satisfactorily withstood very severe and searching tests on the Rand is that invented by James Higham and described by him in the Journal of the Chem., Metall. & Min. Soc. of S. A. for July, 1903. It will be seen from the illustration that it is essentially a short length of curved and slotted pipe that is made to pass across the stream of pulp by means of the screw mechanism which is power driven, and the direction of travel automatically reversible. A break is made in the pulp launder and the stream in dropping from FIG. 26.-Section of Cutter Pipe, Higham Pulp Sampler. the upper to the lower level encounters the slotted pipe which cuts out a section of it and delivers it to a suitable receptacle placed underneath. The normal size of sample taken is about 2000 lb. per day from a 100 stamp mill, or 1 ton from a daily tonnage of presumably about 1000 tons of ore. The only objection raised to this sampler has been the unwieldy size of the resulting sample but this may be obviated by inserting a second sampler below the first to reduce the quantity taken to convenient dimensions, or by the use of the Higham Dividing Box, a trough fitted with a number of knife edged slots, at right angles to it, acting on the same principle as the Jones dry ore sampler. The trough is placed parallel with the direction of movement of the sampling pipe and as the outflowing pulp passes over the slots a definite proportion of the stream enters and is received in a second trough underneath, the remainder or reject flowing away to join the main stream. Any machine, however, will answer the purpose that is so designed as to cut a perfect section of the stream both horizontally and vertically without appreciable acceleration during any part of the stroke, and with a minimum of splash and diversion of the line of flow. When milling in cyanide solution it is difficult to obtain a head assay from the pulp owing to part of the gold passing rapidly into solution. Methods have been suggested on the Rand and are said to have been carried out with a certain measure of success, whereby the true value of a pulp sample taken under these conditions is calculated by ascertaining the ratio of solution to ore in the sample, and the assay value of the solution before entering the mill and also as present in the pulp, and then adding on to the assay of the ore the equivalent of the gold dissolved during milling as shown by the solution assay. This method, while it seems theoretically capable of giving correct results, yet is open to so many possibilities of error and needs so much care and time expended on it that it hardly commends itself as a practicable everyday solution of the problem. The only alternative is to sample the ore before it comes in contact with the solution, and if properly done, this method may be made to yield reliable working results. It is possible to set about it in two ways, first, to install a complete sampling plant in the mill, cutting out a section of the stream of broken ore going to the bin and subdividing it by machines interspersed with crushing devices for reducing the size of the particles after each quartering until a small but representative assay sample remains. This, while satisfactory, is costly, both to install and to run, and some important mills which adopted the system have since abandoned it as not yielding any adequate compensation for the expense involved. The second method is to cut out small individual samples of the ore periodically at some point after it leaves the breakers. If means can be found to do this regularly and fairly, a good representative sample may be obtained. The most reliable system is of course an automatic sampler if it is practicable to insert one somewhere between the rock breaker and the intake of the mortar box, but hand samples faithfully and regularly taken are on the average dependable. An important point to be observed is that as long as the ore is coarsely broken, the aggregate sample for the day or shift must be a large one or it will be of no use. When cutting out, at regular intervals of half an hour, samples of a two-inch rockbreaker product in a mill crushing 100 tons a day, an aggregate sample of 1000 lb. for the twentyfour hours is not any too much, and 2000 lb. would be safer. This should then be all crushed to about 1/4 inch size before being quartered down to a smaller bulk. The mixing and quartering may be done by hand shovelling or one of the various mechanical devices may be used. When the bulk is reduced to about 100 lb. this should then be crushed to pass 1 inch screen and quartering continued till the sample weighs about 20 lb. when it should be ground to pass 10 mesh; in this condition it may be reduced in bulk to one or two pounds and then sent to the assay office where the process is continued until the necessary assay sample of 80 or 100 mesh is obtained. This may seem an unnecessarily expensive and laborious process, but it cannot be shortened with safety if the resulting assays are expected to furnish a basis on which to figure bullion returns. An unusually significant example of this contention, drawn from the writer's own experience, is here given by way of illustra |