The jig, in one form or another, continues to hold a leading place among the machines designed to separate two or more minerals of different specific gravities. It is simple in construction, easily operated, capable of treating large quantities in a short time, and highly efficient under various conditions.

The question, whether the material to be jigged has first been sized, determines the two principal methods of jigging. Jigging preceded by close sizing, generally known as the Continental or German system, involves a more or less elaborate series of screens or trommels, with attendant cost for installation, operation, and repairs. Jigging without sizing, known as the English system, is, according to Munroe, “ a development of the hand-jigging formerly employed in Cornwall and introduced by English miners to this country.” In its simplest form, the method consists in jigging an ore-mixture previously crushed to some maximum size (although, in some cases, even this preliminary is omitted) on a relatively coarse sieve, and then jigging again on a finer sieve, the material passing through the first sieve and bedding. While many modifications have been necessary to adapt it for use in mills of large capacity, where hand-work was necessarily replaced by machines, the principle remains the same; the fact that the English system has been successfully employed, both in this country and abroad, is well known; and arguments have been made for its efficiency and applicability in a wider sphere than it has occupied hitherto.

Testing Minerals for Jig Recoverability

The fact that treating a mixture of minerals under jigging conditions increased the amount of mineral saved; or, as Professor Richards aptly terms it, “ the extra jig-catch,” has long been known. To account for this fact a number of theories have been proposed. The work of Rittinger in this field has, for many years, been a classic in the literature of ore-dressing. For the purposes of my present paper, however, the work of two American investigators, Prof. H. S. Munroe and Prof. R. H. Richards, is chiefly concerned.

Professor Munroe has given the results of an elaborate series of experiments, and his deductions, based largely on theoretical grounds, of this work. After reviewing briefly the two systems of jigging, followed by a discussion of Rittinger’s formulas and the derivation of them, and after a careful study of the behavior of grains (usually shot) in a tube en masse, acted upon by a rising current of water, he is led to conclude that the interstitial currents play a very important role, and are responsible for the high ratios of concentration obtainable in the English system of jigging. Since his conclusions bear directly upon the present investigation, they are given in full, as follows :

1. Bodies falling through water in a tube do not attain as high a velocity as in falling through the same medium in large vessels.
2. The falling velocity is but little affected when the diameter of the body is less than one-tenth that of the tube.
3. The falling velocity is the more retarded as the diameter of the body approximates that of the tube.
4. A sphere four-tenths the size of the tube will develop the greatest falling velocity, and will require a current of maximum velocity to support or raise it.
5. Grains falling en masse are really moving in confined channels, and follow the law of the movement of bodies in tubes. The falling velocity, and the velocity of the current necessary to support or raise a mass of grains, increase and diminish with the distance apart of the grains.
6. The diameter of the channel in which the single grain moves equals the cube root of the volume of the grain with its proportion of the interstitial space.
7. In a mass of grains of different sizes, the large grains move relatively in smaller channels than the small grains. The ratio of the diameters of equal-falling grains of quartz and galena, under such conditions, is 31 to 1, instead of 4 to 1, which latter ratio holds good for free-falling grains only.
8. The formulae for grains moving in tubes, when applied as above to grains moving en masse, enable us to compute the velocity of jig-currents and thus determine the proper length and number of strokes of the jig-piston. The old formulae gave results many times too large.
9. The present investigation demonstrates that close sizing is not necessary for the separation of different minerals by jigging, unless the difference in specific gravity is small.
10. Downward currents are apparently necessary to success in jigging through a bed. This requires confirmation by experiments on a larger scale.
11. Very fine material, less than 1/10 millimeter in diameter, can be treated successfully on jigs, if treated with coarse stuff, the concentration taking place in the small interstitial channels between the grains forming the mineral bed. For the treatment of fine stuff on jigs, close sizing is a positive disadvantage. Jigs work well on mixed stuff, and very badly on fine stuff alone. Stuff less than four-tenths the size of the smallest interstitial channels cannot be treated successfully in this way.
12. The size of the mesh of the jig-sieve has a very important influence, and must be proportioned to the work to be done.
13. The English method of gold jigging without sizing, except possibly so far as is necessary to remove the very finest slimes, has many advantages, and should be more generally adopted.”

Professor Richards, in a very careful and elaborate investigation on the question of jigging relatively small sizes, treats it under four heads:

1. the law of equal-settling particles;
2. the law of interstitial currents;
3. the law of acceleration; and
4. the law of suction.

These four laws are supposed to govern all jigging operations. Practically, Professor Richards’ full conclusions are :

” The two chief reactions of jigging are pulsion and suction.

“ The effect of pulsion depends upon the laws of equal-settling particles, interstitial currents, and, possibly, also of acceleration. The chief function of pulsion is to save the larger grains of the heavier mineral, or the grains which settle faster and farther than the waste.

” The effect of suction depends upon the interstitial factor of the minerals to be separated. If this factor is greater than 3.70, suction will be efficient and rapid. If the factor is less than 3.70, suction will be much hampered and hindered. The use of a long stroke will help to overcome this difficulty. The chief function of suction is to save the particles that are too small to be saved by the laws of equal-settling particles, and of interstitial currents, acting through the pulsion of the jig.

“ For jigging mixed sizes, pulsion with full suction should be used.

“ For jigging closely-sized products, pulsion with a minimum of suction should be used.”

He concludes by saying, in effect:

In jigging minerals having an interstitial factor greater than 3.7, sizing is simply a matter of convenience, although the fine sizes should be removed in some suitable manner. But if the factor is less than 3.7, then the jigging of mixed sizes cannot give a perfect separation, and if this is desired, then close sizing must be adopted, and the closer the sizing the more perfect the jigging. As an expedient, however, there are often cases where a satisfactory separation may be attained without sizing.

The differences in the conclusions of the two investigators above quoted have been chiefly influential in suggesting this present investigation, which was begun in the fall of 1906, and the results of the work done in the Mining Laboratory of the Columbia School of Mines have been embodied in a paper submitted to the Faculty of Pure Science in Columbia University. Since most of the work done then was preliminary to that recently undertaken, I include herewith a resume of my former results and conclusions.

Resume of the Results of Previous Preliminary Work

In the following investigation an effort was made to determine, among other things:

1. the conditions and laws of hydraulic classification;
2. the conditions and limitations of jigging in the pulsion-jig;
3. the effect of varying the length and number of strokes per minute in the Vezin laboratory-jigs;
4. experiments with a large five-compartment Harz jig to determine the limits and perfection of, separation effected in an ore containing galena and sphalerite with a quartzose gangue.

Considered briefly, the results of these tests, in the above-named order, are:

Hydraulic Classification

A number of tests were made with quartz paired with galena, antimony, arsenopyrite, magnetite, sphalerite, etc., in different proportions, and with a velocity varied between wide limits, in order to determine whether a fixed ratio existed as to the diameters of the grains of the two minerals. All tests under this head were made in a Munroe hydraulic laboratory-classifier. Without going into details of the methods, etc., the results indicated that whether or not a more perfect separation was effected in the classifier-tube itself, the manner of drawing off the classified products always resulted in giving a large proportion of mixed products, and after a number of calculations upon different drawings, similar to the manner detailed under the pulsion-jig tests, and described by Professor Richards, proved to my satisfaction that no such ratio existed with classified products under the conditions the above type of classifier was operated and the products removed.

Pulsion Jig Tests

The largest size of Munroe hydraulic classifier was first fitted up in such way that a column of ore 5 to 6 in. long was supported upon a bedding of large grains, and then treated with a pulsating current of water. The tube in which the jigging took place had a diameter of about 1.75 in., and the pulsion was effected by compressing a rubber tube connecting the bottom of the ore-column with a pressure-head of water. The compression of the tube was effected both by mechanical means and by hand, and apparently it made little difference which method was used. The bedding-grains served only to support the ore-column and confine it within the tube; and in drawing off the products this was always first to be removed. The results of jigging under these conditions and the removal of the jigged product—namely, by allowing the jigged material to subside gradually into a rubber tube connected with the receptacle which supported, the bedding, if it may be called such, and which was really the hutch of the jig, were that after drying, screening, weighing, and analyzing the different screen-products from a number of drawings, and finally calculating the ratios between the diameter of the grain of quartz and that of the other mineral paired with it, no such ratio as that given by Richards could be obtained under such conditions, but the tests were in all respects duplicates of the first series run with the classifier operated under the conditions, of hydraulic classification.

It was found, however, that if the jigged products were not removed from the jigging-tube as above described, but, instead, a screen attached to the lower end of the jigging-tube, and the mixture of minerals jigged on this screen, and then instead of drawing off the products through the rubber tube at the bottom the entire apparatus was dismantled, and the jigged products removed from the tube by inserting a piston and forcing the ore-column from the bottom of the tube, cutting sections at equal intervals, that approximate concordant results were obtained. These sections, which were cut off at equal intervals, and usually eight or nine in number, were dried, sized on a nest of sieves, weighed, and analyzed. Ratios of diameters were then calculated for some four or five drawings, in which the mixed grains occurred, according to the method described by Richards, which was as follows: The average diameter of the quartz-grains was obtained by multiplying all the quartz-weights in a particular drawing by their diameters, and dividing the sum of the products by the sum of their weights; and similarly for the other mineral paired with it. The average diameter of the quartz-grain thus determined is divided by the average diameter of the grains of the other mineral, and the quotient is the desired ratio. Table I. gives the ratios that were obtained with the pulsion-jig, the material in nearly all cases being sized between 0.15 and 2 mm. For purposes of comparison I have included the ratios obtained by Professor Richards with a pointed tube, the results of which he considers to hold true for the pulsion-jig as well.

In the tests of Table I., 50 per cent, by volume of each mineral was used. It seems evident, therefore, that under the conditions that exist under the influence of pulsion alone, the free-settling ratios obtained with Rittinger’s formula are increased, but by no great amount.

Vezin Laboratory Jig Tests

Without going into the details of construction of this very useful little laboratory-apparatus, suffice it to say that the piston is driven by a variable-speed shaft, with a disk and friction-wheel, and the number of strokes may be varied from 100 to 300 per min., and, with a double eccentric, the length of stroke from 0 to 1.25 in. (31.7 mm.). The box carrying the sieve is attached to the body of the jig by means of clamps, so that, together with the ore and bedding resting on the sieve, it may easily be removed and the contents examined, or another box with its attached sieve substituted. In all tests with the Vezin jig a sieve of 8-mesh (2.2 mm. square hole) was used. The bedding was in most cases sized between the limits of 2.5 and 3.3 mm., and maintained at a thickness of 0.75 in. (19 mm.). The jig was driven from a counter-shaft by an electric motor, so that a uniform speed was secured. The feed in all cases was sized between the limits of 0.10 and 1.9 mm., and the various mixtures were made up by volume to contain 3 of quartz and 1 of the heavier mineral. From 1.6 to 2.0 kg. represented the amount generally employed in each test. After this quantity had been run over the jig it was stopped, the sieve-box removed, the contents placed in a large pan and dried, the hutch-work drawn off, the water decanted and treated in the same way, and finally the tailings were freed as far as possible from water and dried. The three products were then sized separately on a nest of sieves, each size weighed and analyzed, the material being subsequently used again for another test.

It is evident that in so simple a machine as the Vezin jig there are a number of factors that may be made either constant or variable. Thus the length of stroke and number per min. are easily varied, or may be kept constant; the size of the grains constituting the bed, and its thickness, may be varied within limits, although this is likely to vary with other factors, especially the piston-speed; then the quantities of water used on the piston side, with the feed and the amount discharged from the hutch, as well as the rate of feed, may also be varied. In these tests the length and number of strokes were the principal variables, and also the amount of suction, of which there are a number of degrees, limited as follows :

Full suction

In which the hutch-spigot is fully open, and the water thus discharged is supplied entirely by increasing the amount added with the feed, and, if possible, cutting down the amount supplied to the piston side.

Part suction

In which the hutch-spigot is not fully open, and does not discharge a quantity equal to the extra amount added with the feed.

Balanced suction

In balanced suction the hutch-spigot is closed and the feed-water and piston-water are equal; or the hutch-spigot is partly or fully open, and the amount thus discharged is supplied entirely from the piston side.

The results obtained indicate the following conclusions:

Jig Suction

With full suction, (A), the bed was not mobile, and after a few minutes’ feeding the jig was very badly choked and little or nothing passed into the hutch. After trying a few tests with the same bad results, full suction was considered impracticable. In the case of part suction, (B), the mobility of the bed was decreased in proportion to the amount of suction, and with it a decrease in the amount of coarse mineral passing into the hutch, but with a corresponding increase in the amount of fine material without a noticeable enrichment. The best results were obtained with balanced suction, having the spigot completely closed, although the results with the spigot partly or fully open did not differ materially from those of full suction (A).

Feed Water and Rate of Feed to Jig

These factors were kept as nearly constant as possible, and the effect of varying them was not considered.

Jig Filter-Bed

The thickness and the size of the filter-bed, also, were made a constant. It was found, however, that the shape of the grains of the bedding does influence the ease and rapidity with which the mineral passes into the hutch. Thus with antimony and arsenopyrite, both of which break into long, pencil-shaped grains, the sieve became quickly blinded, which interfered with the free passage of grains below, and required a long, heavy stroke to dislodge them.

Length and Number of Jig Strokes

The results of the tests seemed to show that the character of the separation is not directly dependent upon absolute piston-speed, but that the quick, short stroke was more efficient, and resulted in a cleaner hutch-product, and relatively more of it, than a longer stroke of less frequency, but of the same piston-speed.

Mineral Concentration

If the diameters of the grains of the heavy mineral jigged, and of the bedding-grains (and therefore the diameter of the sieve-hole), do not differ by any large amount, a clean separation can easily be made. With an increase in these ratios, perfect separation is impossible. Stated in other words, with bedding of a definite size, and hence a fixed sieve-aperture, the finer the grain the more difficult is its separation on the jig.

Specific Gravity

Within rather wide limits, the difference in the specific gravity of the heavier mineral paired with quartz did not influence greatly the ease with which it could be separated, or a good concentration attained.

Experiments with a Five-Sieve Harz Jig

Two runs were made as nearly as possible under practical conditions to determine to what extent the conclusions derived from the Vezin-jig tests were applicable to an ordinary jig. The ore used for the work contained 6 per cent, of mineral—about half sphalerite and the balance galena, with a quartzose gangue. The jig was bedded with material sized between 5.2 and 6.6 mm. The first compartment was bedded with a clean galena, the second with sphalerite, and the third, fourth, and fifth with mixtures of sphalerite and quartz. The thickness of the bedding averaged from 20 to 30 mm. at the beginning of the run. All beds naturally tended to increase in thickness, since no products were skimmed off during the run.

The jig differed in no respect from the common type of Harz jig. Each sieve-compartment was 16 by 20 in. (406 by 512 mm.) in section, with pistons of equal area. The lengths of strokes could be adjusted between limits of 0 to 50 mm., and within a considerable range in the number per min.—in the experiments, from 175 to 180. The actual piston-speeds used ranged about as follows: first compartment, 75 mm.; second, 66 to 70 mm.; third, 57 to 67 mm.; fourth, 45 to 58 mm.; and fifth, 45 to 50 mm. per sec. Only the hutch-products and tailings were examined.

The ore, sized between 0 and 4.8 mm., round hole, was delivered to the jig through a centrifugal pump. All products traveled in closed circuits, and were finally returned to the centrifugal elevator or pump to be passed again over the jig. The spigots constantly discharged their products, and from these discharges time-samples were cut out. The run occupied exactly an hour, so that after weighing each of the products—in this case six—with the tailings, data were at hand for calculating the capacities; and after screening, weighing, and analyzing, a complete record of the run was made. The results of these tests showed that the differences in length of stroke, or number of strokes per min., were not sufficient to produce a marked difference in the character of the concentrate; that most of the galena was saved in the first hutch and most of the sphalerite in the second; that the third, fourth, and fifth hutches carried very little galena, but more sphalerite. It was found that the first hutch-product contained 57 per cent, of galena, and of this nearly 70 per cent, was larger than 1 mm. in diameter; and that sizes finer than this contained more quartz and less galena. The results seemed to indicate the necessity of first removing stuff less than 0.4 mm. in diameter in order to increase the richness of the product. The first hutch-product contained no coarse sphalerite, and only when the material was as small as 0.2 mm. was any considerable amount present. This seems to indicate that an almost perfect separation of these two minerals (galena and sphalerite) from each other and quartz, under the conditions with which the jig was operated, was possible if the feed had been sized between the limits of 0.4 and 4.8 mm. The second hutch, which carried most of the sphalerite, shows that the coarse sizes pass through the sieve of the jig less readily than galena. Not until the material was reduced to 1.5 mm. was any marked percentage noticeable in the product. From 0.4 to 1.5 mm. most of the saving was made. Evidently, the cause for so little very fine stuff in the second hutch-product was owing to the fact that most of it was caught in the first. Under the conditions obtaining in the second compartment, a very satisfactory separation could be made on all sizes below 1.5 mm. The results from the third compartment were like those of the second. The fourth and fifth hutches indicated a further saving of sphalerite, but between somewhat different size-limits than in the second and third; the limits in the last two compartments varied between 0.7 and 2.5 mm., with very little fine stuff. This result indicates that in the first compartments more fine material is present, making a denser and more impervious bed, and that the large grains cannot so easily pass through it; and that in the last compartment the bed is more open and porous, and hence larger grains can more readily pass into the hutch. An examination of the tailings indicated that the loss in the fine material was very small, but by far the largest loss was in the four coarsest sizes, which were mixed grains or middlings, and to reduce this loss further crushing must be done. The results indicate that in order to separate sphalerite and quartz, a jig of at least three compartments should be used; since smaller differences in the specific gravity of these minerals require a longer time to effect the separation. In the case of a heavy mineral, such as galena, one or two compartments will effect a perfect separation.

Experiments with the Jarvis Laboratory-Jig

In order to investigate particularly the effect of pulsion and suction upon jigging, and upon accelerated and retarded strokes, I designed a special jig, with which I conducted a series of experiments and obtained the following results:

How to Make/Build a Jig

Figs. 1, 2, and 3 are detailed drawings of the Jarvis laboratory-jig, with the exception of the variable-speed shaft, which is of the ordinary disk-and-friction-wheel pattern. Figs. 4 and 5 show the designs of the cams used. The screen-area in this jig is 8 by 12 in. (203.2 by 304.8 mm.), with a piston of equal area. With an adjustable dam, the height of discharge may be varied from 3 to 4.5 in. (76.2 to 114.3 mm.). In order to study the behavior of the ore-column and bedding during the process of jigging, one side of the jig-box was made of plate glass. Three types of strokes were employed: (1) The eccentric, adjustable within the limits of 0 and 2 in. (0 and 50.8 mm.). (2) Circular-arc cams, where the period of pulsion occupies three-fourths of the revolution of the cam, or eccentric shaft, and suction one- fourth; or by reversing the direction of rotation of the cam-shaft, or slipping the hub and cam off the shaft and turning it end for end, the times or periods are reversed respectively for pulsion and suction. Cams were made having throws up to 2 in. (50.8 mm.), but only the three shortest throws were used— namely, 1 in. (25.4 mm.), 0.5 in. (12.7 mm.), and 0.25 in. (6.35 mm.). (3) Involute cams, in which the periods were divided into thirds, i.e., one-third of the revolution of the cam-shaft devoted to pulsion, and two-thirds to suction; or as noted above, by reversing the direction of rotation of the cam these periods were reversed. All cams were made of wood, and quickly and easily attached to a cast-iron hub, and by means of a set-screw fastened to the shaft, as shown in full in Fig. 4. Circular-arc and involute cams indicate the character of the curves. The circular-arc cams do not give a uniform motion; or in other words, the cam in describing equal arcs in either the pulsion- or suction-period does not cause the piston to travel equal distances. In the involute cams, however, in either pulsion- or suction-periods, equal arcs give equal distances for piston-travel. The manner of communicating motion from the cams or eccentric is clearly indicated in Figs. 1 and 2. These engage with a brass roller attached to a wrought-iron yoke moving between vertical guides. In order to steady and support the yoke still more, a steel rod is attached to the upper end, passing through a hole in a cross-beam, and is attached to the lower end of the yoke of the piston-rod. The roller, yoke, and piston are actuated

positively by the cam on the up-stroke, and to secure a strong and quick down-stroke, a spring of 60 lb. pressure per linear inch of compression was employed. This elastic pressure insured a uniform contact of the roller and cam. It is evident that a large number of styles of cam-curves may be used

with this device, and the period of movement of the piston may be varied almost infinitely. It is to be observed, also, that in this system the piston, in all positions, is perfectly horizontal. The piston is made of a single piece of sole-leather, securely riveted between two heavy plates of galvanized iron. With these materials the piston can be run with very little

clearance, and there is no danger of warping, swelling, or getting out of repair very easily. The hutch-box sloped from three sides, at an angle exceeding 50°, to a single spigot in one side of the jig. It was found that at this angle little or no hutch- work collected on the sides, and its entire removal was easily effected. The jig was driven by a 1-h.p. electric motor through the variable-speed counter-shaft. The sieve was supported in a galvanized-iron skeleton, which was removable from the jig- box itself, and different sized screens could readily be inter-

changed. In the tests hereinafter described, only one size sieve—an 8-mesh one—was used.

Screens

Table II. gives the number and mesh of the screen, and the size of the aperture in inches and millimeters. In all cases the holes were square. The size of the hole in the first five sizes was determined by measuring the wire with a wire-gauge, and counting the number of meshes in a given length. For the remaining screens the diameter of hole was determined by measuring the diameter of the wire and the aperture with a microscopic micrometer, each value given being the mean of several determinations.

Sieve-Sizes

The data pertaining to the sieve-sizes are given in Table II.

Jig Bedding

The bedding used in all the following tests was sized between the limits of 3.510 and 5.326 mm., or through the 4-mesh sieve and on the 6-mesh sieve, and was maintained at the same thickness, 1.5 in. (38.1 mm.), upon the jig-sieve throughout the experiments.

Minerals

The three minerals used were fairly pure. The quartz was kindly furnished by Professor Munroe, and the sphalerite and galena by the Foote Mineral Co., of Philadelphia, Pa.

Specific Gravities

The specific gravity of each mineral was: galena, 6.66; sphalerite, 3.74; and quartz, 2.62.

The low specific gravity of the metallic minerals used indicates that they are not pure, and an examination revealed the presence of included quartz and minute quantities of other minerals. In crushing these minerals, all the quartz particles that could be picked out by hand were removed. The values given, however, are those obtained for the crushed minerals, ready to be added to the feed.

These three minerals were selected since zinc-blende and galena represent about the minimum and maximum limits respectively of the ores usually treated on jigs. In thus examining the two limits, the behavior of intermediate minerals could be closely predicted.

Jig Feed

The feed in all the tests was crushed by stages until small enough to pass the 10-mesh (2.136-mm.) screen. This size represented the maximum, from which it varied to that of the finest dust. Two classes of feed were employed. The first contained 10 per cent., by weight, of heavy mineral (galena or blende), and the second 20 per cent, of heavy mineral. The balance was, respectively, 90 or 80 per cent, of quartz. Table III. shows the screen-analysis of the three minerals constituting the feed.

The values in Table III. represent the mean of three or four different determinations, made after crushing a large lot and thoroughly sampling it down.

Table IV. shows the calculated percentages of galena and quartz in the two classes of feed, based upon the screen-analysis of the pure minerals given in Table III.

The results obtained for sphalerite and quartz are given in Table V.

In Tables IV. and V. the columns for each of the respective feeds show the percentages of each of the two minerals on the different screen-sizes. Thus, in Table IV., with 10 per cent, of galena,the material resting on the 12-mesh (1.66-mm.) sieve contained 94.1 per cent, of quartz and 5.9 per cent, of galena, etc.

With both sphalerite and galena, the screen-analyses, and from these the calculated percentages of the mineral-content of each screen-size, show that more fine material is produced in crushing these softer minerals than in crushing quartz. The finest size of the 10 and the 20 per cent, galena or sphalerite shows a much higher percentage of these minerals than the average of the feed, as shown in Tables IV. and V.

Method of Conducting a Jigging Test

In beginning a series of tests on a given feed, the exact proportion of each mineral was weighed out, so that the total quantity was 35 lb. (15.87 kg.). Meanwhile, the sieve had received its bedding, 1.5 in. (38.1 mm.), and the hutch-box and jig were filled with water; the tailings-trough placed in position, connecting with a large tub in which all the overflow and tailings were caught; the feed thoroughly wetted down (if fresh material); power was turned on and the jig started. In case it was the first run of a series, the jig-box containing bedding only, the feed was rapid until this was filled with the mixture, after which the feeding proceeded at the regular rate.

The feeding was accomplished by filling with the ore-mixture a large flat-bottomed scoop, of a width slightly less than that of the jig-compartment, 8 in. (203.2 mm.), and with a small and constant stream of water washing the material from the scoop on to the jig. While the speed of jigging and the rate of feeding varied, the object always aimed at was to feed the jig just as fast as it appeared able to treat the material. The discharge was watched constantly to see if any particles of heavy mineral were being carried into the tailings. If so, the rate of feeding was reduced. Close watch was also kept on the jig-bed, and if the jig showed symptoms of clogging up, due to rapid feeding, the rate of feed was immediately decreased.

At the end of the run, usually from 8 to 15 min., the jig was stopped, the water-supply cut off, and the hutch-products drawn off into suitable vessels. After allowing the material to settle, the water was carefully decanted and the products thoroughly mixed, and a sample of about 125 g. cut out, which was dried, and later exactly 100 g. of this sample was weighed out on a pulp-balance and sized on a nest of sieves, ranging from 12-mesh (1.66 mm.) through 100-mesh (0.16 mm.), and each size carefully weighed; finally, the percentage of galena or sphalerite in each sieve-size was determined. The analyses of the products were made in several ways. In the first two or three coarse sizes good results were obtained by weighing out 1 or 2 g. and picking out the quartz or other mineral by hand and then weighing again; also, by comparing with standard mixtures of quartz and galena or sphalerite. In the small sizes vanning-tests were made.

After the completion of a run, the tailings, which were given ample time in which to allow the fine material to settle and the water to be decanted off, were again mixed with the product from the hutch and formed the feed for another test. The material was thus used repeatedly until all the tests had been completed for a particular series or class. The material remaining in the jig-box was not cleaned out from test to test, unless another feed was to be employed. The investigations had to do only with what passed into the hutch, and determinations upon the character and nature of what remained on the sieve, except as it could be examined through the glass side of the jig, were not made.

Record of Results

In the following records are five horizontal rows of figures: in the topmost row, the sieve-mesh; in the next lower row, the corresponding size in millimeters of the aperture upon which the material was caught; and three lower rows marked “A,” “ B,” and “ C” respectively. The first of these, A, gives the weights in grams of the different sieve-sizes; and since these are all on a basis of 100 g. the weights, therefore, represent percentages as well. Row B gives the percentage of heavy mineral, galena or sphalerite, in each of the sieve-products, and the balance in every case is quartz. Row C gives the weight of heavy mineral contained in each of the sieve-sizes, and is obtained by multiplying the weights in row A by the respective percentages in the B row. The sum of the products in the C row gives the number of grams of mineral in 100 g. of the concentrate, or in other words, the percentage.

Under the stroke of each experiment are given: (1), the number of revolutions of the cam or eccentric shaft per minute ; (2) the length in inches and millimeters; (3) the kind of stroke; (4) pulsion, in which the fractious ¼, 1/3, ½, 2/3, and ¾ refer to the fractional part of the entire revolution of the cam or shaft in which this movement took place. The smaller this fraction the quicker the movement. The rates or velocities are set opposite. The same is true for the period of suction.

The observed pulsion- and suction-velocities noted in the following tests and elsewhere in this paper are to be understood as the mean piston-velocities, or the velocities of the water-column in the free part of the jig-column, and not the actual current- velocities acting upon a mass of grains constituting the jig-bed.

In studying these experiments, reference should be made to Figs. 6 to 13, inclusive, in which row C is shown graphically, representing the mean diameter of grains.

TEST 1.—Galena 10, Quartz 90 per cent.

Percentage of galena in concentrates : 35.7.
Ratio of concentration based on original feed: 3 57.
Remarks.—All the material on the jig-hed pulsated—the material above having a longer amplitude than the grains deeper down. It was observed that the bedding- grains at the top moved nearly 0.75 in. vertically, while those at the bottom of the bed next to the screen moved about 0.25 in.

Percentage of galena in concentrates : 42.4.
Ratio of concentration based on original feed : 4.24.
Remarks.—Some movement of the bedding-grains, especially near the top, but only a few grains of galena were visible in the interstitial spaces of the bedding.
The ore-column pulsated violently, and between the bedding and the ore was a zone in very active motion, while above the column of ore was quite compact.

Percentage of galena in concentrates : 29.9.
Ratio of concentration based on original feed : 2.99.
Remarks.—The bedding and with it the ore-column pulsated. The grains of bedding were kept in constant circulation. Very few grains of galena could be seen in the interstitial spaces of the bedding, but were free. It was evident, therefore, that if a galena- or quartz-grain got as far as the hedding it had little opportunity of remaining there.

Percentage of galena in concentrates : 24.6.
Ratio of concentration based on original feed : 2.46.
Remarks.—The ore-bed pulsated violently, but not so much so as with the strong suction of the circular-arc cam in Test 3. The grains of the bed did not behave exactly alike, and the middle of the bed contained some grains of galena.

Percentage of galena in concentrates : 34.1.
Ratio of concentration based on original feed : 3.41.
Remarks.—The entire bed pulsated much more uniformly than in Tests 3 and 4. The bedding-grains were free to move, and tended to move in convection-currents. No particles of galena collected on top of the bedding, and few could be seen in the interstitial spaces.

Percentage of galena in concentrates : 50.2.
Ratio of concentration baaed on original feed : 5.02.
Remarks.—The lower third of bedding quite fixed, while the upper two-thirds pulsated, but the grains did not change positions—moving en masse. The ore-column pulsated regularly, and between the bedding and the ore was a zone of great mobility. The action and movement going on in the ore-column resembled very much that taking place in a hydraulic classifier.

Percentage of galena in concentrates : 40.1.
Ratio of concentration based on original feed : 4.01.
Remarks.—The bedding and the ore-column pulsated uniformly—the top having a longer amplitude and extending over a longer time than the grains nearer the bottom. Ore-column very mobile and in active circulation. The upper third of bedding contained many particles of galena.

Percentage of galena in concentrates : 34.3.
Ratio of concentration based on original feed : 3.43.
Remarks.—Only the top third of bedding showed any signs of movement, but tbe interstitial spaces were filled with particles of galena. The particles in the ore-column tended to circulate in two opposite and distinct paths.

Percentage of galena in concentrates : 48.3.
Ratio of concentration based on original feed : 4.83.
Remarks—Both the bedding and the ore-column pulsated—the top having a longer amplitude of vibration and requiring a longer time than the grains nearer the bottom. Many grains of galena in the upper third of bedding and decreasing below. The ore-column very mobile, and line between bedding and ore horizontal and uniformly even.

Percentage of galena in concentrates : 43.0.
Ratio of concentration based on original feed : 4.30.
Remarks.—The bedding-grains were practically stationary—neither pulsation nor movement among themselves, and were filled with particles of galena. The ore-column pulsated, but was not mobile except for a zone 0.5 in. thick on top of the bedding. Evidently too much suction.

Percentage of galena in concentrates : 49.0.
Ratio of concentration based on original feed : 4.90.
Remarks.—The bedding, as a whole, did not pulsate, but the grains in the upper part of the bedding showed some movement, and this portion was filled with particles of galena. The ore-column was very mobile and pulsated regularly and uniformly. The large grains of quartz rested directly upon the bedding, with the finer quartz-particles above.

Percentage of galena in concentrates: 37.0.
Ratio of concentration based on original feed : 3.7.
Remarks—The bedding did not move at all. The ore-bed seemed to be quite mobile immediately above the bedding, but compact close to the top.

Percentage of galena in concentrates : 36.5.
Ratio of concentration based on original feed : 3.65.
Remarks.—The bedding and the ore-column pulsated, but the grains of bedding were not sufficiently mobile to rearrange themselves, although the upper third was much more mobile and pulsated much more than the bottom, and many particles of galena were contained in the interstitial spaces of the upper third of bedding. The entire ore-column was very free and mobile and pulsated uniformly.

Percentage of galena in concentrates : 35.9.
Ratio of concentration based on original feed: 3.59.
Remarks.—No movement in the bedding, although the top, bedding-grains showed some tendency to move, and many particles of galena could be seen in the upper third of the bedding. The ore-column pulsated regularly, and was quite compact.

Percentage of galena in concentrates : 33.4.
Ratio of concentration based on original feed : 3.34.
Remarks.—The bedding was quite fixed in position, and the upper part well filled with grains of galena. It was noticed that when the feed was too fast, an inclined line, beginning at the top of the bedding at the back of the jig-box, and sloping up nearly to the top of the ore-column at the front or discharge was formed. Otherwise the ore-column was mobile, with the coarse particles of quartz resting above the bedding, and the finer particles arranged above.

Percentage of galena in concentrates : 40.6.
Ratio of concentration based on original feed : 4.06.
Remarks.—Both the bedding and the ore-column pulsated regularly—the grains near the top of the bedding having a longer amplitude of vibration than those near the bottom, and the same being true of the grains in the ore-column. The ore-column was very mobile. The jig worked fast.

Percentage of galena in concentrates : 29.0
Ratio of concentration based on original feed : 2.90.
Remarks.—The entire bed pulsated, and the bedding contained many particles of galena and some quartz. As noted before, the top had a longer amplitude of vibration than the bottom, and required a longer time. The ore-column pulsated regularly, and the fine material (quartz) was carried down to the bedding so that it was distributed quite regularly throughout the ore. The ore-column was compact.

Percentage of galena in concentrates : 24.8.
Ratio of concentration based on original feed : 2.48.
Remarks.—The grains of the bedding were not very mobile, and only the top layer of grains showed any indication of pulsating. The base of the ore-column was distinguished by a zone of active agitation. Above this zone, which was only 0.5 in. thick, the ore-column was compact and not mobile. The bedding-grains contained only a few galena-grains in the upper third portion, but the interstitial spaces were filled with quartz. In the middle and lower third portions of the bedding, many more grains of galena were visible, being more abundant in the middle third.

Percentage of galena in concentrates : 59.6.
Ratio of concentration based on original feed : 2.98.
Remarks.—Movement of jig-bed same as Test 1.

Percentage of galena in concentrates : 57.3.
Ratio of concentration based on original feed : 2.86.
Remarks.—Movement of jig-bed similar to Test 2.

Percentage of galena in concentrates : 56.0.
Ratio of concentration based on original feed : 2.80.
Remarks.—Movement of bed similar to Test 3.

Percentage of galena in concentrates : 52.5.
Ratio of concentration based on original feed : 2.62.
Remarks.—The entire bed pulsated, but not so violently as Test 23. The ore-column pulsated much more than the bedding, and the top of the bedding than the bottom. Between the bedding and the ore-column was a zone 0.5 in. thick of great activity. Few grains in the interstitial spaces of the bedding. Jigged rapidly.

Percentage of galena in concentrates : 67.4.
Ratio of concentration based on original feed : 3.37.
Remarks.—The entire bed pulsated, the upper part having a longer amplitude of vibration and requiring a longer time to complete it than the grains nearer the bottom. Difficult to save the finest grains of galena. The bedding-grains were free to change positions during the pulsion-cycle.

Percentage of galena in concentrates : 60.5
Ratio of concentration based on original feed : .3.02.
Remarks.—The entire bed pulsated, and the zone between the bedding and the ore-column was an active one—the grains in the ore-column were kept in constant circulation. The interstitial spaces in the upper third of the bedding filled with particles of galena.

Percentage of galena in concentrates : 62.9.
Ratio of concentration based on original feed : 3.15.
Remarks.—The entire bed pulsated very uniformly, the top having a longer time to complete it than the grains nearer the bottom. Many grains of galena were observed in the upper part of the bedding, but only a few in the lower half.

Percentage of galena in concentrates: 59.0.
Ratio of concentration based on original feed : 2.95.
Remarks.—The bedding-grains did not pulsate, although those near the top exhibited a slight tendency. The ore-column pulsated, but excepting a zone about 0.5 in. thick immediately above the bedding was otherwise compact. The ore- grains circulated in two distinct and opposite paths.

Percentage of galena in concentrates : 67.5.
Ratio of concentration based on original feed: 3.37.
Remarks.—The entire bed pulsated, the upper part, as noted before, having a longer amplitude of vibration and requiring a longer time to complete it than the grains beneath. The upper half of the bedding contained many particles of galena, while only a few were visible in the lower half.

Percentage of galena in concentrates : 53.1.
Ratio of concentration based on original feed : 2.65.
Remarks.—The bedding exhibited a slight tendency to pulsate en masse. The upper part of the bedding well filled with particles of galena, decreasing rapidly in number below. Immediately above the bedding the ore-column presented a zone of active agitation about 0.5 in. thick, while above the particles seemed quite compact and not very mobile.

Percentage of galena in concentrates: 67.7.
Ratio of concentration based on original feed: 3.39.
Remarks.—The upper one-third of the bedding-grains exhibited some tendency to arrange themselves during pulsion, but the lower two-thirds did not move or pulsate. In the upper third were many particles of galena and less below. The ore-column pulsated regularly, the large grains of quartz arranging themselves next to the bedding, the smaller on top. The ore-column was mobile.

Percentage of galena in concentrates : 52.6.
Ratio of concentration based on original feed : 2.63.
Remarks.—The bedding did not pulsate. The upper third was filled with particles of galena and decreasing numbers below. The ore-column was somewhat mobile in spots, but pulsated quite regularly, and on top of the bedding was a zone which exhibited considerable activity.

Percentage of galena in concentrates: 46.5.
Ratio of concentration based on original feed : 2.32.
Remarks.—The upper half and often more of the bedding pulsated. In this part, also, were many particles of galena. The ore-column was mobile, with the large quartz-grains arranged near the top of the bedding and the smaller sizes above.

Percentage of galena in concentrates : 63.2.
Ratio of concentration based on original feed : 3.16.
Remarks.—The bedding exhibited very little tendency to pulsate, nor was there any movement among the grains themselves. The upper half of the bedding was well filled with grains of galena. The particles of ore above the bedding circulated in two opposite orbits, passing down at the front and back end of jig and joining in the center.

Percentage of galena in concentrates : 59.7.
Ratio of concentration based on original feed : 3.
Remarks.—The bedding pulsated slightly, and the upper half was well filled with galena, with decreasing quantities below. The ore-column pulsated regularly, with the largest grains of quartz resting on top of the bedding, decreasing in size above.

Percentage of galena in concentrates : 55.2.
Ratio of concentration based on original feed : 2.76
Remarks.—The entire bed pulsated regularly. The upper part of bedding contained many particles of galena.

Percentage of galena in concentrates : 60.8.
Ratio of concentration based on original feed : 3.04.
Remarks.—The upper third of bedding was quite mobile, and filled with particles of galena, decreasing below. The ore-column seemed quite compact, but pulsated regularly.

Percentage of galena in concentrates : 55.3.
Ratio of concentration: 2.76.
Remarks.—Both ore and bedding pulsated regularly, but violently. Considerable of the finest size of galena could be seen in the tailings. The jig worked very rapidly.

Percentage of sphalerite in concentrates : 24.7.
Ratio of concentration based on original feed : 2.47.
Remarks.—The bedding pulsated very violently, and after the jig was stopped it was found that the surface of the ore-column was 1.5 in. below the tailings-dam.

Percentage of sphalerite in concentrates: 39.3.
Ratio of concentration based on original feed: 3.93.
Remarks.—The bedding-grains were carried up from bottom to top, circulating in that way as by convection-currents. The ore-column was in active agitation, and the bedding and the ore were not separated by a clearly defined and horizontal line.

Percentage of sphalerite in concentrates : 37.6
Ratio of concentration based on original feed : 3.76.
Remarks.—The movement of the jig-bed was very similar to Test 42. The bedding-grains were not only carried from the bottom of the bedding-column itself, but many rose to the top of the ore-column, and a few of the lightest were carried off with the tailings. The grains of quartz could be seen very plainly rolling down with the larger bedding-grains and being carried into the hutch.

Percentage of sphalerite in concentrates : 50.7.
Ratio of concentration based on original feed : 5.07.
Remarks.—Both the bedding and the ore-column pulsated regularly. Each formed distinct and well-defined layers. The jig worked very rapidly.

Percentage of sphalerite in concentrates : 50.5.
Ratio of concentration based on original feed : 5.05.
Remarks.—The upper half to three-fourths of the bedding pulsated regularly, the bottom grains were almost stationary. The lower part of the ore-column consisted of the largest particles of quartz, with smaller and smaller grains to the top.

Percentage of sphalerite in concentrates : 37.0.
Ratio of concentration based on original feed: 3.70.
Remarks.—The entire bed moved en masse, the top of the column having a longer amplitude of vibration and requiring a longer time for its completion than the grains nearer the bottom. The jig worked rapidly.

Percentage of sphalerite in concentrates : 33.2.
Ratio of concentration based on original feed : 3.32.
Remarks.—The bedding pulsated, but not regularly, and tended to thicken in the middle and thin down at the ends. The grains at the bottom of the ore-column were in very active agitation, but it was found that these grains were really describing two distinct orbits.

Percentage of sphalerite in concentrates: 39.2.
Ratio of concentration based on original feed : 3.92.
Remarks.—The upper two-thirds of the bedding and the entire ore-column pulsated regularly. As noted before, the grains nearest the top had a longer amplitude of vibration and required a longer time to complete it. The lower one-third of the bedding was quite stationary.

Percentage of sphalerite in concentrates : 31.0.
Ratio of concentration based on original feed : 3.1.
Remarks.—The bedding pulsated slightly, and the grains shifted positions as in convection-currents. A zone between the bedding and the ore-column moved much as noted in Test 40. The ore-column above this zone pulsated regularly, although the ore-column was not very mobile.

Percentage of sphalerite in concentrates : 32.8.
Ratio of concentration based on original feed : 3.28.
Remarks.—The bedding and with it the ore-column pulsated en masse. Taking the entire column of bedding and ore as a whole, the top had a much longer amplitude of vibration, and required a longer time in which to complete it.

Percentage of sphalerite in concentrates : 46.1.
Ratio of concentration baaed on original feed : 4.61.
Remarks.-Movement of jig-bed very similar to that of Test 48, but to a less extent.

Percentage of sphalerite in concentrates : 47.4.
Ratio of concentration based on original feed : 4.74.
Remarks.—The bedding pulsated in the upper third and half, and was quite mobile as well. The lower part, however, was stationary. The line between bedding and ore was horizontal and straight. The ore-column pulsated regularly— the top for a greater distance, and for a longer time, as before.

Percentage of sphalerite in concentrates : 41.2.
Ratio of concentration based on original feed : 4.12.
Remarks.—The entire bedding was practically stationary, did not pulsate, nor was it mobile. The interstitial spaces in the upper part of bedding filled with grains of sphalerite. The ore-column pulsated en masse and was fairly mobile.

Percentage of sphalerite in concentrates : 52.7.
Ratio of concentration based on original feed : 5.27.
Remarks.—The upper third of the bedding was mobile, but the lower two-thirds was quite fixed. The ore-column pulsated regularly, together with the upper third of bedding. The line between the ore-column and the bedding was clearly marked.

Percentage of sphalerite in concentrates : 33.9.
Ratio of concentration based on original feed : 1.7.
Remarks.—The entire jig-bed pulsated very violently. The feed was very fast, a large amount of hutch-work was made, and the tailings contained considerable fine mineral. The fine quartz could be seen sifting through the bedding.

Percentage of sphalerite in concentrates : 45.2.
Ratio of concentration based on original feed : 2.26.
Remarks.—Behavior of jig-bed similar to Test 41.

Percentage of sphalerite in concentrates : 43.9.
Ratio of concentration based on original feed : 2.20.
Remarks.—Behavior of jig-bed similar to Test 43.

Percentage of sphalerite in concentrates : 41.6.
Ratio of concentration based on original feed : 2.08.
Remarks—Movement of jig-bed similar to Test 44.

Percentage of sphalerite in concentrates : 53.7.
Ratio of concentration based on original feed : 2.7.
Remarks.—Movement of jig-bed similar to Test 45.

Percentage of sphalerite in concentrates : 57.3.
Ratio of concentration based on original feed : 2.8.
Remarks.—Movement of jig-bed similar to Test 46.

Percentage of sphalerite in concentrates: 46.4.
Ratio of concentration based on original feed : 2.32.
Remarks.—The bedding and the ore-column pulsated, and the bottom grains of bedding much more than in Test 66.

Percentage of sphalerite in concentrates: 49.4.
Ratio of concentration band on original feed : 2.47.
Remarks.-Movement of jig-bed similar to Test 48.

Percentage of sphalerite in concentrates: 53.1.
Ratio of concentration based on original feed: 2.65.
Remarks.—The entire jig-bed moved en masse, and was very mobile. As in all cases of this kind, the upper part of the ore-column had a longer amplitude of vibration and required a longer time in which to complete it than the grains (whether bedding or ore) nearer the bottom.

Percentage of sphalerite in concentrates: 45.9.
Ratio of concentration based on original feed : 2.29.
Remarks.—Movement of jig-bed similar to Test 50.

Percentage of sphalerite in concentrates : 46.2.
Ratio of concentration based on original feed : 2.31.
Remarks.—Movement of jig-bed similar to Test 51.

Percentage of sphalerite in concentrates : 62.9.
Ratio of concentration based on original feed : 3.14.
Remarks.—The entire mass except the lower part of the bedding pulsated en masse, and the ore-column seemed quite mobile.

Percentage of sphalerite in concentrates : 65.3.
Ratio of concentration based on original feed : 3.26.
Remarks.—The upper two-thirds of bedding and the entire ore-column pulsated. Ore-column mobile.

Percentage of sphalerite in concentrates: 69.3.
Ratio of concentration based on original feed: 3.46.
Remarks.—The upper third of bedding together with the ore-bed pulsated en masse. The lower two-thirds of bedding was quite fixed in position. The top of ore-column, as before, had a longer amplitude.

Percentage of sphalerite in concentrates : 65.4.
Ratio of concentration based on original feed : 3.27.
Remarks.—The upper third of bedding together with the ore-column pulsated en masse. The lower two-thirds of bedding scarcely moved. The interstitial spaces of the bedding, as with all experiments with the short stroke, filled with mineral.