have become thin fluid and flow bright and uniform. The operation is therefore finished in about half of an hour. The crucibles are now allowed to cool, and the buttons weighing about 22 grammes (339.51 grains) are freed from slag and then cupelled. (In assays of large lots of ore five assays are made.)-Other charges for sweepings: 25 grammes (385.80 grains, about A. T.) of sweepings, the same quantity of minium, 35 grammes (540.13 grains) of flux (prepared by mixing 600 parts of potassium carbonate, 200 parts borax, 100 parts glass-galls, 100 parts soda, 30 parts saltpetre, 30 parts powdered charcoal); or, 25 grammes (385.80 grains) of sweepings, 20 grammes (308.64 grains) of common salt, the same quantity of sodium carbonate and of potassium carbonate, 25 grammes (385.80 grains) of litharge, 10 grammes (154.32 grains) of argol, and the same quantity of powdered glass; or; 25 grammes (385.80 grains) of sweepings, 20 grammes (308.64 grains) of litharge, 25 grammes (385.80 grains) of flux (1 part potassium carbonate and 1 sodium carbonate), and a covering of common salt. The charge is kept in the furnace until quiet fusion, requiring about three-quarters of an hour (Braubach).—Slags: 10 grammes (154.32 grains, about A. T.) of slag, 150 to 160 grammes (2314.85 to 2469.17 grains) of litharge, 2.5 grammes (38.58 grains) of quartz, and 0.25 gramme (3.85 grains) of charcoal powder. The charge is fused in a crucible for 20 minutes after the development of gas in the furnace has ceased, and the buttons of two charges are cupelled together (Pribram).—Freiberg: 7.5 grammes (115.74 grains, about A. T.) of slag are mixed in a crucible with 11 to 15 grammes (169.76 to 231.48 grains) of potassium carbonate and flour, 19 to 30 grammes (293.21 to 462.97 grains) of granulated lead are strewed on top, and the charge is fused for 3 hours in the furnace.-Refuse from stamping mills: 10 grammes (154.32 grains, about A. T.) of substance are mixed in a high crucible (Fig. 39, p. 66) with 60 to 120 grammes (925.94 to 1851.88 grains) of potassium carbonate and flour; upon this is placed 50 to 100 per cent. of borax, next 10 to 15 grammes (154.32 to 231.48 grains) of granulated lead, and finally a covering of common salt. It is fused in the muffle-furnace for 1 to 2 hours.

3. Combined lead and silver assay.-This method of assaying is used for oxidized lead products (litharge, skimmings, dross), and for galena with at least 30 to 40 per cent. of lead, and not over 0.12 per cent. of silver. Such galena, after the assay with potassium carbonate,

as described on page 92, is fused to a lead button, which is then cupelled. This method is not satisfactory after the preliminary assay with iron (p. 86), as the iron sulphide will retain silver in varying quantities in the slag.

Litharge: 20 grammes (308.64 grains, ab. & A. T.) litharge, 15 grammes (231.48 grains) potassium carbonate and flour, and 5 to 6 per cent. of powdered charcoal, with a covering of common salt, are placed in a crucible and fused in the muffle-furnace. If necessary, several of the buttons are concentrated by scorification (p. 136) and cupelled. Skimmings and dross are charged in the same manner; but, should they be very impure, the lead buttons must be scorified with 4 to 8 times their weight of granulated lead before they are cupelled.

B. Cupellation of the argentiferous lead (assaying by the cupel or cupellation).-The lead buttons obtained according to A Nos. 1 to 3 (pp. 132-141) are subjected to oxidizing fusion. During this operation the lead is first oxidized; the lead oxide yields up oxygen partly to the foreign metals, and partly combines with their oxides; and if they are not too refractory (as, for instance, ferric and stannic oxide, etc.), enters with them into the cupel.

If the lead contains much antimony and zinc, it is apt to cause cracks in the cupels; copper colors them green, and the percentage of copper may be quantitatively determined within certain narrow limits by the intensity of the color.

The cupels (Fig. 44, p. 68), first carefully wiped out with the fingers, then all extraneous matter blown out, are thoroughly heated (ignited) in the muffle. They are arranged in two rows, six in each row, in the front third of the muffle.1 The lead buttons are now laid

1 Hempel's gas-furnace with oxidizing apparatus in Fresenius' Ztschr. xvi. 454; xviii. 404.

hold of with a pair of forceps (Fig. 46b, p. 74) and gently deposited, first in the front and then in the back row of crucibles. The mouth of the muffle is closed, the fire is urged on, and the lead fused as quickly as possible. The lead will at first be covered with a dull, dark film. As soon as this disappears, and the lead shows a lustrous, fuming surface, the mouth of the muffle is opened (with the exception of a low piece of charcoal which is left in it) for the admission of air to oxidize the lead, and the temperature is lowered, by ceasing to stir the fire, to lessen the loss of silver. Cold scorifiers are placed in several rows above each other back of the cupels, and, if the argentiferous lead is very rich, the cooling-iron (Fig. 48, p. 74), which should be frequently cooled off in water, is moved to and fro closely over the cupels. The correct temperature is indicated by the arising lead fumes whirling over the assays, and not slowly creeping over them or rising straight up; by the cupels glowing dark brown, by small scales of crystallized litharge (plumose litharge, Federglätte) showing themselves on the inner edge, and by a bright but not too wide border of litharge upon the lead. If the temperature is too low, the fume creeps slowly over the cupels; these become too dark, a dark rim of litharge is formed, and the lead ceases to "drive." This is called a "freezing" of the assay. Frozen assays, if again brought to "driving" by a higher heat, and generally some addition of lead, cause a considerable loss of silver. By too high a temperature, the lead fumes arise up straight, the cupels glow too brightly, neither plumose litharge (Federglätt) nor a rim of litharge shows itself, and the loss of silver increases. At the correct temperature small beads of litharge float upon the surface of the lead button, the heat to which it is subjected causing convection currents, which give the button a motion from below upwards

("driving"), and a convex surface from which the luminous beads and patches of litharge are continually thrown towards the sides, and are there absorbed by the cupel. As soon as these patches upon the diminishing lead become larger towards the end, the scorifiers, which had been placed back of the cupels, are removed, the cooling with the cooling-iron is stopped, and the fire is urged on. Towards

the end the last of the lead is absorbed, and the silver button presents itself colored with all the tints of the rainbow (brightening, coruscation), which gradually disappear, whereupon the button solidifies. (If the temperature has been too low, the surface of the button is dull and yellow, and the unabsorbed litharge forms lumps or scales about it; while otherwise, it is pure silver-white on top and bottom, and very lustrous.) If the bead is large the cupels are allowed to cool off slowly by drawing them to the front of the muffle, to prevent the buttons from “sprouting" or "spitting." They are then taken out upon a piece of sheet-iron, and the buttons (generally of 99.7 to 99.8 per cent. pure silver) are detached by means of a pair of pincers (p. 75), and brushed off with the buttonbrush (p. 75). Faultless buttons brightened at a sufficiently high temperature (smaller ones are round, larger ones hemispherical) should have a silvery lustre on the surface, be dull silver-white and crystalline on the bottom, and without rootlets. They are then weighed. The globule of silver obtained from the separately scorified and cupelled granulated lead, if this was not entirely free from silver, is placed in the scale-pan containing the weights; or, the silver percentage of the lead having been determined once for all, it may be deducted from the results obtained.

Assays of silver should agree very closely, and if properly conducted are of great accuracy. The results of duplicate assays should not differ

from each other more than one-half ounce Troy per ton of two thousand pounds. Should a greater difference be found, an additional assay should be made.

Smaller losses of silver occur in scorifying (which for this reason should be continued as long as possible, so as to obtain small lead-buttons requiring but a short time for "driving") than in cupelling, by the volatilization of silver, and by the silver oxide passing into the cupel with the lead oxide by which it has been oxidized (loss by cupellation, Kapellenzug). The loss increases with the temperature and the size of the button, and for this reason, with the time required for cupelling, as well as with the porosity of the cupels. The percentage of loss is considerably larger (2 to 4 per cent.) in smaller buttons (poorer ores) than in larger buttons (1 to 13 per cent.), but in the first case can generally not be determined by the balance. The smallest loss occurs in gas muffle-furnaces (Fig. 29, p. 55), which have no ventholes in the muffle. In case the ore contains tellurium, the button spouts at the moment of solidification, after brightening, and fine globules of the metal are thrown off and lost.

II. Wet Assays.

Balling's volumetric assay.—2 to 5 grammes (30.87 to 77.16 grains) of galena are fused in a porcelain crucible with 3 to 4 times its weight of a mixture of equal parts of saltpetre and soda. This is allowed to cool off; the contents of the crucible are lixiviated with water, heated in a porcelain dish, and filtered. The residue is decomposed with diluted nitric acid, and evaporated to dryness. The dry mass is then taken up with water acidulated

1 Fresenius' Ztschr. xiii. 171; Oestr. Ztschr. 1879, No. 27.