In assaying, more especially products direct from the mine, there is always found, when the rock is siliceous, a quantity of white sandy-looking substance, insoluble in acids, which is sometimes accompanied by a light gelatinous material very difficult to filter. This is variously described as "insoluble," "sand," "insoluble silicates," "gangue," or "rocky matter." It may be pure quartz; but oftener it is mixed with silicates from the rock containing the mineral. Some silicates, but not many, are completely decomposed by boiling with hydrochloric acid or aqua regia; and others are partly so, they yield a gelatinous precipitate of silica which greatly interferes with the filtering. It is a common practice with assayers to carry the first attack of the sample with acids to dryness, and to take up with a fresh portion of acid. By this means the separated silica becomes granular and insoluble, and capable of being filtered off and washed with comparative ease.
This residue may be ignited and weighed; and be reported as so much per cent. of "silica and silicates insoluble in acids." Unless specially wanted, a determination of its constituents need not be made. When required, the analysis is best made on the ignited residue, and separately reported as "analysis of the insoluble portion."
Silicon only occurs in nature in the oxidised state; but the oxide generally known as silica (SiO2) is common, being represented by the abundant minerals—quartz, flint, &c. Silica, combined with alumina, lime, oxide of iron, magnesia and the alkalies, forms a large number of rock-forming minerals. Most rock masses, other than limestones, contain over 50 per cent. of silica. The following are analyses of some of the commoner silicates; but it must be noted that these minerals often show great variation in composition. This is more especially true of chlorite, schorl, hornblende and augite.
| Silica, SiO2. | Alumina, Al2O3. | Ferric Oxide, Fe2O3. | Ferrous Oxide, FeO. | Lime, CaO. | Magnesia, MgO. | Potash, K2O. | Soda, Na2O. | Fluorine, Water &c. | |
| Potash-felspar | 65.2 | 18.2 | 0.2 | — | — | — | 14.7 | 1.5 | |
| Soda-felspar | 67.0 | 19.2 | — | 0.3 | 1.2 | 1.8 | 2.2 | 7.2 | |
| Lime-felspar | 43.3 | 35.4 | — | 1.3 | 17.4 | 0.35 | 0.5 | 0.9 | |
| Potash-mica | 45.7 | 33.7 | 3.1 | — | — | 1.1 | 7.5 | 2.8 | F (0.8) H2O (4.9) |
| Magnesia-mica | 39.1 | 15.4 | 7.1 | — | — | 23.6 | 7.5 | 2.6 | F (0.7) |
| Hornblende | 40.6 | 14.3 | 5.8 | 7.2 | 12.5 | 14.0 | 1.5 | 1.6 | |
| Augite | 50.0 | 3.7 | 2.4 | 6.6 | 22.8 | 13.5 | — | — | MnO (0.1) |
| Almandine (Garnet) | 39.7 | 19.7 | — | 39.7 | — | — | — | — | MnO (1.8) |
| Chlorite (Peach) | 32.1 | 18.5 | — | — | — | 36.7 | — | — | H2O (12.1) |
| Schorl | 37.0 | 33.1 | 9.3 | 6.2 | 0.5 | 2.6 | 0.7 | 1.4 | B2O3 (7.7) F (1.5) |
| China-clay | 46.7 | 39.6 | — | — | — | — | — | — | H2O (13.4) |
| Talc | 61.7 | — | — | 1.7 | — | 31.7 | — | — | H2O (3.8) |
| Serpentine | 42.9 | — | — | 3.8 | — | 40.5 | — | — | H2O (12.6) |
| Olivine | 39.3 | — | — | 14.8 | — | 45.8 | — | — |
Silicon, from a chemical point of view, is an interesting body. It combines with iron to form a silicide; and is present in this condition in cast iron. Only in the case of the analysis of this and similar substances is the assayer called on to report the percentage of silicon. Silicon is readily converted into silica by the action of oxidizing agents. Silica forms only one series of salts—the silicates—which have in many cases a complex constitution; thus there are a large number of double silicates, which vary among themselves, not only in the relation of base to acid (which is the essential difference), but also in the ratio of the bases between themselves (which varies with almost every specimen).
Silica is detected by heating the substance with a fluoride and sulphuric acid in a platinum-crucible. On holding a rod, moistened with a drop of water, over the escaping fumes, the white crust of silica formed on the drop of water shows its presence. The insolubility of a fragment of the mineral in a bead of microcosmic salt, is also a very good test; the fragment, on prolonged heating, does not lose its angular form.
There is no dry assay for this substance, nor volumetric method; when the determination is required, it is carried out gravimetrically and, generally, by the following plan.
If the sample contains oxides, sulphides, &c., in any quantity, these are first dissolved out by treatment with acid, evaporated to dryness, taken up with hydrochloric acid, and filtered. The dried residue is treated in the same way as the silicates. Some silicates are completely decomposed by such treatment; but it saves time (unless one is sure that no undecomposable silicate is present) to treat these in the same way as the others. On the other hand, there are some silicates which are only attacked with difficulty even by fusion with alkaline carbonates; consequently, it is always well to have the substance reduced to the finest state of division by careful powdering, as this greatly assists the subsequent action. With very hard silicates, the grinding away of the mortar in this operation will be perceptible; the foreign matter thus introduced must not be ignored. Previously igniting the substance sometimes assists the powdering; but it is best to use a steel mortar. The particles of steel can be removed by a magnet, or, where the nature of the substance will allow it, by boiling with a little dilute hydrochloric acid.
The dried and powdered material is intimately mixed with four times its weight of "fusion mixture" in a platinum-crucible or dish. It is then moderately heated over a Bunsen burner, and afterwards more strongly fused over a blast, or enclosed in a clay crucible in the wind-furnace. The action is continued until the fused mass is perfectly tranquil. With very refractory substances, the action must be long continued at a high temperature. When sufficiently cold, the crucible is examined to see that no particles of foreign matter are adhering to its outer surface. It is then transferred to a five- or six-inch evaporating-dish, where its contents are acted upon with warm water for some time. The "melt" will slowly dissolve, but the solution should be hastened by keeping the liquid moderately acid with hydrochloric acid. When the "melt" has dissolved, clean and remove the platinum-dish, and evaporate the solution to a paste. Continue the evaporation to dryness on a water-bath (not on the hot plate), and whilst drying stir with a glass rod, feeling at the bottom of the dish for any unfused particles, which, if present, can be detected by their grittiness. If there is much grit, it will be necessary to repeat the assay; but with a small quantity it will only be necessary to refuse the grit and silica after ignition.