(see Nitre, Chemistry-Index, in this Suppl.) is an article of so much importance, and sometimes so difficult to be had, that it is wonderful more attention is not bestowed in endeavouring to discover some easy method to increase the quantity. Such a method has been long practised by the farmers of Appenzell in Switzerland. In so hilly a country, most houses and stables are built on slopes, one side of the edifice resting on the hill, and the other being supported by two strong polls, elevated two or three feet above the ground; so that the air has a free current under the building. Immediately under the stable a pit is dug, usually occupying both in breadth and length the whole space of ground covered by the building; and instead of the clayey earth which is dug out, the pit is filled up with sandy soil. This is the whole process, and all the rest is done by nature. The animal water, which is continually oozing through the planks of the floor, having drenched the earth contained in the pit for the space of two or three years, the latter is emptied, and the saltpetre is refined and prepared in the usual manner.

That manner, however, is not the best; and the French chemists, during the incessant wars occasioned by the revolution, have, for the sake of supplying their armies with gunpowder, turned their attention to the best method of refining saltpetre. The following are directions given for this purpose by Chaptal, Champy, and Bonjour.

The crude saltpetre is to be beaten small with mallets, in order that the water may more easily attack every part of the mass. The saltpetre is then to be put into tubs, five or six hundred pounds in each tub. Twenty per cent. of water is to be poured into each tub, and Saltpetre and the mixture well stirred. It must be left to macerate or digest until the specific gravity of the fluid ceases to augment. Six or seven hours are sufficient for this first operation, and the water acquires the density of between 25 and 35 degrees. (Sp. gr. 1.21, and 1.306, ascertained by Baume's hydrometer. See Hydrometer, Suppl.

The first water must then be poured off, and a second portion of water must be poured on the same saltpetre amounting to 10 per cent.; after which the mixture must be stirred up, suffered to macerate for one hour, and the fluid drawn or poured off.

Five per cent. of water must then be poured on the saltpetre; and after stirring the whole, the fluid must be immediately drawn off.

When the water is drained from the saltpetre, the salt must be thrown into a boiler containing 50 per cent. of boiling water. When the solution is made, it will mark between 66 and 68 degrees of the hydrometer. (Sp. gr. 1.848, and 1.898.)

The solution is to be poured into a proper vessel, where it deposits by cooling about two-thirds of the saltpetre originally taken. The precipitation begins in about half an hour, and terminates in between four and six hours. But as it is of importance to obtain the saltpetre in small needles, because in this form it is more easily dried, it is necessary to agitate the fluid during the whole time of the crystallization. A slight motion is communicated to this liquid mass by a kind of rake; in consequence of which the crystals are deposited in very slender needles.

In proportion as the crystals fall down, they are fanned to the borders of the vessel, whence they are taken with a skimmer, and thrown to drain in baskets placed on trellises, in such a manner that the water which passes through may either fall into the crystallizing vessel, or be received in basins placed underneath.

The saltpetre is afterwards put into wooden vessels in the form of a mill-hopper or inverted pyramid with a double bottom. The upper bottom is placed two inches above the lower on wooden ledges, and has many small perforations through which water may pass to the lower bottom, which likewise affords a passage by one single aperture. A reservoir is placed beneath. The crystallized saltpetre is washed in these vessels with 5 per cent. of water; which water is afterwards employed in the solution of saltpetre in subsequent operations.

The saltpetre, after sufficient draining, and being dried by exposure to the air upon tables for several hours, may then be employed in the manufacture of gunpowder.

But when it is required to use the saltpetre in the speedy and immediate manufacture of gunpowder, it must be dried much more strongly. This may be effected in a floe, or more simply by heating it in a flat metallic vessel. For this purpose the saltpetre is to be put into the vessel to the depth of five or six inches, and heated to 40 or 50 degrees of the thermometer (or about 135° of Fahrenheit). The saltpetre is to be stirred for two or three hours, and dried to such that, when strongly pressed in the hand, it shall acquire no consistence, nor adhere together, but resemble a very fine dry sand. This degree of dryness is not required when the powder is made by pounding.

From these circumstances, we find that two saline liquids remain after the operation; (1) the water from the washing; and (2) that from the crystallizing vessels.

We have already remarked, that the washing of the saltpetre is performed in three successive operations, in which, upon the whole, the quantity of fluid made use of amounts to 35 per cent. of the weight of the crude saltpetre. These washings are established on the principle, that cold water dissolves the muriats of soda, and the earthy nitrates and muriats, together with the colouring principle, but scarcely attacks the nitrat of potash.

The water of these three washings therefore contains the muriat of soda, the earthy salts, the colouring principle, and a small quantity of nitrat of potash; the amount of which is in proportion to that of the muriat of soda, which determines its solution.

The water of the crystallizing vessels contains a portion of the muriats of soda, and of the earthy salts which escaped the operation of washing, and a quantity of nitrat of potash, which is more considerable than that of the former solution.

The waters made use of at the end of the operation, to whiten and wash the crystals deposited in the pyramidal vessel, contain nothing but a small quantity of nitrat of potash.

These waters are therefore very different in their nature. The water of the washings is really a mother water. It must be collected in vessels, and treated with potash by the known processes. It must be evaporated to 66 degrees (or 1.848 sp. gr.), taking out the muriat of soda as it falls. This solution is to be saturated with 2 or 3 per cent. of potash, then suffered to settle, decanted, and poured into crystallizing vessels, where 20 per cent. of water is to be added to keep the whole of the muriat of soda suspended.

The waters which are thus obtained by treatment of the mother water may be mixed with the water of the first crystallization. From these the marine salt may be separated by simple evaporation; and the nitrat of potash, which they hold in solution, may be afterwards obtained by cooling.

The small quantity of water made use of to wash and whiten the refined saltpetre, contains nothing but the nitrat of potash; it may therefore be used in the solution of the saltpetre when taken from the tubs.

From this description it follows, that a manufactory for the speedy refining of saltpetre ought to be provided with (1) mallets or rammers for pounding the saltpetre; (2) tubs for washing; (3) a boiler for solution; (4) a crystallizing vessel of copper or lead, in which the saltpetre is to be obtained by cooling; (5) baskets to drain the crystals; (6) a wooden cage or hopper for the last washing and draining the saltpetre; (7) scales and weights for weighing; (8) hydrometers and thermometers, to ascertain densities and temperatures; (9) rakes to agitate the liquor in the crystallizing vessel; (10) skimmers to take out the crystals, and convey them to the baskets; (11) syphons or hand-pumps to empty the boilers.

The number and dimensions of these several articles must vary according to the quantity of saltpetre intended to be refined.

Gum-SANDARAC, is said, in the Encyclopaedia, to be produced from a species of juniper. This was long the common opinion; but M. Schouboe has lately proved (a) it to be a mistake. The juniperus communis, from which many have derived this gum, does not grow in Africa; and Sandarac seems to belong exclusively to that part of the world. The gum sandarac of our shops is brought from the southern provinces of the kingdom of Morocco. About five or seven hundred quintals of it are exported every year from Santa Cruz, Mogador, and Safi. In the language of the country it is called el griffa. The tree which produces it is a Thuya, found also by M. Vahl in the kingdom of Tunis. It was made known several years ago by Dr Shaw, who named it Cyprissus fruticosa quadrivalvis, Equifoli inflar articulatis; but neither of these learned men was acquainted with the economical use of this tree; probably because, being not common in the northern part of Barbary, the inhabitants find little advantage in collecting the resin which exudes from it.

M. Schouboe, who saw the species of thuya in question, says that it does not rise to more than the height of twenty or thirty feet at most, and that the diameter of its trunk does not exceed ten or twelve inches. It distinguishes itself, on the first view, from the two other species of the same genus, cultivated in gardens, by having a very distinct trunk, and the figure of a real tree; whereas in the latter the branches rise from the root, which gives them the appearance rather of bushes. Its branches also are more articulated and brittle. Its flowers, which are not very apparent, they themselves in April; and the fruit, which are of a spherical form, ripen in September. When a branch of this tree is held to the light, it appears to be interspersed with a multitude of transparent vesicles which contain the resin. When these vesicles burst in the summer months, a resinous juice exudes from the trunk and branches, as is the case in other coniferous trees. This resin is the sandarac, which is collected by the inhabitants of the country, and carried to the ports, from which it is transported to Europe. It is employed in making some kinds of sealing-wax, and in different sorts of varnish. In 1793 a hundred weight of it cost in Morocco from 13 to 13½ piastras, which make from about L.3, 5s. to L.3, 7s. 6d. sterling. The duty on exportation was about 7s. 6d. sterling per quintal.

Sandarac, to be good, must be of a bright-yellow colour, pure and transparent. It is an article very difficult to be adulterated. Care, however, must be taken, that the Moors do not mix it too much sand. It is probable that a tree of the same kind produces the gum sandarac of Senegal, which is exported in pretty considerable quantities.