It only remains to fill a deep pan or boiler, three parts full with Saltpetre water, and to throw upon it one fifth by weight of the water obtained as abovementioned, the solution is then to be well mixed, and left to settle, when it is to be drawn off from the surface, leaving the sediment.

The expense of this operation is a very small matter, compared to the quantity of nitrate of potass recovered which would otherwise be lost-and the saltpetre will be obtained of a fine white colour.

On the decomposition of the earthly nitrates by means of ashes. Troughs the same as are required to wash the sa tpetre earth, or wine barrels with double bottoms pierced with small holes, may be used for this purpose.

A small quantity of long straw is to be placed between the two bottoms, and also over the false bottom, covered by a piece of coarse flannel or serge.

Having prepared fresh wood ashes, by passing them through a sieve to separate the charcoal, they are to be moistened with water very equally, so that when taken in the hand they may form in lumps, in this state they are to be spread over the flannel, in the first înstance to the depth of six centimetres, (about 2 inches) and pressed down lightly at first, and afterwards more strongly by a flat beater, continue to add fresh layers of ashes pressed down as beforementioned, until the trough or cask, is about half full, the surface being levelled each time with a trowel. It is necessary to place over the trough in which the ashes are placed a hurdle, or basket with some straw in it, to prevent the loss of water.

All that is necessary then, is, to pass the Saltpetre water through the ashes, in the proportion of fifteen measures of water, for one of ashes. After which the ashes which remain may be mixed with fresh saltpetre earth, to disengage by washing any saltpetre remaining in them.

If the use of ashes does not altogether supercede the employment of potass, still they will assist, and diminish the quantity of alkali required to be used, according to the proportion of alkali, contained in the ashes.

To ascertain the quality of the ashes to be used, take about 25 lbs. of them, and wash with water, till no salt remains in the ashes; filter this water, and then evaporate it to dryness in an iron pan, taking care to remove the different salts which may precipitate, during the operation, and to constantly stir the salts at the time of desiccation, in order that they may not adhere to the bottom of the

pan-the produce thus obtained will shew the quantity of alkaline matter, changed with a part colouring and extractive, and a portion of foreign salts. If this produce is calcined in a crucible, the pure potass alone will remain.

Mode of conducting the evaporation.

over

To proceed with advantage in this operation, it is necessary, first. That the saltpetre water should be of sufficient strength to prevent the unnecessary expenditure of fuel-second. That the water should have been sufficiently treated with potass, to admit of the evaporation being continued to the extent required, without that swelling or o boiling which is hurtful-and lastly. That there should be a sufficient quantity of water collected, before the operation is commenced upon, not only to fill the boiler, but to fill it up, as the water evaporates, and by continuing this operation without ceasing to obtain by one process as much Saltpetre as possible.

In general it is advisable to stop the evaporation at the point when the nitrate of potass and the water are in equal parts as the operation can be better conducted, with less obstruction from the earthy salts, and other refuse more easily and exactly separated, than if pushed further.

Accordingly it is only by knowing the capacity of the boiler, that a calculation can be made. If for instance the water shews a density of 15 degrees by the Aréometre for nitre-which indicates that it contains 0.12 of nitre- and if the boiler contains 1200 Kilogrammes then this quantity of saltpetre water, may be considered to consist of 1020 Kilogrammes of water, and 180 Kilogrammes of salts, or about 144 Kilogrammes nitrate of potass, and 36 of Muriates, in order therefore to bring the nitre and water to equal parts, 876 Kilogrammes of that salt must be added, this quantity of nitre is represented by 7,300 Kilogrammes of saltpetre water at 15 degrees, or 12 per cent.

Accordingly it is this quantity of 7,300 Kilogrammes equivalent to 68 hectolitres (about 1,743 English gallons) with which it is necessary to fill up the evaporating boiler; when all this additional water has been evaporated the operation may be brought to a conclusion.

Having collected a sufficient quantity of saltpetre water to carry through one operation, the boiler is to be filled, and brought to boil, and replenished gradually as the water evaporates, this is best done by placing a pan, kept constantly full of this water, near the surface of the boiler and by means of a stop cock, to let off the water gradu

ally as required, so that the boiling may not be checked by a large addition of water at any time-this pan may be heated by a flue from the furnace, by which means a considerable expenditure of fuel will be saved.

As a considerable quantity of earthy matter &c. will soon begin to fall to the bottom of the boiler, and form a hand cake, which is very injurious to the metal, a small pan ought to be placed in the centre of the solution, raised two inches from the bottom of the boiler, in which this earthy matter will collect, this pan is to be suspended by means of iron chains, connected by a rope to a pulley above, so that it can be raised up from time to time, and its contents emptied into a box or basket placed at the edge of the boiler, to allow the water to drain back. It is essential that the solution should boil at an equal steady temperature, increasing or diminishing the fire as may be necessary, and the skum ought to be carefully and frequently removed.

When all the extra water has been added, the pan is to be withdrawn from the bottom of the boiler-The common salt will then begin to fall, and must be removed with a scoop ladle from the bottom, and placed, in a basket as beforementioned to drain.

It is proper at this period to decrease the fire which delays the operation, but the results are thereby obtained with more regularity. When the common salt falls to the bottom in abundance, the solution is approaching the proper degree of concentration, which may be easily ascertained by taking out a small portion of the liquid and placing it in a vessel to crystallize. &c.

When it is judged proper to stop the operation the fire is to be decreased, and the solution left to repose, till the common salt ceases to fall, or about five hours, during which time the solution ought to be kept steadily at the temperature of about 88 Centigrade thermometer or 190 of Farenhiet when it may be drawn off from the surface into crystallizing pans, taking every care to leave undisturbed, the salt or sediment at the bottom. It will require three days to complete the crystallization in winter, and longer in summer, and during hot weather fewer crystals will be obtained.

The mother water is then to be removed from the pans, and they are to be put up to drain, two and two, inclining towards each other in a channel.

In the manufactories of the Administration the crystallization is determined, by a much more easy and expeditious operation, the water is run off, into a long and large reservoir, lined with sheet lead or copper, the interior dimensions of which is about 12 feet

long, by 7 feet broad at the surface, the bottom is formed with a double slope of four inches towards the centre, one from the sides, the other longitudinally, so that the depth at one end is 16 inches and at the other end 20 inches.

The water is kept constantly in motion by means of wooden rakes to assist in cooling it, and the crystals as they form are raked to one end of the reservoir heaping them up to drain, and removing the most elevated parts, as they whiten perceptibly, to be carried to troughs or baskets and left to drain. In thus removing successively the small crystals as they form, the agitation, of the water with rakes must be continued incessantly to prevent the Saltpetre from forming in large cakes.-When the water has fallen in temperature to within 4 degrees of the place, or in about six hours all the Saltpetre will have been obtained, and the water will remain collected in the centre and towards one end of the reservoir from whence it is easily removed..

On the treatment of the water of crystallization.

This water will be left fully saturated with nitre and common salt, with a portion of earthy nitrates which may not be decomposed and ought to be again treated with a small proportion of potass, or passed over ashes, and then evaporated-or this water may be mixed with other Saltpetre water, before the latter has been treated with potass. If it is evaporated separately, the water which again remains may be abandoned or sold.

In France the earthy substances removed from the bottom of the boiler, is washed to recover the nitre remaining in it, the skum is boiled up in good water for the same purpose, and the refuse salt, which also contains a considerable proportion of Saltpetre is treated as follows.

* A boiler is to be filled three parts full of water and brought to boil; pure common salt is then thrown in till the water is fully saturated, which will be known by some of the salt falling to the bottom, the excess is removed with a scoop. Have in readiness some

baskets filled with the refuse salt, one of these baskets, is to be immerged in the boiling solution, which cannot dissolve any more common salt, and can only act upon the nitre contained in it, in a short time, this basket may be withdrawn, and another put in, and so on till the water in the boiler is fully saturated with nitrate of potass, when it may be drawn off to crystallize. The baskets are put into the boiler by means of ropes, and a pulley fixed to a Crane. I have made trial of this system of treating the refuse salt, and I find that it will amply re-pay the cost, by the quantity of Saltpetre recovered.

VII.-Assay of Saltpetre according to the principle established by M. M. Botté et Riffault in France.

This process consists in using a dissoluent of salts of earthy bases and of muriate of Soda which at the same time will not act upon the nitrate of potasse, and in leaving it undisturbed would shew the exact quantity of pure nitre contained in any given quantity of grough Saltpetre; this dissoluent is water fully saturated with pure nitre; if care is taken to obtain the saturated solution at the time of using it, in a fully saturated state.

The following is the method to be pursued.

Method of preparing the saturated

solution.

pure

nitre

Take about two or three pounds of pounded and sifted, and throw it by degrees into a copper vessel tinned inside about 12 inches deep, and 5 inches diameter pouring in at the same time, good water heated to 126 degrees of Fahrenhiet, to be well stirred, and left for three or four hours or during the night, there ought always to be a portion of pure nitre not dissolved at the bottom, and the solution ought never to be used without stirring it well up from the bottom, and allowing it only one minute to settle, each time of using it."

The Saltpetre received is to be collected in a heap, and well mixed, and a quantity taken from different parts of the mass, from which a sample; or any number of samples of 100 drams avoirdupois each is to be taken, the saltpetre taken for assay must be well

bruised in a mortar.

One of these samples of 100 drams is to be placed in a glass vessel (a common finger glass or cooler is well adapted for this operation) and half a pint of the saturated solution poured upon it, and constantly stirred with a glass rod for 15 minutes, then allow it to settle for an instant, and pour off the surface water into a paper filter placed in a funnel over any vessel, this paper filter is necessary in order to retain any nitrate of potasse which may escape by carelessness in pouring off the surface water.

The Saltpetre in the glass vessel is then to be washed a second time with a quarter of a pint of the saturated solution, and stirred with a glass rod constantly for 15 minutes, allowed to settle and

In France where this mode of assay is pursued, an instrument called an Aréometre, with a thermometer are used to ascertain, that the solution is fully saturated up to the temperature of the place before using it, but the assay may be conducted with great accuracy without these instruments if the rules here laid down are strictly observed.