To produce mineral waters that will yield, by analysis, the same component parts combined in the same proportions and manner as in the natural springs, is an operation that requires considerable chemical skill, and should not be attempted by in-experienced persons. It is rarely possible to reproduce a natural water by the addition of the exact ingredients found in it by analysis. It is usually necessary to use chemicals which, upon combining, produce those ingredients. For instance, carbonate of magnesia and carbonate of lime, unless freshly precipitated, will not make a clear solution. Therefore, in order to produce this, it is customary to employ other substances which will dissolve at once and produce upon combining those ingredients.

A glance at the results of the analyses will convince that hardly all the detected constituents of a natural mineral water, however they may be grouped together, are readily soluble in water. Especially is this true of the compositions of magnesia, lime, barita, strontian, iron, manganese, alumina with carbonic acid, phosphoric and silicic acid, partly also with sulphuric acid. These are difficult of solution or absolutely insoluble in water under ordinary circumstances and in natural spring water only by the influence of free carbonic acid, pressure or warmth; also by chemical action or the influence of the other constituents. It would be wasted time and difficult to try and add all the ingredients the analysis revealed, in their dry state, to water charged with carbonic acid, and induce their solution by constant agitating. Their entire solution could not be accomplished. Those constituents which are difficult to dissolve are best produced at the moment when they are wanted from solutions that produce them upon combining with others. This is done in such combinations, the chemical components of which permit the momentary production of hydrates, otherwise they must be added to the water in their ready-made state. For instance, if artificial water needs the addition of the insoluble carbonate of lime, say 50 grains, it is necessary to take about 55.5 grain calcium chloride and 53 grain dry carbonate of soda, both in solution. A fine precipitate of carbonate of lime forms, which is easily and quickly soluble in the carbonated water.

Another product which would form is sodium chloride, 58.3 grains. The production of carbonate of lime from those salts is therefore only allowed if the artificial water is to contain sodium chloride also, otherwise the carbonate of lime must be'added as such so soon as precipitated in its freshly precipitated and moist state. In a similar way all the other constituents that are difficult or absolutely insoluble under ordinary circumstances must be produced and con-verted into a state of solubility, if this can possibly be done. This requires considerable chemical knowledge.

To enable the manufacturer of mineral waters not familiar with chemistry to imitate all the known and analyzed natural mineral waters prominent in the trade by such substitutes as yield, by analysis, the same component parts combined in the same proportion and manner as in the natural springs, is the aim of the appended formula?.

For some of the calculations we are indebted to Dr. Hager's Adju-menta varia chemica et pharmaceutica clique subsidia ad parandas aquas minerales, and to Dr. Hirsch's Die Fabrication der Kunstliclien Mineral Wasser, which we have altered to suit our purpose, and allow the employment of the chemicals as manufactured for commercial and pharmaceutical application. The calculations for American mineral waters we have made very carefully on the foundation of the latest analyses of well-known chemists, published in the "Bulletin of the United States Geological Survey, No. 32".

All the proposed imitations and substitutes we explain expressly hereafter, and append special directions for practical compounding. On the basis of these calculations, which we may term "the recipes for imitations," any manufacturer is enabled without any trouble and chemical knowledge to make a true substitute of the natural mineral water, if he follows instructions closely.

The substitutes are classed in groups in such a way that none will separate or precipitate the other if added to the water in the succession stated, and in either the dry or liquid form as specially directed, and each group is separated by a line from the other, making it more easily distinguishable.

The analyses refer to dry substances, and the calculations for artificial combinations refer to the same rule, except where such ingredients are required that appear in commerce, principally in a solid, but hydrated state, containing several equivalents of water, viz: Alum, barium chloride, iron sulphate, magnesium carbonate, magnesium sulphate, manganese sulphate, sodium carbonate, sodium sulphate, and the specially prepared calcium sulphate praecipitat. All the calculations refer to avoirdupois weight. (1 lb. =16 ounces; 1 ounce =8 drachms or 480 grains).

All the chemical substances are products of chemical manufacturing, and bought cheaper than the carbonator could make them; however, we have given some directions for their preparation in either dry, liquid or precipitated state, where we deemed it necessary for the benefit of the trade, and where the necessary solutions are better and more practically prepared directly in the bottler's laboratory.

It is quite convenient to have those salts, the chemical properties of which allow them to be kept any length of time in solution without being decomposed by oxidation or otherwise, in stock in solutions, handy for immediate use. We have, therefore, in the next part of this Chapter given special directions for the preparation of ready-made solutions of most of the components, where it is advisable. These solutions are prepared to contain either one, five or ten per cent., or one part in one thousand parts of the salt in solution, as specially directed. Instead of weighing the required quantities, which is sometimes inconvenient where but minute quantities are required, they are then measured (fluid measure). The proportions are then:

Of a one per cent. solution, . . . 100 to 1

" five " " 100 " 5

" ten " " .. 100 " 10

Or, what is the same, . . . . 10 " 1

If one part be dissolved in 999 parts of water to make it 1000 parts, as, for instance, directed for carbonate of lithium, the proportion is 1000 to 1, - that is, instead of one grain of the salt 1000 grains of the solution are employed; and if, for instance, 0.32 grain of chloride of strontium is required, a solution of which is made by dissolving one grain or one part in 99 parts of water to make it 100 parts, or a one per cent, solution, we employ 100 times 0.32 = 32 grains (fluid measure), which is more conveniently measured in a little graduate than weighed on the scales.

This is easily understood. Each bottle is marked by a label indicating its contents and strength of solution, viz.: "Chloride of sodium solution 10 per cent; proportion 10 to 1". The solutions can be prepared at leisure, and are handy for use. They ought to be filtered before stored in the laboratory. As the proportions of water combined with many commercial chemical ingredients of artificial mineral waters are varying, it is in most cases necessary to prepare all solutions with the hydrometer and regulate the specific gravity of the solutions, in order to determine the quantity of salt dissolved in them.

We append hereafter the standard specific gravities of the different solutions where we deem it necessary, when prepared either from anhydrous salts or commercial hydrates. Some salts are dissolved for immediate use, as their solutions would be decomposed by oxidation or otherwise, such as bromides, iodides, etc., also silicate of soda, when kept in solution; in regard to this we have appended special remarks to the descriptions of the chemical properties of the components.