PROCEEDINGS

of the Tomsk Technological Institute

of Emperor Nicholas II.

Vol. 10. 1908. No. 2.

III.

D. P. Turbaba.

On the Composition of Siberian Mineral Waters.

1–6.

On the Composition of Siberian Mineral Waters.

At the end of July 1904, I took water from Lake Shira and from the nearby lakes Itkul and Bilë for chemical analysis. At the same time, thanks to the kind assistance of A. A. Barrok (to whom I offer my sincere gratitude), I received two ordinary bottles of water from Lake Uchum, situated approximately 80 versts from Lake Shira. On this Lake Uchum, a (primitive) resort had also been opened a year or two earlier. On site almost no tests of these waters were carried out; their analysis was begun only in mid-August 1904, after my return to Tomsk.

The methods of analysis were the usual ones (see, for example, B. A. Gemilian’s Guide to the Examination of Water, Warsaw, 1906); departures from the above-mentioned guide were, for example, the determination of magnesia by Neubauer’s newest method (see Treadwell, Quantitative Analyse, 1902), and the first operation in determining the alkalies—the addition of barium chloride to remove all sulfuric acid.

I. Lake Shira.

Both water from the bathing place was taken, i.e. near the very shore, and water drunk by visitors to the resort, i.e. drawn at a considerable distance from the shore and filtered. The specific gravity of both waters is the same; at 21.9° C it is 1.0161.

Only the drinking water was examined.

1 liter of it contains, in grams (direct data of the analysis):

Note. The reaction of the water is alkaline. The combined carbonic acid was determined by prolonged—over several hours—boiling of a definite volume of water, and then, as usual, by the action of hydrochloric (or sulfuric) acid and absorption of the carbonic acid by caustic lime.

Particular attention was given to the determination of the alkalis. Their sum—in the form of chlorides—was found to be 11.190 g per liter, as the mean of four closely agreeing determinations. This number was further checked by evaporating a definite volume of water with an excess of sulfuric acid, igniting the residue to constant weight, and from this weight subtracting the other sulfates (CaSO₄ and MgSO₄) and SiO₂ + Al₂O₃ + Fe₂O₃, according to the analysis given above. Knowing the quantity of potassium, I find from the weight of these alkaline sulfates that the sodium and potassium chlorides should give, from 1 liter, 11.150 g. Student Privalov also obtained a fairly close number: from 1 liter there is obtained KCl + NaCl—10.92 g; but he himself acknowledges that the barium salts still retained some quantity of alkalis.

The question of which acids in the water are combined with precisely which bases—a question that is insoluble (for example, in the present case, where there are 3 acids and 4 bases)—falls away according to the newest views, since free ions are present in the water in the greatest quantity. I therefore express it in ions. Then 1 liter of water of Lake Shira contains ions in grams:

*) In view of the small weight, it is not worth converting into ions.

Agreement with the weight of the ignited dense residue is excellent. Almost the same sum, namely 19.823 g, is obtained if one assumes that all the chlorine is combined with sodium, and the remaining quantities of bases are expressed as anhydrous oxides. My analysis (1904) comes closest to the analysis of N. S. Kastorsky in 1901 (see A. Bychkov’s brochure, “Lake Shira as a Curative Lake”) and shows, if one compares the analyses of Prof. Leman in 1890, Prof. E. V. Werner in 1895, N. S. Kastorsky in 1901, and D. Turbaba in 1904, that the water of Lake Shira is fairly rapidly enriched in its chief constituents: chlorine, sulfuric acid, sodium sulfate, and magnesium. The clear difference between my analysis and the analysis of N. S. Kastorsky consists in the fact that 1) potassium is unquestionably present in the water of Shira according to my determinations, whereas N. S. Kastorsky says that “neither in the precipitate nor in the filtrate was any potassium found at all”*) (A. Bychkov, p. 13), and 2) the quantity of combined carbonic acid according to me is approximately three times less than according to N. S. Kastorsky.

II. Lake Bilë.

It consists of 2 lakes connected by a rather narrow channel. The water from the farther Lake Bilë was investigated, if one travels from Shira.

The reaction of the water is clearly alkaline. The specific gravity at 22° C is equal to 1.0063.

1 liter of water contains, in grams (direct analytical data):

Note. By ammonia and ammonium oxalate the oxides of aluminum, iron, calcium, and silicon were precipitated together and were not separated, in view of their insignificant quantity compared with

*) Prof. Leman found potassium there.

with the remaining constituents of the solid residue. My analysis is very close to F. V. Ludwig’s analysis of 1899 (“Materials for the Study of the Chemical Composition of Some Bitter-Saline Lakes of the Steppes of the Minusinsk District,” Yuriev, 1903); only here, over 5 years, an increase in the solid residue is observed; the amounts of chlorine, sulfuric acid, and sodium have increased, although only slightly. Potassium was not determined, even qualitatively, through an oversight of mine, and the alkali chlorides obtained were taken as pure sodium chloride (however, this, in all probability, introduces only a negligible error, since all the lakes around Shira contain far less potassium than sodium).

Expressing the principal constituents in ions, I have:

1 liter contains ions in grams

This sum agrees very closely with the weight of the ignited solid residue.

III. Lake Uchum.

The reaction is distinctly alkaline. The specific gravity at 22.0° C is 1.0195.

1 liter contains, in grams (direct analytical data)

*) As before, because of the small quantity I do not convert it into ions.

Note. Again, as in the analysis of the water of Lake Shira, I paid special attention to the determination of the alkalis; the sum of the potassium and sodium chlorides, equal to 20.400 gr., is taken as the mean of three well-agreeing determinations. Expressing in ions, we have:

1 liter contains ions in grams

IV. Lake Itkul.

Its water serves, as ordinary drinking water, for the visitors of the Shira resort. It lies approximately three versts from Lake Shira. The water was taken from a water-pump set near the very shore. The water is colorless, transparent, and has no odor.

1 liter contains in grams

Note. The amount of chlorine—0.0346 g—determined by the weight of silver chloride, I have given here in the form of sodium chloride (NaCl), i.e., all the chlorine was combined with sodium; the remaining amount of sodium I have given as sodium oxide (Na₂O). Unfortunately, I took little water—about 3 liters—and, when I accidentally spoiled the determination of magnesia, I therefore could not repeat it; the given figure for magnesia—0.1323 g—represents only a minimum, i.e., some of the magnesia was lost by me. The water of Itkul, unfortunately, is polluted: cattle, entering the lake, drink near the water pump. While taking the water, I myself observed this. That is why there are more organic substances than are ordinarily accepted for normal drinking water*) (see Hemilian, “Guide to the Investigation of Water,” p. 10). Originally, in the water of Lake Itkul the weight ratio of magnesia to sodium oxide is: in Lakes Shira and Belyo sodium strongly predominates over magnesia; in Lake Itkul magnesia predominates.

In addition, I determined the electrical conductivity in ohms of the waters of Lakes Shira and Uchum. It turned out that cylinders of water 1 meter long and 1 square millimeter in cross-section have a conductivity in ohms at 18.2° C:

*) Normal water, for the oxidation of its organic substances, should not require more than 0.0025 g of oxygen.