CHEMICAL COMPOSITION OF THE WATER OF SALTON SEA. 



41 



remained practically constant. Some of the other constituents which occur in small 

 quantities have varied somewhat irregularly. This may be due in part to experimental 

 error in analysis, but principally to the method of collecting the samples, as it was found 

 very difficult to collect samples perfectly free from silt. 



The samples for the first analyses were collected and transferred to the laboratory 

 in a large carboy. The water, as needed for analysis, was siphoned off without disturbing 

 the container, which gave any sediment in the water an opportunity to sink to the bottom. 



Table 9. — Composilion of the yearly residues in percentages of the total anhydrous soUds. 



Sodium, Na 



Fotansium, K 



Lithium, Li 



Calcium, Ca 



Magnesium, Mg 



Alumina, AlsOi 



Ferric oxide, FejOi. . . . 



Silica, SiO. 



Chlorine, CI 



Sulphate radicle, SO>. . 

 Carbonate radicle, COj 

 Phosphate radicle, POi 

 Nitrate radicle, NOi. . . 

 Oxygen consumed 



1907. 1808. 1909. 1910. 1911 



31.289 

 0.648 



trace 

 2.803 

 1.812 

 0.016 

 0.002 

 0.259 



47.826 



13.411 

 1.854 

 0.003 

 0.051 

 0.026 



100.000 



31.496 

 0.652 

 0.003 

 2.785 

 1.791 

 0.015 

 0.002 

 0.218 

 47.868 

 13.310 

 1.796 

 0.003 

 0.047 

 0.014 



100.000 



32.027 

 0.647 

 0.003 

 2.5.S7 

 1.790 

 0.023 

 0.003 

 0.218 

 48.122 

 13.158 

 1.466 

 0.002 



6.014 



100.000 



32.569 

 0.607 

 0.004 

 2.352 

 1.693 

 0.013 

 0.002 

 0.174 

 48.339 

 13.139 

 1.098 

 0.002 



o.dos 



100.000 



32.671 

 0.646 

 0.003 

 2.240 

 1.678 

 0.024 

 0.007 

 0.171 

 48.678 

 13.147 

 0.829 

 trace 



6.009 



100.000 



Tabus 10. — Analysis of sill deposited hy Ih* 

 collected sample. 



In 1909 it was thought that this method of sampling, which would certainly give low results 

 for iron and aluminium, on account of their partial precipitation through loss of carbon 

 dioxide by the water on standing, might also give rise to some loss of calcium and mag- 

 nesium in the same way. The residue which had settled out in the bottom of the demi- 

 john from the samples collected in the year referred to was accordingly filtered off and 

 analyzed. The constituents are expressed in table 10 

 in parts per 100,000 of the water in the demijohn. 



These figures confirm the view already expressed 

 that iron and aluminium separate out from the Salton 

 Sea water on standing, which explains the somewhat 

 irregular results obtained for these constituents from 

 year to year. The amounts of calcium and magnesium 

 carbonates lost in this manner, however, are insignifi- 

 cant. Instead of being formed by precipitation from 

 the water, the greater part of the residue analyzed is 

 due to silt which settled out from suspension. 



The sample for 1911 was collected and filtered on the spot into graduated fla.sks. 

 The total contents of a flask were then taken for a determination. In this way it was 

 thought that inaccuracies would be avoided through loss of any of the constituents by 

 precipitation. It was found, however, that the water of the lake was quite noticeably 

 turbid and could not be obtained perfectly clear by filtration through filter paper. The 

 duplicate results obtained for silica, iron, and aluminium did not for this reason agree very 

 closely, and all are no doubt slightly high. 



In order that the analyses might represent the composition of the anhydrous inorganic 

 matter remaining when the water was evaporated to dryness, a gravimetric determination 

 was made each year of the combined carbon dioxide in the residue. As shown in the table, 

 this amounted in 1911 to 5.78 parts as CO, in 100,000, equal to 4.24 parts of CO,. The 

 bicarbonates in solution were also determined by titrating with potassium acid sulphate, 

 using methyl orange as indicator. The value found amounted to 17.14 parts as HCOi 



