CHEMICAL COMPOSITION OF THE WATER OF SALTON SEA. 37 



For the direct estimation of the iron the mixed oxides were digested in dilute sul- 

 phuric acid until all were completely dissolved. The solution was transferred to a small 

 Erlenmeyer flask and oxidized by boiling with a little bromine water. The excess of 

 bromine was expelled by boiling, and the solution then poured into a Nessler tube. 



In other tubes were placed known amounts of a standard iron solution made by dis- 

 solving a weighed amount of iron in sulphuric acid and oxidizing with bromine water, 

 after which the solution was made up to a known volume. To each of the tubes contain- 

 ing the standard solution was added an amount of sulphuric acid equal to that in the tube 

 containing the unknown amount of iron ; 5 c.c. of a 5 per cent solution of potassium sulpho- 

 cyanide were then added to each tube. The tubes were finally filled to the mark with water 

 and mixed. By thus making up standard tubes until one was obtained which agreed in 

 color with the one containing the unknown amount of iron, the amount of iron originally 

 present in the water was determined. The aluminium was calculated by difference. 



MANGANESE. 



The filtrate from which the mixed oxides of iron, aluminium, and phosphorus were 

 precipitated was concentrated somewhat and acidified with a slight excess of hydrochloric 

 acid. Bromine water was added until the solution was colored dark brown, after which 

 it was treated with a considerable excess of ammonia and the solution was heated to boil- 

 ing. The negative result obtained in this way for manganese was confirmed by making 

 a qualitative test with lead oxide and nitric acid on the residue from a separate portion 

 of water. 



CALCIUM AND MAGNESIUM. 



The solution in which the manganese was found to be absent was used to determine 

 calcium and magnesium according to the usual methods. 



SULPHATE RADICLE. 



For the determination of sulphates a separate portion of 500 c.c. of the water was 

 taken in 1907, but as the water became more concentrated a correspondingly less amount 

 was used. This was evaporated to convenient volume and the sulphates determined by 

 the usual gravimetric method. 



SODIUM, POTASSIUM. AND LITHIUM. 



The usual methods were likewise employed in determining sodium and potassium in 

 the filtrate from the sulphate determination. 



For the estimation of lithium a separate preparation of the mixed alkali chlorides 

 was prepared. Part of the sodium was removed by dissolving in a small amount of water 

 and adding alcohol. The filtrate from the sodium precipitate was washed with 80 per 

 cent alcohol, evaporated to dryness, and the lithium line (6708.2^) given by the residue 

 in the spectroscope was compared with the lines given by mixtures of sodium and lithium 

 chlorides. Knowing the ratio of sodium and lithium in the mixture which gave the same 

 intensity of line as the residue containing the unknown amount of lithium, the extent to 

 which this constituent occurs in the Salton Sea water could be calculated. 



COPPER. 



Two liters of water were evaporated to dryness in a porcelain dish and the silica 

 removed, as already described. The filtrate was then heated almost to boiling and hydrogen 

 sulphide passed in for an hour. The slight precipitate formed, mostly sulphur, was filtered 

 off, washed with hydrogen-sulphide water, and dissolved by digesting with concentrated 

 nitric acid. The excess of acid was expelled, and the residue taken up in a solution con- 

 taining 0.5 c.c. each of nitric and sulphuric acids in 25 c.c. of water. The solution was 



