HOT SPRINGS, ARKANSAS. 19 



10 c. c. of oxalic acid run in, which destroyed the color, and the solution 

 titrated with the .standard permanganate to the appearance of a pink 

 color. From the total number of c. c. of permanganate used was sub- 

 tracted the number of c. c. equal to 10 c. c. of oxalic acid. The result 

 gives the number of c. c. of permanganate required for 200 c. c. of 

 water. 



Total solids. — Measured amounts of the water were evaporated to 

 dryness in weighed platinum dishes on the steam bath. The dishes 

 were dried for twelve hours at the temperature of boiling water, cooled 

 in the desiccator, and weighed. The increase in weight of the dish 

 gives the amount of solids present in the volume of water used. 



To determine chlorine, iron and aluminum, manganese, bromine, 

 iodine, arsenic, and boric acid large quantities of the water were evap- 

 orated to dryness after the addition of a small amount of sodium car- 

 bonate. The residue thus obtained was boiled with distilled water, 

 transferred to a filter and thoroughly washed with hot water. The 

 residue in the paper was dried and transferred to the dish in which the 

 evaporation was made, the paper burned and added, and the whole 

 kept for the determination of iron, aluminum, and manganese. The 

 filtrate was made to a definite volume and aliquot portions taken to 

 determine the constituents mentioned above other than iron, aluminum, 

 and manganese. 



Chlorine. — An aliquot portion from the above filitrate was treated 



with a few drops of phenolphthalein and t-x HKS0 4 added at the rate 



of a drop eveiy few seconds until the red color had entirely disap- 

 peared, thus showing that all of the carbonates had changed to bicar- 

 bonates. a A few drops of potassium chromate indicator were then 

 added and the chlorides titrated with a solution of silver nitrate each 

 c. c. of which would precipitate 1 milligram of chlorine. 



Iodine and bromine. — The qualitative tests for the presence of iodine 

 and bromine were very much the same as those used in Fresenius. 

 Another aliquot portion from the above filtrate was evaporated to dry- 

 ness on the steam bath. Two or 3 c. c. of water were added to dissolve 

 and soften up the residue and enough absolute alcohol added to bring 

 the percentage of alcohol down to about 90 per cent. This was boiled 

 and filtered and the treatment with 90 per cent alcohol repeated once 

 or twice. Two or 3 drops of sodium hydrate solution were added to 

 the filtrate and it was evaporated to dryness. The same process of 

 extracting with 90 per cent alcohol was repeated on the new residue 

 and the extract filtered off from the undissolved portion. A drop of 

 sodium hydrate was added to the filtrate and it was evaporated to dry- 

 ness. The residue was treated with a little distilled water, dilute sul- 

 phuric acid added to acid reaction, the liquid transferred to a test tube, 

 and a little carbon disulphide added. Three or 1 drops of potassium 

 nitrite solution were then added and the test tube shaken. The pres- 

 ence of iodine was shown by a pink color in the carbon bisulphide. 

 Chlorine water was then added until the pink color due to the iodine 

 had disappeared, then a little more chlorine water. 



The presence of bromine was shown by an orange color in the car- 

 bon bisulphide. 



In no case did a sample of spring water give nearly as distinct a reac- 



a See Cameron's paper in Amer. Chern. Journal, 23, 481, 1900, 



