Iodine in the Analysis of Alkalies and Acids. 457 



ture, and cannot be pushed to completion except under condi- 

 tions that make its application in analysis impossible. In the 

 case of barium and potassium hydroxides both the original pro- 

 cedure of Phelps and the modification above described were 

 employed. The modified method was found to be the more 

 convenient and speedy of the two. The results obtained in 

 both cases agreed with one another, but were invariably lower 

 by a small nearly constant amount than those obtained by both 

 the gravimetric and the Groger processes. This error of the 

 Phelps process and its modification is possibly due to the 

 action of atmospheric carbon dioxide on the hydroxide solu- 

 tion during the short time it is exposed. While it will affect the 

 value of the method as a means of accurately determining the 

 absolute amount of hydroxide present in a given volume of solu- 

 tion, it cannot so affect the accuracy of any differential method 

 founded on the original Phelps process or its modification. 

 This is demonstrated by the work of Phelps in the case of 

 carbon-dioxide, and by the present investigation in the case of 

 hydrochloric and sulphuric acid. Analyses of these two acids 

 were made by adding the solution to be analyzed to a measured 

 volume of barium or potassium hydroxide, previously stand- 

 ardized by the modified Phelps method. The small excess 

 of hydroxide remaining was then estimated by the same 

 method, the results agreeing with those already obtained by 

 both the gravimetric and the Groger processes. It seems 

 probable that other acids and compounds for which there is 

 now no rapid iodometric method may be analyzed by a method 

 similar to this, that has given good results with carbonic, hydro- 

 chloric and sulphuric acids. 



Decinormal solutions of the alkali hydroxides were prepared, 

 and kept with great care in trapped bottles, from which por- 

 tions for analysis were measured by means of a self- feeding 

 burette, which was also fitted with a trap. All vessels and 

 water used were made as free as possible from carbon-dioxide, 

 and the operations were conducted as rapidly as possible. 



In the analyses by the Phelps method a carefully measured 

 excess of decinormal iodine was drawn into a small ether wash- 

 bottle, and the desired amount of alkali was rapidly run into 

 it. The stopper, to which had been sealed a Will and Varren- 

 trapp absorption bulb, was placed in the bottle and the bulb 

 was charged with a 5 per cent solution of potassium iodide to 

 catch any escaping vapors of iodine. The apparatus was 

 placed over a low flame and the contents heated to boiling or 

 slightly longer, and then cooled in a stream of water. The 

 contents of the bulb and connecting tubes were then washed 

 into the flask, and the excess of free iodine remaining was 

 titrated with arsenious acid, in the presence of 5 cm3 of starch 



