Iodine in the Analysis of Alkalies and Acids. 459 



neck of the beaker was then closed by a little trap, made of 

 one of the halves of a double end calcium chloride drying 

 tube, to prevent appreciable loss by spattering. The beaker 

 was then placed over a low flame, and the contents boiled until 

 the last trace of the excess of iodine had volatilized from the 

 solution and the trap. The volume was carefully regulated 

 before a-nd during the boiling, being kept as small as possible, 

 usually amounting to about 100 cm3 at the start and 35 cm3 at the 

 close. In the case of barium hydroxide care had to be taken 

 to keep the dilution sufficient to prevent the separation of the 

 crystal Hue barium iodate, which is soluble with difficulty. To 

 steady the ebullition a little spiral of platinum was introduced 

 into the beaker. After the boiling had ceased, the colorless 

 solution, containing a neutral mixture of iodate and iodide, 

 was cooled in running water, and treated with 10 cm3 of dilute 

 (1 : 3) hydrochloric acid or (1 : 3) sulphuric acid. The liber- 

 ated iodine was titrated directly with sodium thiosulphate, in 

 the presence of 5 cm3 of starch emulsion. In the case of barium 

 hydroxide, the iodine was liberated with dilute (1 : 3) hydro- 

 chloric acid to save the inconvenience of working in the pres- 

 •ence of precipitated barium sulphate; with potassium hydrox- 

 ide, however, dilute (1 : 3) sulphuric acid was employed. In 

 view of a statement by Pickering* that titrations with sodium 

 thiosulphate in the presence of acid involve an error, a series 

 of blank analyses was made which showed conclusively that no 

 such error exists under the conditions which obtain in the 

 process under consideration. Care was also taken, as in a 

 former case, to guard against the possible presence of carbo- 

 nates or other impurities in the reagents employed. 



In Table III are given the results of a series of analyses of 

 barium hydroxide by the modified method just described. 

 They agree fairly well with those of Table I. 



The analyses of potassium hydroxide by the modified method 

 are given in Table IV, and are found to agree well with those 

 of table II. 



A gravimetric analysis of the barium hydroxide solution in 

 which the barium was weighed as the sulphate, gave as a result 

 0-1411 grin. Ba(OH) 2 for each 20 cm3 taken. An^analysis of the 

 same solution by the Groger process gave for the same volume 

 0-1420 grin. The result by the Phelps process, however, was 

 0-1398 grm., and by the modified process 0*1390 grm. That 

 the difference of 2 mg. between the results by the gravimetric 

 and the Groger processes on one hand, and the Phelps process 

 and its modification on the other, may be due to atmospheric 

 -carbon dioxide, has already been pointed out. A gravimetric 



* Jour, of the Locd. Chem. Soc, xxxvii, 134. 



