338 F. A. Gooch and P. E. Browning — Determination 



errors of both processes of treatment are reasonably small, (re- 

 sjDectively 0001 grm. — and 0*0004 grm. — in the average) 

 with what difference there is in favor of our treatment of the 

 residues. Our process is incomparably the more convenient 

 and rapid. "We do not doubt that Messrs. Friedheim and 

 Meyer could have obtained equally good results had they 

 worked with pure reagents. 



Messrs. Friedheim and Meyer disagree with us also as to the 

 degree of concentration of the liquid necessary to insure the 

 complete expulsion of iodine and as to the effect of the con- 

 centration upon bromides present — points to which we gave 

 particular attention in our former paper. We showed that, in 

 general, upon boiling a mixture of iodide with sulphuric acid 

 and the arseniate diluted with water, tlie amount of iodine re- 

 moved depended upon the proportion of the sulphuric acid to 

 the final volume of the whole liquid, it being plainly shown 

 that even after the liquid had lost the color of free iodine in 

 the process of concentration it was still possible to detect com- 

 bined iodine. Further concentration and, to some extent, 

 dilution and repetition of the concentration to the former 

 point tend to set free the residual iodine. In experimenting 

 upon the proportions of acid most convenient for the work we 

 found that a solution made up to contain 0*5 grm. of potassium 

 iodide, 2 gram, of dihydrogen potassium arseniate, and 20 cm 8 

 of the sulphuric acid mixture (1 : 1 by volume) retained when 

 boiled down to 40 cm 3 no determinable free iodine and but the 

 faintest trace of combined iodine, while at 35 cm 3 the liquid 

 was free from iodine in any form. Upon experimenting as to 

 the behavior of mixtures of chlorides and iodides we found 

 that 0'5 grm. of sodium chloride added to the mixture contain- 

 ing 0*5 grm. of the iodide occasioned at 40 cm 3 a maximum 

 loss of arsenious chloride equivalent to - 0004 grm. of the 

 oxide, or to 0"0008 grm. at 30 cm 3 ; and that the loss as a mat- 

 ter of course, is less as the amount of iodide present (and con- 

 sequentty the arsenious oxide formed) is diminished. We 

 found that 0*5 grm. of potassium bromide treated in similar 

 manner occasioned no loss of arsenious bromide but did induce 

 at 35 cm 3 reduction of the arsenic acid amounting to 0'0005 

 grm. of iodine. Smaller amounts of bromide produced propor- 

 tionately smaller effects, but concentration even a little below 

 35 cm 3 was likely to be productive of serious error. We fixed 

 upon 35 cm 3 therefore, as the ideal concentration for removing 

 iodine from unknown mixtures with chlorides and bromides, 

 but expressly stated that failure to concentrate below 40 cm 3 

 introduces no appreciable error, while great care should be taken 

 not to press the concentration below 35 cm 3 on account of the 



