Brown — Work on the Biter action of Hydrochloric Acid. 51 



and the Yoit flask connected with a Drexel washing-bottle 

 containing about 300 cm3 of an approximately two per cent 

 solution of potassium iodide, the Drexel washing-bottle being 

 joined to Will and Varrentrapp bnlbs also containing a strong 

 solution of potassium iodide to absorb escaping iodine fumes. 

 The chlorine retained in solution in the flask was driven into 

 the potassium iodide by means of a strong current of air dried 

 and purified in the usual manner, the liberated iodine titrated 

 by means of an approximately twentieth-normal solution of 

 sodium thiosulphate, and its equivalent in permanganate cal- 

 culated. In those experiments in which the digestion was 

 made with hydrochloric acid alone, and in the presence of 

 cadmium chloride and chromic chloride, the " oxidizing mate- 

 rial not chlorine " was determined by adding potassium iodide 

 to the digestion liquid after the removal of the chlorine, and 

 titrating the liberated iodine with sodium thiosulphate. Its 

 permanganate equivalent was then estimated as indicated in the 

 table. By adding together the " KMn0 4 equivalent " of the 

 chlorine retained in solution during digestion and the " KMn0 4 

 equivalent" of the " oxidizing material not CI formed during 

 digestion " we have the " total KMn0 4 found." The difference 

 between this last value and the "KMn0 4 before digestion" 

 gives the u KMn0 4 lost" as indicated in Table V. Since, 

 however, potassium iodide could not be added to the solution 

 containing ferric and auric chlorides because of the reduction 

 of these salts with liberation of iodine, an action probably 

 indefinite under the conditions of experimentation, the average 

 of the permanganate equivalent of the " oxidizing material 

 not CI" obtained with hydrochloric acid alone and with cad- 

 mium chloride was taken as indicating the same value with 

 ferric and auric chloride. The results of Table II have shown 

 that the residual oxides of manganese are substantially con- 

 stant with hydrochloric acid alone and with cadmium, ferric, 

 and auric chlorides. So no error is introduced by the pro- 

 cedure here adopted. Calculations of " KMn0 4 found " and 

 "KMn0 4 lost" were then made as with hydrochloric alone, 

 and with cadmium chloride. The results follow : 



