190 Gooch and Browning — Method for the 



tated. Kerosene was chosen in preference to other solvents of 

 iodine on account of its lightness, which makes it float npon 

 the mixture, and its high boiling-point, which permits the ap- 

 plication of heat to hasten and complete the reaction. Its dis- 

 advantage is the persistency with which it adheres to the walls 

 of the test-tube, so that washing with alcohol (or other solvent) 

 after the completion of each test is necessary to prevent the 

 transfer of the iodine of one test to the test next succeeding. 

 The data of these experiments are indicated in the headings. 



It will be noted that in Series A, in which the absolute 

 amount of iodine employed, its proportion to the entire volume, 

 and the amount of the arsenic salt remained the same, the pro- 

 portion of sulphuric acid being the variable element, it is 

 shown that the proportion of sulphuric acid should reach at least 

 twelve parts by volume in one hundred of the solution in order 

 that the maximum distinctness of the test may be developed. 

 An excess of sulphuric acid beyond this proportion is not dis- 

 advantageous. 



In Series B the proportion and absolute amount of iodine 

 vary as well as the proportion of acid, while the quantity of 

 arsenic remains invariable. The results of this series confirm 

 those of the previous series as to the proper proportion of sul- 

 phuric acid to be used, and the sensitiveness of the test is 

 shown to reach (in round numbers) one part by weight of 

 iodine -in six hundred thousand parts of the solution. 



The tests of Series C indicate plainly that it is the sulphuric 

 acid which is the potent agent in liberating the iodine, the experi- 

 ment in which acetic acid was substituted for sulphuric acid 

 being particularly noteworthy in this connection. The pres- 

 ence of arsenic acid increases the sensitiveness of the reaction, 

 but its addition beyond a very moderate amount does not ap- 

 pear to be necessary or advantageous. The presence of a 

 chloride or bromide does not impair the delicacy of the test. 



The quantities of iodine taken in the experiments just de- 

 scribed were necessarily small, and the question arises naturally 

 as to whether the course of action would be similar in the 

 presence of larger amounts of that substance and the corre- 

 spondingly greater amount of arsenious oxide which is pro- 

 duced with its tendency to reverse the reaction according to 

 which the elimination of iodine proceeds. The solution of this 

 question was reached in the following experiments : 



To 50 cm 3 of liquid containing 10 cm 3 of sulphuric acid 

 [1:1], and 1 cm 3 of a decinormal solution of iodine in potassium 

 iodide (0*001265 grm. of the former in 0*0018 grm. of the lat- 

 ter) was added 1 cm 3 of a decinormal solution of arsenious 

 oxide (0*00495 grm.), an amount ten times as much as would 

 be necessary to convert the iodine into hydriodic acid were 



