Icdometric Determination of Nitrates. 119 



In the first attempts to apply this reaction to the quantita- 

 tive estimation of nitrates, 10 cm 3 of the manganous mixture, 

 the pure weighed nitrate, and an excesss of potassium iodide 

 were put in a tubulated retort fitted with a hollow ground stop- 

 per drawn out at both ends so as to serve for the introduction 

 of carbon dioxide evolved in a Kipp's generator charged with 

 marble and acid previously boiled. The neck of the retort 

 passed through a rubber stopper nearly to the bottom of a side- 

 neck Erlenmeyer flask, used as a receiver, in the mouth of 

 which the stopper fitted tightly. The side-neck of the receiver 

 was joined by a rubber connector with a bent glass tube pas- 

 sing through a rubber stopper and reaching nearly to the bot- 

 tom of a side-neck test tube into the mouth of which the stop- 

 per was fitted. 



The first receiver contained hydrogen sodium carbonate in 

 excess of the amount needed to neutralize the acid in the 

 retort as it should distil over, a considerable quantity of potas- 

 sium iodide (about 3 grm.) to aid in dissolving condensed 

 iodine, and arsenious oxide in known amount and in excess of 

 the quantity necessary to convert to hydriodic acid the free 

 iodine evolved. The second smaller receiver was partly filled 

 with a dilute solution of potassium iodide and hydrogen 

 sodium carbonate. The current of carbon dioxide was started 

 immediately upon introducing the contents of the retort, and 

 the air was safely removed before the darkening of color, 

 which begins to appear very soon, had spread through the 

 liquid. Heat was applied, and the evolution of nitric oxide and 

 later that of iodine began. The distillation was continued until 

 nearly all the liquid had passed over. Finally, the contents 

 of both receivers were united and titrated against deci normal 

 iodine. The excess of arsenious oxide remaining unoxidized 

 was taken as the measure of the iodine liberated and, accord- 

 ingly, of the nitrate decomposed, upon the presumption that 

 two molecules of the nitrate liberate ultimately six atoms of 

 iodine according to the equation 



2HN0 3 + 6H-I = 4H 2 + 2N O + 3l-I. 



The choice of the solution for the retention of the halogen 

 evolved was dictated by the consideration, in the first place, 

 that very little iodine could pass through the alkaline arsenite 

 to come into contact with the rubber stopper of the receiver 

 on the way to the second absorbing liquid, and, secondly, that 

 higher oxides of nitrogen reformed by the action of traces of 

 air possibly introduced with the carbon dioxide or imperfectly 

 removed by it could not liberate iodine from an iodide in 

 in alkaline solution. 



