99- 



•0024 



100-5 



•0043 



99-2 



•0150 



99*4 



•0020 



100-8 



•00355 



98-9 



•0008 



Roberts — Reduction of Nitric Acid by Fevroiis Salts. 133 



tween the Will and Yarrentrapp tube and the Hempel burette 

 to prevent the possibility of any iodine being carried over 

 mechanically into the gas receiver. In all cases the Will and 

 Yarrentrapp tube was filled with potassium iodide, and the 

 amount of iodine liberated estimated by titration with sodium 

 thiosulphate. 



ill. 

 Per cent reckoned Amount of 



from total vol. of gas. iodine in grams. 



(i) 



(2) 



(3) 



W 



(5) 



(6) 



In the last five experiments of this series, the caustic soda 

 solution was thoroughly boiled to eliminate any error which 

 might arise from oxidation of the nitric oxide by air dissolved 

 in the liquid, but with no marked difference in the results. 



In all except the last one of these experiments, the nitrate 

 and ferrous chloride were brought together in a more or less 

 hot solution instead of being heated gradually, and although 

 the amounts of iodine vary, it will be seen that they are on 

 the whole greater in this series than in the experiments in 

 which the reacting bodies were brought together in the cold. 

 This would indicate the possibility of some slight local action 

 in the hot solution by which higher oxides of nitrogen are 

 generated, but any analytical error from this source is corrected 

 by the use of the potassium iodide which converts them into 

 nitric oxide. 



In experiment (3) of this series, where a very large amount 

 of iodine is liberated, only one half as much ferrous chloride 

 was used in proportion to the amount of nitrate as was taken 

 in the other experiments, so that the amount of iron present 

 must have been very near the theoretical limit. Here the 

 amount of iodine would indicate the formation of a higher 

 oxide of nitrogen, an error corrected however by the use of 

 potassium iodide so that the percentage as calculated from the 

 total volume of gas compares well with the results obtained 

 in the other experiments. In experiment (6) also, a rather 

 small excess of the FeCl 2 was taken (-255 grms. Fe being 

 weighed out, whereas -16ti is theoretically required), but in 

 this case the liquids were almost cold when brought together, 

 and the amount of iodine set free is very small. 



In several cases experiments were made to test the solubility 

 of the nitric oxide in caustic soda. In experiment (2) the 

 gas was allowed to stand over the liquid, with frequent shak- 

 ing, for about an hour when the volume had decreased from 



