132 Roberts — Reduction of Nitric Acid by Ferrous /Salts. 



other experiments, but, as would naturally be expected, the 

 volume of gas incapable of absorption by potassium perman- 

 ganate is much increased. 



An interesting continuation of the supposition that small 

 quantities of air introduced into nitric oxide confined over 

 caustic soda would not influence the total volume of gas was 

 found in some experiments performed later. After the nitric 

 oxide in the Hem pel burette had been accurately measured, 

 air was introduced through the upper stop cock, a cubic centi- 

 meter at a time, and the burette lightly shaken. It was found 

 that more than twenty cubic centimeters of air could be intro- 

 duced in this way without materially affecting the total volume 

 of gas, but that the volume began to increase when about one 

 cubic centimeter of nitric oxide was left unoxidized, as subse- 

 quently determined by potassium permanganate. 



In the fifth experiment of Table II, supposing the 2 -43 

 cubic centimeters of unabsorbed gas to represent nitrogen left 

 from air which had oxidized the nitric oxide to nitrous acid 

 2*5- cubic centimeters of nitric oxide would thus have been 

 made to disappear, and the amount of nitric oxide left should 

 be 89 per cent, 2 per cent higher than that obtained. On the 

 same supposition, in the four preceding experiments, the *6 

 cubic centimeters of unabsorbed gas represent a loss of '6 

 cubic centimeters of nitric oxide which corresponds to a loss 

 of about 3 per cent on the nitrate, so that the results reckoned 

 from the absorbed gas should be about 97 per cent instead of 

 95 per cent as found. Here, again, is an unexplained loss of 

 about 2 per cent. Moreover, with the use of potassium iodide, 

 it is plain that, in whatever way these higher oxides of nitrogen 

 may be formed, they should be entirely broken up into nitric 

 oxide, so that the amount reckoned from the absorbed gas 

 should be 100 per cent, whatever the total volume may be. 

 We find instead that we get an average of 97'6 per cent, a loss 

 of a little more than two per cent. It seemed probable, then, 

 that as these experiments were done, there was a loss of from 

 2 to 2*5 percent of nitric oxide by solution in 'caustic soda. 

 This loss would mean the solution of *5 to *6 cubic centimeters 

 of the gas after being thoroughly shaken with about 75 cubic 

 centimeters of strong caustic soda, an amount of solubility 

 which did not seem at all improbable and which would be 

 quite in accordance with the results of Warington's experi- 

 ments, previously mentioned. 



The following set of experiments was therefore undertaken 

 with the double object of studying the solubility of nitric oxide 

 in caustic soda, and noting any conditions of temperature or 

 proportions which might influence the course of the reaction. 

 The same apparatus was used as in the preceding experiments 

 except that a Geissler tube filled with water was inserted be- 



