Section 79. Errors in the Solutions and their Correction. 203 



This difficulty can be partly removed, of course, by pumping the air 

 out as completely as possible. This causes no appreciable loss of ammo- 

 nia, since its partial pressure above a 0.1 normal solution at 18 is only 

 1.34 mm ;* hence the ammonia present in 20 c.cm. of the vapor above the 

 solution is only 0.01 per cent of that in the solution. But some oxygen 

 still remains dissolved in the solution and in the platinum, and causes 

 oxidation of the ammonia. Hence the most feasible plan was to pump out 

 the air before the heating till the pressure became 2 or 3 cm., and to 

 determine after the heating the strength of the solution, by titrating the 

 solution left in the bomb. 



There is some error in the titration of an ammonia solution so dilute as 

 0.01 nonnal, as the end point is not sufficiently sharp. The method used 

 was to adopt a standard color, add an excess of acid to a portion of the 

 solution, and titrate to the standard color with the residue of the solution. 

 The percentage error of the titration was at the same time determined by 

 titrating similarly a portion of the unheated solution, whose concentra- 

 tion was known. Solutions 6.1 and 6.2 were titrated with phenacetolin, 

 the correction for the titration-error being -J- 1.0 per cent ; the other dilute 

 solutions were titrated with Congo red, for which correction was 0.3 

 per cent. 



In the first experiments with ammonium acetate, the conductance at 

 18 was also found to have decreased from one to two per cent as a result 

 of the heating at 218. Experiments with acetic acid showed that this was 

 not alone due to the oxidation of ammonia, but that the acetic acid itself 

 had decreased both in conductance and concentration after being heated 

 to 218. This effect was not sufficiently marked to be taken account of in 

 the work of Noyes and Cooper, probably because in their small tempera- 

 ture-bath the solution could be heated to constant temperature much more 

 quickly, and also because they used unplatinized electrodes, platinum black 

 being a catalyzer of the decomposition, according to the work of Sabatier 

 and Senderens.f In the heatings to 306 this effect was found to become 

 greater with increased concentration of the ammonium acetate solution. 

 It seemed possible that it might be due to the formation of acetamide at 

 the high temperature and the continued existence of this in the solution 

 at 18 owing to the rapid cooling. If this were the case, it should be pos- 

 sible to reconvert it to ammonium acetate by prolonged heating at about 

 100. Two hours heating of one of the solutions at 110 - 120, however, 

 produced only a slight decrease, instead of an increase, in the conductance 

 at 18. The existence of acetamide in the solutions at the high tempera- 

 tures would give rise to an error in the calculated hydrolysis. That it does 



*Locke and Forsell, Am. Chem. J., 31, 268 (1904). 

 fAnn. chim. phys. (8), 4, 319, 433. 



