1824.] the Aeriform Compounds of Nitrogen. 347 



the constitution of ammonia, taken by M. Gay-Lussac, repre- 

 sents it as consisting of 1 volume of nitrogen and 3 volumes 

 of hydrogen condensed into the space of 2 volumes. 



In order to satisfy myself on a point, the determination of 

 which is so essential to a just view of the atomic constitution 

 of the compounds of nitrogen, I have lately made fresh expe- 

 riments on the decomposition of ammonia by electricity, using 

 every precaution that occurred to me as likely to insure the- 

 accuracy of the results. The gas was collected over recently 

 boiled and dry mercury, and was transferred for decomposition 

 into graduated tubes, filled with mercury, which had been 

 heated in the same tubes and still remained hot. To prevent 

 any ammoniacal gas from lodging beneath the surface of the 

 quicksilver in the tube, the flame of a spirit lamp was passed 

 slowly along the part containing mercury, a precaution which 

 was shown not to have been unnecessary by the ascent of a 

 few bubbles of gas. 



In four experiments, conducted with a degree of caution, to 

 which I am not aware that any thing could have been added,, 

 the volume of the ammoniacal gas was fully doubled. In the 

 first, 44 measures became 88 + ; in the second, 157 became 

 320 ; in the third, 60 became 122 ; and in the fourth, 120 be- 

 came 240. The evolved gases, carefully analyzed by combus- 

 tion with oxygen, were found in each case to consist of 1 vo- 

 lume of nitrogen and 3 volumes of hydrogen. I repeated, also, 

 with the greatest attention, a process for analyzing ammonia, 

 which, with various other methods capable of being more 

 quickly executed than that of electrical analysis, I have de- 

 scribed hi the Philosophical Transactions for 1809. It consists 

 in firing, by the electric spark, a mixture of the alkaline gas- 

 with nitrous oxide, the latter being employed in rather less 

 proportion than would be necessary for perfect decomposition, 

 in order to prevent the formation of nitrous acid vapour, which 

 is always generated when the nitrous oxide is in excess. For 

 example, 10 measures of ammonia were deflagrated with 12 or 

 13 of nitrous oxide, the full proportion of the latter being, if 

 pure, 15 measures. All the oxygen of the nitrous oxide was 

 transferred to the hydrogen of the ammonia, water was formed, 

 and the whole nitrogen of both gases remained as the aeriform 

 product, mixed with a small quantity of hydrogen gas, for the 

 combustion of which the nitrous oxide had not supplied suffi- 

 cient oxygen. This quantity of hydrogen being too small to 

 form a combustible mixture, it was necessary to make an addi- 

 tion of that gas, and to employ, for the second combustion, 

 more oxygen than was requisite to saturate the hydrogen added. 

 The quantity of hydrogen, originally in the mixture, was thus 

 easily determined, and, when added to the volume of pure ni- 



