E. J. Russell and E. H. Richards 537 



any absorption or evolution of gas could thus be measured. As oxygen 

 was absorbed in quantity it was essential that a pure supply should 

 be available, free not only from sulphur dioxide, carbon dioxide, etc., 

 but also from traces of nitrogen : this was ensured by sealing on to the 

 apparatus an electrolytic vessel G, charged with baryta, by the electro- 

 lysis of which very pure oxygen was obtained. The volume of oxygen 

 was determined by measuring the volume of hydrogen in the nitro- 

 meter H. 



The manure having been put into the flask A, and the last joints 

 blown, the pump was started working and sufficient air was extracted 

 to allow the auxiliary mercury lifting pump to come into action. The 

 system was then completely closed, the circulation commenced, and the 

 pressure on the manometer read. 



The volume of the apparatus now had to be determined. This 

 was done by letting in a known volume of oxygen from the electrolytic 

 vessel and measuring the drop in pressure thus caused. 



Knowing the volume of the apparatus and the percentage composi- 

 tion of the air contained therein, it was easy to calculate the volume of 

 nitrogen initially present. 



The circulation was then kept up for several hours a day, the mercury 

 being lifted by the auxiliary water pump so that the operation was 

 automatic. The acid in the one set of bulbs caught any trace of ammonia 

 carried over, while the alkali in the other set took out the COg: the 

 movements of the manometer showed the net change in pressure. 

 Fresh oxygen was periodically admitted from the electrolytic vessel. 



An occasional sample of gas was pumped out, measured and analysed 

 and the manometer was read. Thus the composition of the air inside 

 the apparatus could be known whenever desired. 



When the experiment had continued long enough the gas was 

 pumped out, measured and analysed. This gave the volume of nitrogen 

 finally present. , 



Lastly the manure was collected from the flask, weighed and analysed, 

 and the acid from the bulbs was also distilled to collect the ammonia, 

 which was added to that in the manure. Thus we know the change 

 in amount of nitrogen in the manure, and by a totally different process 

 the change in volume of the nitrogen present. In our two experiments 

 these two distinct methods gave concordant results. 



In the first experiment the changes in the manure (bullock manure) 

 were: 



