40 PLANT RESPIRATION 



After the absorption of the COo the gas is passed into the 

 measuring tube and again measured. The difference between 

 the two readings gives the volume of carbon dioxide. After 

 this measurement the gas is forced back into the analysis tube. 

 An analysis tube filled with pure hydrogen is transferred to the 

 bath, placed on the end of the measuring tube and the neces- 

 sary amount of hydrogen drawn into the measuring tube, the 

 amount of hydrogen being computed approximately according 

 to the probable oxygen content of the gas to be analysed. If 

 the quantity of carbon dioxide be less than 8%, the amount of 

 hydrogen used for analysis must amount to about ^g of the 

 original volume of the gas to be analysed. The exact volume 

 of hydrogen is measured wdth the aid of the reading microscope 

 and the hydrogen is mixed in the analysis tube with the gas to 

 be analysed. The contents of the analysis tube is now trans- 

 ferred to the explosion pipette. The explosion is fired off by 

 setting the zinc-carbon element for a spark. With this the 

 bulb G is lowered as far as possible and the stopcock p is 

 opened at the same time, whereby a stream of mercury is pro- 

 duced from the bath C into the explosion pipette. After the 

 explosion the gas is at once forced back into the analysis tube 

 by raising the bulb G. Then the analysis tube is transferred to 

 the end of the measuring tube, the gas drawn into the measuring 

 tube and its volume accurately measured. If a were the volume 

 of the gas after the absorption of COo, b, the volume of hydrogen, 

 and c, the volume of the gas mixture after the explosion, the 



, ^, . a -\- b — c ^. . . ,.,. 



volume of the oxygen is if the data obtamed tor 



CQo and oxygen is then expressed in percentages, the nitrogen 

 is computed according to the difference. If the oxygen content 

 of the gas to be analyzed is very small, it can easily happen that 

 no explosion takes place when the spark is passed. In this 

 case some illuminating gas is added and the explosion repeated. 

 The volume of the illuminating gas must be about }^ that of the 

 gas to be analysed. With the use of an excess of illuminating 

 gas, part of the nitrogen may be oxidised as a result of too 

 violent an explosion. Of course this introduces a considerable 

 error in the analysis. 



