PAPERS ON GEOLOGY AND CHEMISTRY 331 



startling and so parts of this work have been repeated 

 by one of us confirming Spiel's results as closely as 

 could be expected considering the use of a different in- 

 duction coil as source of power. 



Very briefly the method used in the Purdue work has 

 been to pass dry, carbon dioxide free, metered air down 

 through that part of the discharge outside of the dielec- 

 tric and up inside around the inner electrode where the 

 most of the discharge occurs. The gas then passes into 

 an apparatus containing broken glass tubing and stand- 

 ard alkali on the top of which are placed various traps 

 to diminish the amount of entrapped liquid as much as 

 possible. A large number of runs has been made at dif- 

 ferent velocities and diiferent pressures. The results 

 show a maximum in each case at a velocity of about 6 

 liters per minute. Beyond that the curve yield drops 

 more or less rapidly. In the discharge there are at least 

 two reactions going on simultaneously, an oxidation of 

 some of the nitrogen and the reverse decomposition. The 

 rate of the second reaction depends upon the concentra- 

 tion of the oxidized nitrogen. By increasing the flow of 

 the gas this concentration is diminished and the amount 

 of the reverse reaction approaches zero. According to 

 this the greater the flow rate the greater the total fixation 

 should be. Of course practically this reaches the limit. 

 Here the limit is reached when the velocity of the gas 

 becomes so great that the absorption is incomplete. How- 

 ever there are some other points that have arisen in con- 

 nection with this maximum. From other reasons there 

 is a suggestion that there might be a true optimum point 

 at this maximum. An interesting side phenomenon which 

 tends to accompany the absorption is the formation of 

 fog above the absorption liquid. This fog forms about 

 the molecules of nitrogen pentoxide or about the addi- 

 tion compound noted by Hautefeuille and Chappius. The 

 fog reaches the greatest density at this velocity and until 

 a small Cottrell electrostatic precipitator was included 

 it escaped with resulting lowering of the observed yield. 

 The appearance of this fog varies with the velocity and 

 probably with the relative temperatures of the gas, liquid 



