June 2,1923 
Nitrogen Fixation by Azotobacter 
765 
These results led to the conclusion that no satisfactory data as to 
changes in reaction could be secured in a medium so poorly buffered. 
The quantity of KH2PO4 was accordingly increased to 0.5 per cent. 
Higher concentrations of phosphate were sometimes found to be toxic. 
This increase in phosphate increased the buffer effect approximately 
ten times, the media containing 0.5 per cent requiring 0.05 cc. N/i 
NaOH to effect a change of o.i Ph in 100 cc. The buffer effect was 
still low, but an examination of experiments 16 to 20 will show that 
changes in Ph produced by the growth of the various cultures were 
practically eliminated. If acid or basic metabolic by-products are pro¬ 
duced under these experimental conditions the quantities are inap¬ 
preciable. 
Fred (j) recorded a change in Ph from 7.2 to 5.1. The media em¬ 
ployed by him, however, contained only 0.02 per cent phosphate, and 
the actual quantity of acid necessafy to produce the recorded change 
in Ph was probably very small. Stoklasa (7) recorded the produc¬ 
tion of as high as 3.3 cc. N/i acid per 100 cc. media. This quantity 
would have produced very marked changes in the Ph of our media. 
However, the purity of Stoklasa’s cultures has been questioned by 
Bonazzi (2), 
In Table IV are recorded in parallel columns the initial reaction, 
growth, final reaction, and milligrams of nitrogen fixed for four experi¬ 
ments. This table is included in order that these various factors may 
be compared one with the other without the necessity of examining 
several tables. In figure i the influence of the reaction on nitrogen 
fixation is shown graphically. 
