24 
AMERICAN JOURNAL OF BOTANY 
[Vol. 8 
experiments, under which marked fixation occurred, have merely amplified 
those already existing in the field, for soluble carbohydrates and a certain 
amount of nitrates would most certainly be available for the development 
of these organisms, especially those occurring naturally on soil. If this is 
actually the case, it is highly probable that fixation occurs also under natural 
conditions. 
The evidence that no fixation occurred on media lacking nitrates is 
quite conclusive. When nitrogen was supplied as glycocoll, asparagine, 
urea, or ammonium sulphate, the increase or decrease in the nitrogen 
content of the culture flasks above or below that of the checks was in all 
cases less than a milligram, even though the growth, especially on urea 
with glucose, was as luxuriant as that produced on nitrates in the presence 
of glucose. The results were uniform for all species grown on media con- 
taining these nitrogen sources and were not altered by the presence eithei 
of glucose or of mannite, the latter being used only with ammonium sulphate. 
Although some of the cultures of these series showed higher nitrogen con- 
tents than the checks, there were no consistent gains, and in most cases 
the figures were lower than those of the checks. However, the differences 
in either case are so small as to be without significance, since most of them 
are within the usual experimental error. 
A comparison of the results presented here with figures secured by a 
large number of investigators working with the legume nodule bacteria and 
Azotobacter forms shows that the amount of fixation produced by the algae 
per unit volume of medium, under the conditions of the experiments, equals, 
and in some cases surpasses, the average fixation by the colorless forms. 
The legume bacteria especially give only slight increases when grown 
in pure culture outside the host plant, Fred (1912) gives a table of the 
results of 25 investigators, from which it is apparent that the average 
fixation with this organism is about 0.9 to 2.2 mg. per 100 cc. culture 
medium. Fred himself secured fixations ranging from 0.4 to 1.58 mg. per 
100 cc. culture solution. With the Azotobacter forms the amounts of 
fixation are considerably larger. Azotobacter chroococcum, which has been 
extensively investigated in pure culture, shows increases ranging from a 
few milligrams to 20 or 30 mg. per 100 cc. culture medium, the average 
fixation, however, being usually about 6 or 10 mg. Thus Gerlach and 
Vogel (1902) report 4 mg. to be the average, while Freudenreich (1903) 
secured only 1.6 to 2.4 mg. per 100 cc. culture solution, but on gypsum 
plates the increases were higher, amounting to 16 mg. per 100 cc. solution. 
Krzemieniewski (1908) secured increases ranging from i to 13.5 mg. per 
100 cc. solution, an average of 3 to 6 mg, being fixed for each gram of 
glucose respired. In the presence of humus the ratio was 5 to 10 mg. 
nitrogen fixed per gram of glucose respired. Hoffmann and Hammer 
(1910) report fixations of from 4.55 to 14.4 mg. per gram of carbohydrate 
consumed, and in the same year Krzemieniewska (1910) pubHshed results 
