Jan., 1921] 
WANN — FIXATION OF FREE NITROGEN 
7 
sembled, the delivery tubes at the end of each rack were connected in a 
single series by means of T-tubes so that aeration of all ten series could be 
accomplished by one operation. Each of these delivery tubes was provided 
with a screw clamp which was kept tightly closed except during the process 
of aeration, when it served to control the volume of air passing through the 
series. With a few exceptions aeration was continued for an hour every 
morning, it being considered that this would entirely replace the air in the 
apparatus. The 191 7-1 8 experiment as completely assembled is shown in 
figure I, Plate I. 
At the end of the first experiment, titrations of the contents of the acid 
tubes showed that no appreciable change had taken place in the concentra- 
tion in any instance. It was assumed, therefore, that with these species 
and with the conditions realized in the experiment there was no loss of 
ammonia from the culture flasks. For this reason the tubes of acid were 
omitted from the second experiment, and in the process of aeration the air 
passed directly from one culture flask to the next in the series. Because of 
the expansion of the air in the culture flasks during the hot summer days, 
the liquids in the gas-washing bottles were frequently forced out through 
the intake tubes of these bottles, — making it necessary to place safety 
flasks outside the acid bottle, and between the acid and water bottles, to re- 
ceive those liquids. During the process of eration the acid and water 
were drawn back into the proper bottles so that no air ever entered the 
series without first passing through the liquids. Aeration was continued 
for an hour every other morning during the growing period. 
Cultural Conditions. — Soon after the inoculation of the culture flasks 
the ten racks were transferred to the greenhouse where more uniform condi- 
tions of light and temperature prevailed than in the laboratory. Since 
preliminary tests showed that agar cultures of the organisms used were 
soon killed by direct sunlight, the bench occupied by the apparatus was 
covered with a canopy of black cloth, which reduced the actinic light 
intensity to about one eighth that of the normal greenhouse illumination. 
On cloudy days, however, this canopy was rolled up on both sides, thus 
permitting better illumination ; on clear days the west side was open during 
the morning only, and during the afternoon the east side only was exposed. 
The arrangement of the apparatus as finally assembled in the greenhouse 
is shown in figure 2, Plate I, which is a photograph of the 191 9 experiment. 
Growth of the Cultures 
Length of Growing Period. — The approximate length, in days, of the 
growing period of each series is indicated in the headings of the tables which 
follow. Inoculations in the first experiment were completed on August 31, 
191 7, and the analyses were begun in April, 191 8. The final inoculations of 
the second experiment were made in May, 191 9, and analyses were started 
in November of the same year. 
