[Vor. 7 
76 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
rate of revolution of round-bottomed culture flasks held in a hori- 
zontal position. A rotating device was constructed, which 
held a vertical wooden wheel, to the rim of which were attached, 
horizontally, four one-liter Ipuddéeh ind flasks. The wheel 
revolving at the rate of one revolution in five minutes kept the 
culture medium in a slow, uniform motion. 
Only one experiment was conducted, and, as pointed out 
above, it was lacking in the second improvement.  Carbo- 
hydrate only was determined at short intervals, and the only 
renewal of this material was in two of the flasks on the rotator. 
The results obtained were so consistently contrary to expectation 
that the data from even the one experiment is of some interest; 
that is, as measured by carbohydrate consumption, the cultures 
showed a declining instead of an increasing metabolic activity. 
The Azotobacter culture used was a later generation of the 
same strain used in previous work. The nutrient solution was 
prepared as follows: 
Solution A: 
Dipotassium phosphate ...........o.ooooooocooooooo.. 0.2 gm. 
ciii oos RS da 0.2 gm. 
Vds AAA Aa 20.0 gms. 
"e. “distilled Po nae OT EEA TNR ERAT ORO 500.0 cc. 
Solution B: 
Magnesium sulphate ............o00oo.ooooocooo ooo». 0.2 gm. 
Calejum eUlphstO.......-..... «ono enn 0.1 gm. 
ZEN A NEREREREREPEXREREXEERISS QUE .0 cc. 
Ferric chloride (10 per cent sol.)............c0ceaeee 3 drops 
Solutions A and B were mixed, and 100-ce. portions then 
placed in each of four one-liter round-bottomed flasks and in 
each of two two-liter Erlenmeyers, all flasks receiving in addi- 
tion 1 gm. calcium carbonate. The flasks were then plugged 
with cotton and capped with beakers in the usual way. After 
sterilization by the intermittent method, all except one round- 
bottomed flask were inoculated with a spiral of a suspension 
prepared from an agar slant, the round-bottomed flasks placed 
in position on the rotator, and the Erlenmeyers on a shelf near 
by. The whole experiment was set up in a warm room kept at 
28-30? C. by a thermostat. 
The experiment was started March 14, 1919. Two days later 
a distinct turbidity appeared in the inoculated flasks on the 
