Feb. 10, 1923 
Life History of Azotobacter 
4 i 7 
very unstable; they either grow up to large cells or reproduce rods either 
by germination or by simple stretching. But if they are once definitely 
established (cultivation on beef or on potato agar proved best for this 
purpose) they may cling tenaciously to their now micrococcus-like habit 
of growth. White growth was most frequent (Table I, p. 415), yellow 
next, and pink much more rare. Changes from white to yellow and 
from yellow to pink were observed in some cases, in full agreement with 
the results recorded with Micrococci (25, p. 31, 43). As A. chroococcum 
as well as A. agile produce all three types of growth, it is easily under¬ 
stood why these differences in pigmentation can not be constant. 
The white coccoid growth is very similar to that ascribed to Micro¬ 
coccus concentricus Zimm. On agar smooth circular colonies are formed, 
which have a finely granular structure, resembling under the micro¬ 
scope very much those of the Bacterium pneumoniae Friedl. group, 
except that the radiate stripes are missing. On gelatine they are dis¬ 
tinctly zoned and have lobate edges. The growth on beef agar, as 
well as on mannite-nitrate agar, is varying from thin dew-like to thick 
aerogenes-like layers. Gelatine is not or only very little liquefied. 
Beef broth becomes slightly turbid and has no film or ring, but heavy 
sediment. Milk remains unchanged, brom-cresol-purple is slightly 
reduced; occasionally a slow peptonization takes place, which, however, 
was always accompanied by a change from the coccoid to the fungoid 
cell type. On potato a thick, creamy yellowish grayish brownish 
growth, similar to that of Bacterium aerogenes , was observed. 
The yellow coccoid growth differs from the white one practically only 
by the yellow pigment produced in the cells on agar, gelatine, potato, 
etc., and the pink growth behaves in an analogous manner, except that 
the growth on potato was always very scant and colorless. Gelatine 
liquefaction was always absent or very moderate; motility was never 
seen. The pink cocci gave sometimes a brown hue to gelatine as well 
as to potato agar. According to the customary diagnosis, these yellow 
and pink coccoids would have to take their places close to Micrococcus 
sulfureus Zimm. emend. Lehm. et Neum. and to Micrococcus roseus 
(Bumm) Lehm. et Neum., respectively. 
Nitrogen fixation was not noticeable with these coccoid strains. 
That, however, occasionally positive results may be obtained is proved 
by the fact that a culture, once determined as Micrococcus sulfureus 
var. tardigradus and found to be able to assimilate free nitrogen vigor¬ 
ously (27), has shown itself to have been the vegetative growth of the 
regenerative bodies of a yellow large sporulating rod, which stands very 
close to those developed from Azotobacter. 
3.— DWARFED GROWTH 
As this type of growth is the result of the vegetative propagation of 
the gonidia, it is always at the beginning, and not infrequently for a 
very long time, rather inconspicuous and easily overlooked. Very small 
whitish circular colonies of about ^-mm. diameter, of finely granular 
yellow structure in the center and of white, smooth appearance at the 
edge, are most typical. If the growth is more vigorous, a flaky struc¬ 
ture may become visible in the center, as with Bacterium septicaemiae 
haemorrhagicae Hueppe (i8 i Table 18 , VI). It often took two weeks 
before the colonies became visible. Transferred to agar, the growth is 
again frequently very thin and dewlike, and it may remain so on all 
