Feb. io, 1923 
Life History of Azotobacter 
425 
assume the character of this organism, as was already indicated in 
those earlier papers. With regard to B. danicus it was said in 1908 that 
all cell forms characteristic of Azotobacter were seen except the sarcina 
form; with the help of our newly won knowledge this could now be easily 
developed from secondary colonies on beef agar. The curious beet¬ 
shaped large cells of B . malabarensis can now also be identified with the 
gonidangia of the sporulating Azotobacter strains, and that the original 
isolation produced in mannite solution large globular spore-free Azoto- 
bacter-like cells (^7) is also easily understood at present. 
Bacillus luteus Baker et Smith seems to be identical with B. petasites, 
according to the detailed description published by Garbowski (ji). The 
three types of growth (white, yellow, or brown), the inclination to pro¬ 
duce in secondary colonies yellow spore-free cells (coccoids), and the 
various cultural marks are the same in both cases. The ability to fix 
free nitrogen is, of course, variable; B. malabarensis and B. danicus were 
found to assimilate 1 to 2 mgm. per 100 cc. 1 per cent mannite soil 
extract. 
The potato cultures, which proved most helpful for reestablishing the 
spore-free typical Azotobacter growth, gave also the only opportunity to 
distinguish with some accuracy the sporulating strains developed from 
Azatobacter chroococcum and A . Beijerinckii and those of A. agile. For 
one or two weeks all of them displayed first the whitish slimy, later 
the yellow glossy growth characteristic of Bacillus petasites; but then the 
growth became distinctly brown when derived from A. chroococcum or 
A. Beijerinckii , while it turned gray with A. agile as with Bacillus sil - 
vaticus. However, a much better differentiation was secured if use was 
made of the conspicuous inclination of these potato cultures to give rise 
to other types of growth. The cultures developed from A. chroococcum 
and A. Beijerinckii were found inclined to resume the character of B. 
pumilus , or transferred to beef gelatine they even went back to the very 
characteristic white-pink fungoid growth, while those of A. agile returned 
on beef gelatine to the white, small nonsporulating nonliquefying rod 
closely related to Bacterium putidum. Still more convincing, although 
more difficult to attain, is of course the reestablishment of the typical 
nonsporulating large cells; continued culture on potato alternated with 
transplants to alkaline mannite-nitrate agar and solution gave compara¬ 
tively the most satisfactory results in this respect. 
That the large sporulating rods of theMegaterium type are much inclined 
to become smaller and to assume other cultural characters was often 
observed, and it is equally well known that stock cultures of the smaller 
sporulating rods not infrequently lose their ability to produce endospores 
and become similar to Bacterium coli and related forms. With the anae¬ 
robic nitrogen-fixing bacilli of the Amylobacter group these changes are 
even more frequent, and the ability of these organisms to produce 
aerobic nonsporulating coccoids brings them into closer contact with the 
aerobic nitrogen-fixing bacilli of the Azotobacter group. All these 
relations should be properly considered and more thoroughly studied. 
DISCUSSION OF RESULTS 
The observations recorded concerning the different types of vegetative 
growth and of reproduction of Azotobacter permit a much more complete 
and much more accurate characterization of this organism than has been 
given thus far to any other genus of bacteria. In fact, all form genera of 
