Feb. io, 1923 
Life History of Azotobacter 
403 
mentioned in Part I (25, p. 29, 34) and others will be discussed on the 
following pages. On the other hand, the instability of a new form may 
prove vexatious. Even if a new type of growth is represented by a 
sufficient number of cells in a culture, and its separation by plating, there¬ 
fore appears to be promising, nevertheless a complete failure may result, 
due either to an immediate return to the original type of growth or to 
the disinclination of the new form to grow in pure culture on the substrate 
used. Occasionally the stabilization of the new form proceeds very 
slowly and gradually; its colonies may appear very late on the plates, 
or they may at first simulate those of the mother form so closely that only 
microscopically can a differentiation be made. However, these particular 
difficulties are rather an exception to the rule; usually the new type of 
growth will be isolated if it is represented in the original culture by a 
sufficient number of cells. To reduce the chances for contamination as 
far as possible we prefer for plating flat bottles with a neck not much 
wider than a test tube. The dilutions are made directly in the bottles 
wherein the substrate was previously sterilized. 
Besides beef agar, beef gelatine, beef broth, milk, and potato the sub¬ 
strates most frequently used in our Azotobacter experiments were man- 
nite-nitrate solution and mannite-nitrate agar. The composition of 
these media was given in our preliminary communication ( 28 , p. 686). 
Our recipe was changed by E. R. Allen (2) with very unsatisfactory 
results; it is, however, only his substrate, not, as he asserts, the ‘‘medium 
of Iybhnis and Smith'* which proved to be unsuitable for Azotobacter 
growth. We studied for example, the alternation between the varied 
cell life and the symplastic stage of this organism in the original cultures, 
made in mannite-nitrate solution, for more than a year and then kept 
these cultures four years longer (most of the time sealed); transferred to 
new solution at the end of the whole period, 60 per cent of the cultures 
gave new vigorous growth. The hydrogen-ion concentration of these 
substrates was usually kept approximately at Ph 6 .8; occasionally it 
was increased to 6.0 or lowered to 7.5-8.0. Potato agar (15 gm. agar 
and 10 gm. peptone added to 1,000 cc. filtered potato water, prepared 
by boiling 200 gm. potato in 1,200 cc. water, to which an excess of CaC 0 2 
was added) was also found to be very useful, as was the case, although 
less frequently, with soda agar (beef agar with 34 P er cent NajCO,), 
phosphate agar (beef agar with pi per cent Ca(H 2 P 0 4 ) 2 ), salt agar 
(beef agar with 2 to 8 per cent NaCl), and water to which 5 cc. beef 
broth per 1,000 cc. were added. Most instructive results, however, 
were secured from cultures kept in sterilized soil (about 5 gm. in test 
tubes heated in the autoclave for 1% hour at 20 pounds pressure) mois¬ 
tened with 0.5 per cent mannite solution. Within a few weeks all 
developmental stages were passed in this substrate with greatest regu¬ 
larity. 
Some experiments upon the effect of symbiosis of Azotobacter and 
Radiobacter were also included. It was pointed out in an earlier paper 
(26) that especially the regeneration of large spore-free Azotobacter cells 
from the sporulating growth is favored by the presence of Radiobacter. 
The following cultures of Azotobacter were used in our experiments: 
A. chroococcum Beij. (Laboratory Numbers 1, 2, 10, 11, 12, 14, 17 to 
25)- 
A. Beijerinckii J. G. Lipman (Laboratory Numbers 3 to6,13, and 15). 
A . agile Beij., syn. A . Vinelandii J. G. Lipman (Laboratory Numbers 
7 to 7c, 16 to 16c). 
