424 
Journal of Agricultural Research 
Vol. XXIII, No. 6 
repeatedly from No. i). A very similar strain was received from the 
New York Museum of Natural History as Azotobacter chroococcum (col¬ 
lection No. 522). Another subculture received from the same source 
under the same number in 1915 had not shown endospore formation 
but grew very similarly to our dwarfed Azotobacter strains. It died 
out after a while, and therefore we can not decide whether the sporu- 
lating growth later received from New York was a legitimate offspring 
or a contamination. On account of the fact that the analogous change 
occurred repeatedly under our eyes, the former possibility seems to us 
more probable. The New York strain showed all features of B. circulans 
Jordan (77, p. 5x9), and our own culture (No. 2) was fairly similar ex¬ 
cept that it was more inclined to become a large sporulating rod like 
those which we still have to discuss. 
The small sporulating rods of Azotobacter agile , while resembling very 
closely the smallest type developed from A. chroococcum and A. 
Beijerinckii , did not show any tendency to pass over to the Bacillus 
pumilus type, and this is again in good agreement with the recorded 
absence of the white-pink fungoid growth in the cultures of A. agile . 
In cell form and colony type these sporulating rods first remained very 
close to those of the small spore-free rods of A. agile (that is, Bacterium 
putidum) ; later however they did not exhibit any sharp difference from 
the Bacillus terminalis-jusiformis type. They, too, could be trans¬ 
formed into large sporulating rods. 
7 . LARGE SPORULATING RODS 
Here again Azotobacter chroococcum , A. Beijerinckii , and A. agile pro¬ 
duced very similar growths. The strains derived from the first-named 
species proved to be identical, morphologically as well as culturally, with 
Bacilluspetasites A. M. et Gottheil (12), while those obtained from A. agile 
exhibited more the character of B. silvaticus A. M. et Neide (37), but these 
differences are rather inconspicuous. An old stock culture of B. petasites 
received from New York was somewhat reduced in size, but otherwise quite 
typical. Like all strains developed from our Azotobacter cultures, it grew 
on beef agar either white, yellow, or brown. These three types of growth 
are not constant on account of the instability of the cells causing the 
different pigmentation. The typical, very large, broad, granulated rods 
produce the characteristic yellow growth; the weakly staining oval cysts, 
which were discussed above, make a whitish grayish layer; and in the 
brownish material, which also may give a brown color to the agar, sporu¬ 
lating long rods and threads of more or less regular shape are most fre¬ 
quent (fig. 73, 74, 76, and 77 on Plate 7). Temporarily, of course, one 
or the other mode of growth may predominate, and in short-termed tests 
such cultures are liable to be erroneously classified as different “species.” 
Form, color, and structure of the colonies, although very characteristic in 
their typical development, may vary considerably, too. Occasionally 
colonies were seen which macroscopically as well as microscopically (low 
magnification) were indistinguishable from those of the large nonsporu- 
lating cells of A. chroococcum; they contained nothing but spores. 
Illustrated descriptions, of two large, sporulating, nitrogen-fixing or¬ 
ganisms {Bacillus malabarensis , and B. danicus) } which were published 
by the senior author some years ago (27, 29, jo) can now be accepted as 
descriptions of the large sporulating cell type of Azotobacter. A re¬ 
newed thorough study of our stock cultures left no doubt that they may 
