Feb. io, 1923 
Life History of Azotobacter 
411 
52 to 54 on Plate 5, as well as figure 83 on Plate 7, represent the charac- 
acteristic picture of cultures inclined to begin or to resume endosporulation. 
Heating of such strains first to 70° C., later to 75°-90° C., provided that the 
tendency to produce such terminal bodies could be stabilized, leads even¬ 
tually to the production of typical polar endospores or exospores. Nega¬ 
tive results are frequent in such experiments, but they do not disprove 
the less numerous positive ones. As was pointed out before, these newly 
evolved small sporulating strains often grow very slowly and weakly, a 
fact which also deserves full attention. When completely developed, the 
endospores of the small-cell type withstood 95 0 C. for4 to 5 minutes, 98° C. 
for 1 minute, and those of the large rods 98° C. for 2 to 4 minutes; the 
germination was mostly polar in both cases. Budding exospores were 
most clearly visible with the large sporulating rods, as was illustrated by 
figure 20 on Plate D of our preliminary paper (28). The spores do not 
always germinate directly; sometimes they first swell up to fairly large 
weakly staining ovals which are indistinguishable from microcysts. On 
the other hand it is no rare occurrence, especially when the small sporu¬ 
lating rods are cultivated in broth or milk, that the normally produced 
spores do not germinate but reproduce, either by budding or by dividing, 
2 to 4 gonidia or regenerative bodies, which again may multiply as such 
(fig. 70 on PI. 6). As was pointed out in Part I {2 5, p . 141 ), similar observa- 
ions were made before; they demonstrate anew that the formation of 
gonidia is in fact the fundamental mode of bacterial reproduction. 
The transformation of vegetative cells to microcysts is most conspicu¬ 
ous with the large-cell types. The large sporulating rods produce weakly 
staining ovals (sometimes discarded as “shadows’') which germinate 
readily to darkly staining, short, broad, mostly pointed rods (fig. 74 on 
PI. 7). With the large spore-free cell type the microcysts are equally 
common; here they are the well-known thick-walled cells of globular, 
oval, or pointed shape, frequently united to two or four, whose germina¬ 
tion was carefully studied by Prazmowski (40) and others. That the 
germ may be of coccoid or of threadlike shape is illustrated by figures 93 
and 94 on Plate 8. With the small-cell types microcysts are less con¬ 
spicuous. The coccoids are either thin-walled or thick-walled; in the 
latter case they may act as microcysts. With the dwarfed growth small 
pale ovals are not infrequent, perhaps a counterpart to those of the large 
sporulating cells. The large microcysts of the spore-free as well as of the 
sporulating cell type were found to be able to become gonidangia, as was 
to be expected. 
3.— Formation of the sympeasm and the regeneration of ceees 
In figures 7 to 18 on Plates B and C of our preliminary paper (28) the 
gradual dissolution of Azotobacter cells and the regeneration of various 
types of cells was first illustrated. Other pictures were published in 
Part I of these Studies (25, Pis. XVIII to XXI), here together with 
analogous findings of other authors. On Plate 9 of the present paper 
some additional photographs are reproduced, all pertaining to the sym- 
plastic stage of Azotobacter. Figures 21 (PI. 2), 33 to 36 (PI. 3), 38, 47, 
and 48 (PI. 4), 56, 59, and 60 (PI. 5), 66, 71, and 72 (PI. 6), and 84 (PI. 7) 
are also of interest in this connection. 
All growth types of Azotobacter, vegetative as well as reproductive 
cells, were seen to enter the symplastic stage regularly. If vigorously 
growing vegetative cells were transferred, practically without exception 
25622—23 - 2 
