412 
Journal of Agricultural Research voi. xxm, no. & 
and on all substrates used the dissolution took place after io to 20 days, 
and about a week later the regeneration became general. If old material 
is transferred, typical symplasm or regeneration of cells may be seen, of 
course, from the start. If the transfers are repeated at short intervals 
the characteristic change between cellular and amorphous stage may 
occur much more frequently. For instance, transfers were made from a 
young culture of normal spore-free Azotobacter cells on mannite-nitrate 
agar and repeated each morning and each evening for several weeks. 
For a few days normal development took place, the next 6 to 8 transfers 
gave only very little dewlike transparent slimy growth made up of typical 
symplasm, later normal development became again visible, and so on in 
regular alternation. The tubes which first showed only symplasm gave 
normal growth after 2 to 3 days' delay, while with transfers made from 
vegetative cells no lag was noticeable (23 } p. 189). Occasionally sym¬ 
plasm was also found in young colonies. On the other hand, cultures 
kept without change in mannite-nitrate solution for several years ex¬ 
hibited the regular alternation between symplastic and cellular stage 
quite clearly as long as the microscopic tests were continued. 
As was described in our preliminary paper ( 28 ), the newly formed 
symplasm may be either homogeneous and not stainable with aqueous 
dyes or of a more or less hairy structure and easily stainable. These 
differences may persist until the new cells are formed, or changes from 
one to the other type may occur. Amoeboid motility, recorded by sev¬ 
eral authors (25, p. 183 ), was never observed, but strong inner move¬ 
ments in the amorphous clumps could be easily seen in the hanging 
drop. Encapsulated symplasm, as was described by Lankester in 1876 
(23, p . 270), was only found in very few instances (in some, but not in 
all, milk cultures of the small sporulating rods). Figure 66 on Plate 6 
shows three such “macroplasts,” as they were called by the British 
author, intact but of relatively small size (most of them were twice as 
large as those photographed), and one which has liberated its sarcina-like 
mass of small globular cells, which on account of its thickness could not 
be properly focused in the picture. The streptococcus-like chains also 
visible are the next step in the development which led to a uniform 
coccoid growth of typical regenerative bodies. The regenerative units 
which first become visible in the symplasm may either gradually increase 
in size, as illustrated by figures 98 to 100 on Plate 9, or may agglomerate 
to full-sized cells (figs. 56 and 59 on PI. 5; fig. 84 on PI. 7; figs. 103 and 
106 on PI. 9). The regeneration of small sporulating rods is shown in 
figure 102; it should be compared with figure 15 on Plate C of our 
preliminary paper (28). Figure 104 illustrates the possibility of the 
newly-formed rods growing in a radiate arrangement, while figure 105 
indicates that in other cases the new cells may appear along and parallel 
to the edges of the lobes of the symplasm. If the lobes are torn apart, 
as easily happens in making a preparate of such growth, a picture results 
which very closely resembles Bacillus pediculatus A. Koch et Hosaeus, 
while the appearance of the intact symplasm of this kind is very similar 
to B. vermiformis Ward and practically identical with Newskia ramosa 
Famintzin {34, pp. 33-33 , PL I). There is no doubt that a thorough 
study of this type of bacterial development will release these species and 
the so-called genus Newskia from their isolated position in the system of 
bacteria. The bacteria visible at the edges are not the cause of the 
slime, as was assumed, but it is the (slimy) symplasm which produces 
