July 31, 1916 ■ Life Cycles of the Bacteria 679 



direct outside connections between the larger and the smaller forms, as 

 indicated by the six arrows: I-E, E-B, B-G, and I-F, F-L, I-H; but 

 they are rather rare exceptions.' The symplastic stage D, on the other 

 hand, can be observed in all cultures. Even if a strain shows little or no 

 inclination to change from one subcycle into another, it regularly passes 

 through type D. By frequent successive transfers (made each morning 

 and evening), we tried to prevent this "breakdown." For some days 

 we were successful, but then this disintegration again took place. After 

 another couple of days the tendency to produce "normal" forms was 

 once more very pronounced, which in its turn was again supplanted by 

 the formation of type D. We have followed this rythmic alternation 

 for some weeks with B. azotobacter as well as with B. subtilis. 



It goes without saying that the arrangement and naming of the different 

 types is merely a matter of convenience. Several of them perhaps could 

 be split up into two or more. However, we found them so suitable 

 in their present form for our work that we do not see any necessity for 

 making an alteration. 



Type A represents the normal, well-known, large Azotobacter cells of 

 globular or oval form (usually 2 to 3^ broad and 3 to 5/x long). By 

 further stretching they pass over into type L (fig. 1 and 2 of Pi. A). 



Type B indicates the thick-walled, rather resistant " arthrospores," 

 regularly formed from type A. 1 When they germinate, either globular 

 or oval forms are liberated. Some cells of type B, however, are produced 

 occasionally by D, or still less frequently by a direct enlargement of small 

 cells of type E or E, which will be discussed later. In the latter case the 

 germ developing from B shows the character of type G. Probably 

 no B formed by A germinates in this manner. This fact, then, would 

 distinguish these morphologically identical forms. 



Type C comprises all large forms in the stage of granular decomposition 

 heretofore generally considered to be dying "involution" forms. If the 

 granules are very small, they are nearly always easily stained. The larger 

 ones, on the other hand, usually remain entirely unstained when treated 

 with aqueous solutions of anilin dyes. Owing to the degenerated condi- 

 tion of the cell wall, the form of the cells frequently becomes quite irregu- 

 lar, and the granular content may become partially or entirely free (fig. 

 7 to 10 of PI. B). Undoubtedly these granules are of different nature. 

 Some may be fat, glycogen, or other metabolic products. Most of them 

 however, are living entities, as is clearly shown by their further behavior, 

 if not by their motility. Sometimes these granules develop to full-sized 

 cells before being liberated. (See type J.) In this case they behave 

 exactly like the "gonidia" in iron bacteria, as described by Cohn (4). 

 However, in most cases they either leave the cell entirely before they 



1 We think it best to reserve the name "arthrospore" exclusively for those cases -where the whole cell 

 acquires the character of a spore. If only parts of the cell (either at its end or side) show such transforma- 

 tion, we call them "regenerative bodies" or "exospores," according to their special character. 



