680 Journal of Agricultural Research voi.vi.No.is 



start growing, or they grow out of it, piercing the cell wall. With all 

 small bacteria we have observed only the two latter types of growth. 

 Almquist (i), who made some similar observations, called these granules 

 "conidia." In our opinion the term "gonidia" — that is, seed — is prefer- 

 able, as these granules in every case act as organs of propagation and mul- 

 tiplication, whatever may be their special mode of growth. 



Type D is in most cases the dissolution product either of the large forms 

 (types C and M) or of the small cells (types E, F, and H), but it can also 

 be formed by typical spores of type L, by regenerative bodies (I) and by 

 gonidia. Its inclination to take the stain varies widely. If the cell 

 walls participate considerably in its formation, it is readily and deeply 

 stained. The same holds true when the gonidia are small and easily 

 stainable. The large unstainable gonidia, on the other hand, which are 

 frequently produced in type C, as well as in type H (see below), naturally 

 give a rather pale or entirely unstained D. The structure, too, varies 

 accordingly. Small cells, or small gonidia, cause a finely granulated, 

 somewhat "hairy" structure; especially in the case of small slender rods 

 like B. fluorescens, B. radiobacter, etc., the term "woolly" perhaps would 

 be applicable. Large gonidia, on the other hand, as well as spores, clearly 

 melt together when entering this stage of growth. Figures 7 to 12 (PI. B), 

 18 (PI. C) , and 19 (PI. D) illustrate the different possibilities. Like type C, 

 type D has been considered by some investigators — for example, by K. B. 

 Lehmann — as an occurrence indicating the death of the bacteria. 1 Usu- 

 ally, however, it has been passed as some uninteresting "slime" or "dirt." 

 As it is made up by a thorough mixing or melting of a frequently large 

 number of cells, spores, or gonidia, the term symplasm or symplastic stage 

 seems to be a correct and convenient name for this stage. 



Some time after the symplasm has been formed, very small granules, 

 regenerative units (0.2 to 0.3^), become visible. If the symplasm does 

 not take the stain, the appearance of these organized well-stained forms 

 inside the amorphous pale mass is very surprising (fig. 12 of PI. B). 

 Such a preparate indeed first turned our attention into this new direc- 

 tion. The regenerative units increase in size until they show the form 

 of type E, F, I, or even A or B (fig. 13 and 14 of PI. C). All these cells 

 are easily stained, their cell wails being usually comparatively thick. 

 At last, practically all the symplasm is reorganized, leaving sometimes 

 only very few pale small " flakes." 



Type E represents a miniature counterpart of types A and B. The 

 size of the cells varies between 0.3 and 1 p.. Only with the latter forms are 

 the thin and the thick walled cells clearly discernible. In some cases at 

 least, we were able to observe germinating arthrospores of this type. 

 If necessary, both subtypes may be indicated conveniently by appending 

 to the E a Greek letter a or /3, respectively. 



1 The absurd name "zoogloea," which means "animal slime," has been repeatedly applied to this product 

 of bacterial "autolysis." and the fact that the walls of the cells are dissolved has been considered as indicat- 

 ing the death of the content of the cells. 



