298 THE STRUCTURE OF RRIZOBIUM LEGITMINOSARUM, 



distributed. A large loop or a small drop of the suspension is 

 spread over a cover-glass. The film is allowed to dry in the 

 air, it is then passed through the flame and decolourised with 

 0-5 % acetic acid (0 - 5 c.c. of the glacial acid in 100 c.c. of water), 

 after which it is dried between sheets of blotting-paper and 

 imbedded in balsam. 



The method is simple, but the result depends largely 

 upon the time that the stain is in contact with the cells. Six 

 hours is generally correct. With less than this the contents of 

 the cells are not sufficiently stained, while with more they become 

 overstained and the structure cannot be seen. Stronger acetic 

 acid than that advised does not remove the excess of stain and 

 only produces a grey colour. Every culture will not give good 

 films; much appears to depend upon the amount of slime. It 

 was only from cultures giving over 8 % of slime that good films 

 were obtained. The cells then appear to be very diffuse or 

 swollen, and the contents are readily stained and observed. It 

 is not advisable to have too much of the culture in the dilute 

 stain as the slime forms a coagulum with the fuchsin. 



Instructive films were obtained from cultures of Robinia and 

 Bean races which produced over 8 % of slime upon plates of 

 maltose- citrate- ammonium sulphate -agar. Overstained cells 

 appeared to contain one or two spongy portions (fig.l) 5 but when 

 properly stained the rods were seen to consist of a number of 

 bipolar staining spherules or coccoid structures (fig. 7). Once 

 these bodies are recognised it is not an easy matter to find cells 

 which suggest any other form of structure. Here and there in 

 the films were deeply-staining free cocci (fig. 2;. These were 

 also frequently seen within rod-shaped cells or capsules (tig.3) r 

 sometimes singly and central, but generally paired and terminal. 

 There may be two or more (figs.3,4). The occurrence of the large 

 diffusely-staining spherules and the deeply-staining cocci in the 

 same capsule enables the latter to be recognised as a condensation 

 of the former, and this is made manifest by the diplococcoid 

 structure of the larger deeply-staining units (figs. 5, 6). The poles 

 of the large spherules are frequently absent and a uniform outline is 



