300 THE STRUCTURE OF RHIZOB1UM LEGUMINOSARUM, 



A similar distortion occurs in figs. 4, 5, 6 and 12. Most of the 

 other cells figured in the plate show the method of formation of 

 the bacteroids; the multiplication of the spherules within the 

 capsule causes a pressure which is laterally applied, probably 

 through the division of the spherule at right angles to the 

 direction of the rod. Without the pressure, the dividing spherule 

 would arrange itself along the capsule. In this particular set. of 

 Bean race films the condensed forms of the spherules were rarely 

 seen, but one with a lateral outgrowth is shown in fig. 17. 

 When the cells are overstained and decolourised with alcohol, the 

 denser crescentic portions of the spherules alone retain the stain, 

 and forms like figs. 18, 19 are seen. 



Although a maltose medium was used for growing the slimes 

 in these two cases, other sugars gave similar results. With 

 saccharose there was rather more condensation of the spherules, 

 and especially of the terminal units. These were at some 

 distance from the central spherules, and gave the appearance of 

 terminal buds, which is so common a feature of the cells grown in 

 or on ordinary media. 



The coccoid* structure of Rhizobium has been advanced by 

 other observers, and this could only be expected in view of the 

 frequent occurrence of the condensed terminal cocci. In a 

 critical review Vogelf mentions Paratore, who saw small cocci in 

 the plasma of the branched forms. That the bacteroids are 

 sporangia was brought forward by Brunchorst and Moeller, and 

 confirmed by Hiltner and by Hartleb. Siichting considered that 

 the proof of the sporangium nature of the bacteroids was not 

 convincing enough, as it was possible that the contents of the 

 sporangia were either artificial products or they were the Babes- 

 Ernst granules which are met with in other bacteria, e.g., Bac. 

 diphtherice. 



With regard to these granules, Marx and Woithe J consider 

 that they are the products of the maximum condensation and of 

 a typical localisation of the chromatin of the bacterial cell. 



* Maze (Ann. de l'Inst Past.xii.) figures cocco-bacteria. 

 tCent. fur Bakt. 2te.Abt. xv. 185. 

 J Cent, fur Bakt. lte.Abt. xxviii. 100. 



