conspectus: action on the plant 



43 



mitted to the seedling, whose leaf-serratures, infected probably 

 through their water-pores (Fig. 29), in turn become the chief 

 focus of the bacterial multiplication. Apparently the bacteria 

 are always present, and we do not know what would happen to 



Fig. 31. In,, 32. 



Fig. 31. — Cross-section of leaf of Pavetta angustijolia showing a veinlet and a 

 small bacterial nodule. The veinlet has a closed cylinder of xylem. The bacterial 

 pockets in the young nodule are drawn in solid black. At c.c. are masses of coUen- 

 chyma. It looks as if the organisms entered from the upper surface very early 

 through the palisade tissue. Drawn November 21, 1914, from an unstained sec- 

 tion lying in water. 



Fig. 32. — Structure of one of the leafknots shown in Fig. 30^4. Stained with 

 anilin blue. 



Ardisia plants grown without them, nor do we know how to 

 obtain such plants (1915). It would be an interesting experi- 

 ment to see if they could be produced without the bacteria and 

 to watch their behavior. 



