no 



BACTERIA IN RKlvATlON TO PIvANT DISEASES. 



this solution they are transferred to the dye which is prepared by mixing an alcoholic solution of 

 aniline blue with orseillin, drop by drop, until a violet solution is obtained. The mixture is acidu- 

 lated with a few drops of glacial acetic acid. The sections remain in the stain for two hours and are 

 then transferred to dilute glycerin and finally mounted in glycerin. By this method the rodlets 

 were plainly differentiated. 



Where swellings on the filaments occur these rodlets are very numerous and finally the tube 

 bursts and the rodlets are liberated into the cell-cavity. The bursting of the filaments, or tubes as 

 Miss Dawson calls them, is a normal phenomenon. A transverse section of the nodule showed a 

 filament crossing the cell-wall, the figure given resembles an ordinary sieve-plate, but the relation 

 of the bacteria to the plate is rather obscure. She thinks the rodlets actually pierce the wall, absorb- 

 ing only the middle lamella. Further confirmation of the general results was obtained by staining 

 with methyl violet and fuchsin, though the former method was the more successful. 



Fig. 36.* 



In some cases the filaments were in close contact with the nucleus but she did not find this 

 relation constant. She says: 



"In sections of older tubercles the thicker filaments crossing the cortex are no longer to be seen 

 but those in the main tissue of the tubercles persist until decay has set in." 



She says further : 



"The tube, therefore, is actually formed by the parasite as it grows down the hair, and does not 

 arise from the plasma of the host plant." 



A variety of opinions exist as to the presence and constitution of a membrane bounding the 

 filaments. This author maintains that her results confirm Marshall Ward's claim that a membrane 

 is present, but she failed to detect in it cellulose or chitin. The presence of mucilage she considers 

 doubtful. She used Wisselingh's method for the detection of chitin. This is as follows : Sections of 

 alcholic material are heated in concentrated potash to i6o° C. for two hours. After cooling they are 



*FiG. 36.— Longitudinal section (9) of root-nodule of Pisum sativum stained with aniline blue and orseillin, 

 showing rodlets of Bact. leguminosarum within filaments. Longitudinal section (10) of root-nodule of Pisum sativum 

 stained with methyl violet and fuchsin, showing liberation of rodlets from filaments. After Maria Dawson. 



