244 



BACTERIA IN RELATION TO PLANT DISEASES. 



Subsequently I tried again with somewhat better results, using Pitfield's stain (fig. 133). 

 All the flagella appeared to be polar. I am therefore inclined to think that the description 

 of Bacillus nicotianae Uyeda may not have been drawn wholly from one organism. 



HONING'S SUMATRAN STUDIES. 



Since completion of the preceding account of the tobacco-diseases, which includes 

 notice of one paper by Honing (p. 224), 11 additional Dutch papers (1911 to 1913) dealing 

 with the Sumatran disease have been received from Dr. Honing. These papers add much 

 to our knowledge of that disease. They render it probable that it is identical with the 

 Japanese disease and that both are due to Bad. solanacearum, first described by the writer 

 from the United States, although some discrepancies remain to be explained. These 

 Dutch papers are summarized as follows : 



(I) THE CAUSE OF THE SLIME-DISEASE AND ATTEMPTS TO COMBAT IT II. 



If two bacteria agree culturally, in reaction to stains, to substrata, etc. (making due allowance 

 for variable culture influence), but do not attack the same plants, then we may assume that they are 

 two nearly related but not identical species. Uyeda was warranted in separating his tobacco bac- 

 teria from Bad. solanacearum only on the ground that the latter does not attack tobacco, a supposition 

 which must now be abandoned, as Smith himself has shown and as has been confirmed in Sumatra. 



Our experiments with different strains of the bacteria have shown a great difference in virulence. 



The following experiments (table 25) show the infectious nature of the bacteria from different 

 sources and the effect of a previous infection of the soil. Each of the six rows contained 100 plants, 

 of which the first 50 in each were inoculated by soaking the seedlings for a few minutes in water to 

 which a bouillon-culture had been added. 



Table 25. Honing s Field Results with Bacteria from Various Sources. 



Rows 4, 5, and 6 were on land infected about 4 months before, and row 3, which also showed 

 many cases in the control, was adjacent to the previously infected land; 2 of these 5 cultures were 

 more virulent than the others. 



This experiment was repeated in pots with the same variable result, as shown in table 26. 



Table 26. Honing s Pot Results with Bacteria from Various Sources. 



"Why the bacteria from tobacco a were not able to attack the seedlings I do not know. For 

 the bacteria out of Pluchea the ease is different, because proof of their identity must yet beobtained." 

 With the bacteria from tobaccos b, c, and Ageratum, the disease was visible in 5 or 6 days. The other 

 plants were entirely sound after 24 days. 



