260 BACTERIA IN RELATION TO PLANT DISEASES. 



the benefit of those who still believe in the degeneration of the Deli tobacco, it is perhaps 

 of service to state that the first Deli planter, fifty years ago, lost a whole field through 

 "the death." At that time the tobacco was immune to the slime-disease even less than 

 now. From this we may conclude that the slime-disease bacteria were here earlier than the 

 culture of European tobacco. 



Always and everywhere goes hand in hand with extensive culture the multiplication of 

 parasites, animal as well as vegetable, as the result of the one-sided activity, and always 

 and everywhere it has cost much labor to keep a crop on the same level and still more to 

 improve it. Therefore, especially in recent years, men have begun to select examples 

 more resistant to disease, and have found these in forms which, in other respects, are unfit 

 for culture, so that they have tried through crossing, followed by systematic selection, to 

 produce the desired combination. 



In both ways, that is by selection and by crossing, the attempt must be made to produce 

 from the Deli tobacco a less susceptible race. The selection can begin immediately; for 

 the crossing one must wait the discovery of immune varieties, which must not be far re- 

 moved from the Deli tobacco in quality. And because a variety, which elsewhere is little 

 or not at all susceptible, may be extremely sensitive when subjected to the different soils and 

 climatic conditions in Deli, one must not rely upon the literature, but all varieties must 

 be investigated in Deli in regard to their susceptibility. 



This year tests have been made with 87 sorts or varieties, among which were good 

 varieties from Java, the United States, Cuba, Mexico, Brazil, Paraguay, Santo Domingo, 

 the Philippines, Japan, and Hawaii. In a later report these experiments will be described 

 in detail. The following are the results in brief: 



(1) All of the 87 sorts or varieties investigated were susceptible to the slime-disease. 



(2) The 7 strains of Deli tobacco, which were inoculated at the same time as the foreign 

 tobaccos, did not make a bad showing. They grew better after the inoculation in the stem 

 than a great number of foreign sorts. 



(3) Two sorts of Philippine tobacco and one Japanese variety, although diseased, 

 gave the impression of being less severely attacked, so that it is worth while to repeat the 

 experiment on a larger scale with these three varieties another year. 



(4) The disease progressed more rapidly when the inoculation was made with a great 

 number of bacteria than with a few. 



This last-mentioned fact affords perhaps the explanation of the remarkable phenomena 

 which occurred in the disinfection-tests in 191 1 . In the different experiments the results of 

 the disinfection were at first gratifying, but in the end, when the plants were examined, it 

 appeared that not any or only a very slight advantage had been gained. But in the mean- 

 time in one of the experiments on a trial plot which had been kept under longer and closer 

 observation there were on each plant five to eight leaves more ("scrubs" and basal leaves). 

 Thus disinfection was insufficient, but the important lessening of the number of bacteria 

 had been advantageous. 



Now it appears that the rapidity with which the disease progresses in a plant is depen- 

 dent on the number of invading bacteria [see pi. 31, middle and right figs.]. This affords 

 perhaps the explanation of the above-described phenomena in the disinfection-tests of 191 1. 



The chance of finding among foreign tobaccos one which is immune in Deli appears to 

 be not very great. However, the fact that among the 1 19 examined varieties or sorts not 

 a single resistant one was present does not mean that none exists. The investigation, which 

 fortunately does not consume an undue amount of time, must be made with still other sorts. 

 Only, one must not raise too high his expectations of finding among the known varieties one 

 which in Sumatra is immune to Bacillus solanacearum. And this is surely an additional 

 reason for immediately subjecting our Deli tobacco itself to a severely conducted selection. 



There are various reasons which make it more than probable that the Deli tobacco 

 is not a clearly circumscribed type, but that it consists of a mixture of several types how 



