BEHAVIOR OF CHECK-PLANTS. 



diseased, then the experiments must be done over with more care and times enough 

 to remove all possible chance of error. When check-plants become diseased, 

 especially in any number, there is always room for grave suspicion. Either the 

 experimenter has been grossly careless, assuming that he used the right organism in 

 his inoculation-experiment, or else he is working in a locality where the cause of 

 the disease is naturally abundant. In either case, however well convinced he him- 

 self may be, his readers will generally have a lingering suspicion that even his inocu- 

 lated plants succumbed not to what he inserted into them, but to some entirely differ- 

 ent cause naturally present and overlooked by the investigator. The remedy for the 



first is to learn to use infectious 

 material with more caution, and 

 for the second is to make the in- 

 oculation-experiments in localities 

 or under conditions where the 

 plant shall be less subject to natu- 

 ral infection. 



If the experiments must be per- 

 formed in localities where the dis- 

 ease is naturally present, then a 

 large number of plants must be 

 selected for inoculation and for 

 control, and such a high percent- 

 age of infections secured in the 

 inoculated plants that the few 

 cases occurring naturally in the 

 control-plants may be neglected 

 as not casting any doubt on the 

 general result. For example, if, 

 in a region subject to the given 

 disease, too plants were reserved 

 for control and 100 similar plants 

 were inoculated, and out of this 

 number 50 of the latter and 40 

 of the former should contract the 



Fig. 9.* 



disease, it is manifest that no deductions of any value could be made from the 

 experiment. All might be the result of some cause totally different from the 



*Fic. 9. Cross-section of a s<ma'll part of a cucumber stem, showing the parasitism of Bacillus 

 tracheiphilns in one of the inner bundles. As yet there is no bacterial cavity, the bacilli being con- 

 fined -to .the spiral vessels and a very few of the adjacent pitted vessels. .Material taken from a field 

 near Washington, D. C., in 1893. Sectioned from paraffin. Drawn from a photomicrograph. 

 X 50. Introduced for comparison with plate 3. Beginning at the top, the tissues occur in the fol- 

 lowing order: (i) Outer phloem, showing sieve plates; (2) cambium; (3) immature xylem; 

 (4) mature xylem, consisting of pitted vessels and pitted connective tissues; (5) spiral vessels em- 

 bedded in non-lignified living parenchyma, which is finally destroyed by the bacteria ; (6) pseudo- 

 cambial layer; (7) inner phloem ; (8) large-celled parenchyma to either .side, separating this bundle 

 from its neighbors. 



