TOBACCO WILDFIRE. 73 



diseased bits of leaves were placed. IVIost of the diseased pieces were en- 

 tirely consumed and all of them more or less eaten. The insects were 

 then transferred to cages in which healthy plants were growing. Although 

 numerous holes were eaten through the leaves, no wildfire lesions de- 

 veloped. The flea beetle experiments variously modified were repeated 

 five times, but always with negative results. Wolf and Foster (5) had 

 the same results with thrips. Other insects have not been found on 

 tobacco in sufficient numbers to indicate that they could be the agents of 

 dissemination. From all evidence which has been secured up to the 

 present time we maj^ conclude that insects are of little or no importance 

 in the spread of the disease. This conclusion is also corroborated by a 

 study of field conditions. If insects were responsible, one would expect 

 to find scattered throughout the field plants which had only one or a few 

 infections on the upper leaves, and spread from a single infected plant 

 would be more rapid. But such is not the case. As ■wdll be explained 

 below, most of the infections can be traced from the lower leaves, and 

 can be readily explained bj^ splashing and wind-borne rain. 



Origin of the Disease in the Seed-beds. 

 Wildfire makes its first appearance in the seed-beds when the plants are 

 yet very small. It is very essential that we should know the source of the 

 bacteria which start this initial infection. This involves the whole problem 

 of the overwintering of the bacteria. The possibilities are that the bacteria 

 pass the winter on (1) the seed, (2) the chaff with the seed, (3) in the soil, 

 (4) in the sash, frames, covers, etc. Experiments are now in progress to 

 solve this problem by determining the longevity of the bacteria in various 

 en\aronments, but since these experiments have not as yet reached their 

 conclusion, nothing can be stated as demonstrated. Fromme (2) in Vir- 

 ginia believes that the bacteria overwinter with the seed. The fact that 

 infections are not positively kno"mi to start in the field from soil which is 

 splashed onto the leaves leads one to suspect that the soil is not the 

 source of early infection in Massachusetts. In North Carolina, however. 

 Wolf and Moss (6) found that soil or the cloth covers which had been used 

 the previous season could serve as carriers in overwintering the bacteria. 

 We expect to have more definite infomiation in regard to this point before 

 next season. When once started in the bed, the bacteria are easily spread 

 bj^ splashing during watering, contact of plants, weeding, etc. 



Source of Infection in the Field. 

 Careful field observations have been made during the past season on 

 the origin of the field infections and the spread in the field. In the vast 

 majority of cases it has been found that the plants came from infested 

 seed-beds, and in all other cases the seed-bed has been suspected, but it 

 has not been possible always to prove that it w^as the source. Almost 

 always, throughout the season, when a diseased plant was found the spots 



