6 MASS. EXPERIMENT STATION BULLETIN 213. 



The above data furnish very strong evidence that the pathogen may 

 be carried from one plant to another or from one crop to another by means 

 of the soil. The failure to get infection in some of the experiments by 

 planting in infested soil shows, however, that infection will not always 

 result necessarily because the soil was infested. 



None of the experiments just quoted furnishes evidence of the length 

 of time during which the bacteria may remain aUve in the soil or indicates 

 whether they will Uve through the winter in this habitat. The following 

 experiments and observations throw some hght on the latter point : — 



Experiment 3. — On July 1, 1921, Erlenmeyer flasks of soil were sterilized and 

 later inoculated with the bacteria. Part were plugged only with cotton, others were 

 parafiined to prevent drying out. At various times during the winter, soil was 

 taken from these flasks and plated out. Then, when bacteria developed about the 

 particles of earth, they were shaken in a suspension of water and atomized on 

 healthy plants. In the flasks which did not have paraffined plugs, the soil became 

 very dry, while in the others it remained muddy. Heavy infection resulted when 

 inoculations were made March 10 and others on March 20, 1922, from the dry 

 flasks, but none from the tightly closed wet flasks. These flasks were kept in the 

 laboratory and were not frozen. In this case the bacteria were still able to produce 

 infection after eight months. 



In two instances in Connecticut, wildfire was found starting in the edge 

 of the beds in soil which had been outside the pans when the remainder of 

 the beds were steamed. In both cases wildfire was present in the beds in 

 1921. The fact that the planks were new and the sash had been sterHized 

 with formaldehyde eUminated these as the source of infection. 



In a number of cases, in both States, it was found that those parts of the 

 field which were diseased in 1921 showed the heaviest infection in 1922. 



On the other hand, fields have been observed which were badly diseased 

 in 1921 and on which tobacco was free from wildfire in 1922. 



On one of the fields at the Connecticut Experiment Station the 1921 

 crop which was badly infected with wildfire was cut late in September 

 and left lying on the ground over winter with a view to getting data on the 

 overwintering under natural conditions. In this case both leaves and 

 stalks were left to weather. In 1922 this field was planted with Havana 

 and Broadleaf wildfire-free seedhngs, the stalks and leaves of the 1921 

 crop having been disked and plowed under two weeks prior to setting. 

 Throughout the season close examinations were made by Slagg and Chap- 

 man for wildfire in this field. Wildfire was not found on this particular 

 field during the growing season, but at harvest an occasional wildfire 

 spot was found, yet nothing to what should have developed if any con- 

 siderable amount of direct infection occurred as a result of the refuse being 

 left on the field. A careful estimate of the wildfire plants on this plot, 

 made at harvesting, showed that infected plants were not more than one- 

 half of 1 per cent of the total number, and on all of these the infection 

 was light. This sUght infection may have come from plants in the wildfire 

 experimental field, since all the station plots — except for the experimental 

 field — showed about tliis same percentage of infection late in the season. 



