Observations were made with tobacco plants which, after being naturally infected in the field by the virus of "vira-cabeça", a disease identical to spotted wilt (3), not only resisted to it, but exhibited complete recovery. Several trials were carried out in order to ascertain the following points : 1) whether the recovery means a greater degree of individual constitutional resistance ; 2) whether it represents an acquired immunity or ; 3) whether it is just of pure accidental nature. Plants showing recovery were selfed and seeds collected from them sown in separate rows. The same procedure was always applied with the new recovered plants. Thus, strains of plants were obtained representing several succeeding generations of recovered plants. In 1940-41 recovery progenies were planted at two different transplanting periods. During the next season 1941-42 these progenies, which represented already one more generation in comparison with those of 1940-41, were planted in two districts where the average occurrence of "vira-cabeça" was different. The percentages of "vira-cabeça", observed in all trials, were submitted to statistical analysis, comparing the errors obtained for difference of districts, transplanting time, varieties and progenies representing several generations of recovered plants. Comparing the calculated errors it was found that the greatest effect in the occurrence of ' Vira-cabeça" observed should be attributed to different transplanting times or different districts. The selection of recovered plants showed no influence in lowering the incidence of * vira-cabeça" : hence the impossibility of getting types of tobacco resistant to "vira-cabeça" through selection of recovered plants. Plants were observed in the field that after the first recovery exhibited symptoms of the disease, and then completely recovered from it again. This is in accordance with the results of Smith (11) in the reinoculation of tobacco plants after recovery from spotted wilt and of Costa (4) who reinoculated "vira-cabeça" - recovered tobacco plants with it. Therefore, as far as our knowledge is concerned districts should be selected where the disease does not appear very frequently; transplanting periods should be chosen, tc guarantee the lowest possible incidence of the disease
