16 NATUKAL CONTROL OF WHITE FLIES IN FLORIDA. 



Unexplained mortality: 



5 lowest percentages averaging 66.1 per cent, 15.7 per cent siu-vived. 



5 highest percentages averaging 76.1 per cent, 11.7 per centeurvived. 

 Fungous parasitism : 



5 lowest percentages averaging 10.2 per cent, 14.7 per cent survived. 



5 highest percentages averaging 20.5 per cent, 12.6 per cent survived. 



In the above summary the comparatively small difference between 

 the five highest and five lowest records in each case makes the results 

 less striking than are the results of the previous year. However, it 

 is noteworthy that a difference of 10 per cent in unexplained mor- 

 tality shows a corresponding difference of 4 per cent in the number 

 surviving, while a difference of 10 per cent in the fungous parasitism 

 shows a difference of 2.1 per cent in the number surviving. 



A fair estimate of the results produced by either unexplained 

 mortality or fungous diseases must include a consideration of the 

 increased degree of benefit from each if it had been the only factor 

 concerned with the mortality of the larvae and pupje on the leaves. 

 This point has been discussed elsewhere as to fungous diseases. 

 Assuming that, in the 10 groves considered in Table III, the 12.4 per 

 cent recorded as infected by parasitic fungi in 1908 and the 15.4 per 

 cent in 1909 were actually destroyed by the fungi,^ a large part of 

 those infected by fungi would have died fi'om unexplained causes if 

 the fungi had not been present. Eighty-eight of every one hundred 

 larvse and pupee were not infected by fungi in 1908 and 85 of every 

 100 were not infected in 1909. Of these 84.1 per cent (73.7/87.6) and 

 83.8 per cent (71.0/84.7), respectively, died from unexplained causes. 

 It must therefore be assumed that if no fungous parasites had been 

 present 84 and 83.8 per cent of the 12.4 and 15.4 per cent recorded 

 in Table III would have died from unexplained causes, giving a total 

 efficacy for unexplained mortality of 84.1 per cent and 83.9 per cent, 

 respectively, for the years 1908 and 1909. This efficacy, combined 

 with the effects of fungous diseases and overcrowding, did not result 

 in a condition of satisfactory control in the average gi'ove in 1908, 

 with an average of about 24 live pupse per leaf, nor in 1909, with the 

 average reduced to 1 1 live pupae per leaf. In each year there was a 

 satisfactory condition of control in two or three of the groves under 

 observation or a promise of such a condition the following season. 

 In the opinions of the authors the data here given, representing a 

 small selection of the large amount of similar data at hand, covering 

 all sections of the State of Florida, conclusively show that the fluctua- 

 tions from year to year in the proportion of white flies dying from 

 causes as yet unexplained are of first importance in the periodical 

 "cleaning up" of infested citrus groves. 



More attention should be given to a study of the cause or causes 

 contributing to the unexplained mortality herein discussed. Attempts 



lAs shown elsewhere, the brown fungus is known to infect dead as well as live insects. 



