16 



MASS. EXPERIMENT STATION BULLETIN 213. 



Tests of Fungicides for the Control of Wildfire. 



1 The high percentage of infection in this experiment is explained by the long intervals between 

 applications and the fact that the bed was watered every day and inoculated twice a week. 



Experiment I4. — In similar experiments at Windsor the beds were on soil which 

 had grown a heavily infected crop of tobacco in 1921. The beds were not artificially 

 inoculated as in the preceding experiments. The fungicides used were Sanders 

 Dust No. 1, Niagara 20-80 copper-lime dust, Dosch 15-85 copper-lime dust, orchard 

 brand Bordeaux lead and Bordeaux zinc. Seven applications were made at inter- 

 vals of three to five days. A natural infection developed on the untreated plot and 

 in one corner of a plot next to it. No other wildfire developed on the treated plots. 



Conclusions from the Experiments and Practical Applications. 



Frequency of Application. ^ The writers recommended in 1921 (1: 81) 

 that the fungicide be applied once a week. Later experiments indicate, 

 however, that this is not sufficient under the following conditions: — 



1. When the plants are watered very frequently. On some soils it is 

 necessary to water the beds heavily every day. Most of the fungicide is 

 washed off before the end of a week. This factor was tested in Experi- 



