69 



trees, and making" counts of scales and other observations of the 

 results. 



Secondary KEvSults of the Experiments. 



It was the principal object of these experiments to test the ef- 

 fects of rains on the two washes used, but other important results 

 appeared in the outcome besides those immediately aimed at. 

 Counts of dead and living- scales on the check trees not treated and 

 on the experimental trees before treatment, showed a surprising- 

 percentage of half g-rown scales already dead, the ratio of dead 

 young- to living- scales varying- on different trees and on different 

 parts of the same tree from twenty-one per cent, to sixty-nine per 

 cent This fact had already been observed in other localities where 

 our insecticide work was in progress, and had, indeed, been noticed 

 and reported as early as 1898 by another assistant of the office, Mr. 

 E. B. Forbes, engag-ed in distributing to infested trees in southern 

 Illinois the spores of a fungus parasite of the San Jose scale. 



This spontaneous death of many of the scales which mig-ht 

 have been expected to pass the winter alive, was apparently due in 

 g-reat measure and in both instances to a severe drouth of the pre- 

 ceding- year. Consistently with this explanation the dead scales 

 were most abundant on trees worst affected by the drouth, and on 

 parts of trees to which the flow of sap would naturally be least. 



Another observation of importance to the investig-ator was 

 made with reference to the action of the insecticide in loosening- 

 the scales of the insects killed by it. In most cases where the ap- 

 plication took fatal effect the scales were so far loosened from the 

 bark that they were easily rubbed off, and might be washed away 

 in larg-e numbers by an ordinary rain. Asa consequence, if counts 

 were made of dead and living- scales upon a tree before treatment, 

 and ag-ain after a treatment and after a heavy rain had fallen, the 

 ratio of living- to dead mig-ht be as g-reat in the latter case as in the 

 former. It will be seen that by overlooking- this circumstance an 

 investig-ator mig-ht easily be led to very erroneous conclusions as to 

 the effects of moisture on the insecticide. 



General Features of thr Experiments. 

 The actual effect of rains was experimentally ascertained by 

 heavily spraying- the trees with water at selected intervals after 

 treatment with the wash, and by making- careful counts of dead 

 and living- scales in each case and comparing the ratios so arrived 

 at with those found in the beginning-. The trees sprayed with 

 each mixture were treated exactly alike except as to the subsequent 

 application of water, and in this latter respect the different trees 



