366 



The third series was made December 15, from 4 to 5 p.m., including 

 experiments 10 to 14; and on December IG, from 11 a. m. to 2 p. m., 

 in(;lnding experiments 15 to 30. A light rain of ten or fifteen minutes' 

 duration occurred on the evening of December 1(5, about 9.30 p. m., 

 with two or three days thereafter without rain. 



The final series of experiments was made January 23, 1895, from 4 to 

 6 p. m., including experiments 31 to 42; and on January 24, from 10 to 

 12 a.m., including experiments 43 to 40. On January 25 a light rain 

 with sleet occurred in the evening, and on the 28th there was a heas^y 

 fall of snow. 



Varying resulU on different parin of the same tree. — Later examina- 

 tion of the trees treated exhibited a marked difl'erence in the results 

 of many of the washes on different parts of the trees. It was found 

 that where tlie scales were most thickly massed on the older and lower 

 parts of the trunk the wash had been most efficient, and frequently in 

 these situations scarcely a living scale could be found, while on the 

 extremities of the branches, where the scales were scattering, the per- 

 centage of living scale insects greatly increased. This is evidently 

 due to the fact that the smooth terminals, especially those of peach, do 

 not retain much of the wash, and lose it more readily under the action 

 of rains; also that the scales in these situations are better nourished 

 and perhaps more vigorous than where more thickly clustered. The 

 dense incrustation of the scales on the bark j^roduces also a roughness 

 which holds the emulsion, and the emulsion naturally accumulates on 

 the basal portion of the branches. With many of the more satisfactory 

 experiments a certain patchy result was noted. In other words, on 

 particular branches or sides of branches every scale would be destroyed, 

 while elsewhere a considerable percentage of living scales would be 

 found, sometimes confined to small areas. This would seem to indicate 

 either a failure to drench the tree evenly, or else that the patchiness 

 was due to the action of subsequent rains and snows. The former is not 

 likely, because especial care was taken to thoroughly wet the trees, 

 and careful examination was made after each spraying to see that 

 this had been accomplished. The latter theory is probably the correct 

 one, and it is supported by the fact that the experiments of December 

 15 and 10 were followed by a light rain of a few minutes' duration from 

 the northwest, while those of January 23 and 24 were affected by the 

 irregular melting of subsequent snows, which lodged to a considerable 

 extent in the trees. 



The facts noted oft'er an additional reason for making the treatment 

 early in the fall, viz, that the wash may act over as long a period as 

 possible. That treatment is best made as soon as the leaves fall has 

 already been suggested by Mr. Howard, on the ground that the insects 

 at this time are more susceptible than later in the fall. (Insect Lipe, 

 vol. VII, p. 295.) A third advantage which would accrue in the East is, 

 that at this season rains are much less frequent than they are in mid or 

 late winter or early spring. 



