eo 
ate application when diluted. I consequently, last winter, instructed 
Mr. E. P. Taylor, the assistant in charge of orchard inspection and 
insecticide operations, to carry on a series of experiments in a care- 
fully chosen orchard, intended to test a number of these insecticides 
under identical conditions and in a way to enable us to give the re- 
sults in the exact form of percentages of benefit for each. 
For this purpose, control was obtained of two small orchards 
near Richview, 111., both belonging to Mr. Edward Tucker, one con- 
taining 780 apple- and peach-trees badly infested in the main, and the 
other, 680 peach-trees infested throughout but much less heavily so. 
These orchards were divided into plots of varying size, those in- 
tended for the more important insecticides containing from 47 to 141 
trees each. Minor experiments were made on smaller lots, of from 
3 to 33 trees, and single trees here and there were also used for a 
few preliminary tests. 
Insecticides were applied, partly early in January, from the 3d 
to the loth, and partly from the 21st to the 24th of March, 1905. 
Twenty-seven kinds and forms of insecticides were used in all, of 
which seventeen were various compounds of sulphur and lime, six 
were kerosene preparations, three were soap solutions, and one was a 
simple alkali. The results were tested by three critical examinations, 
one made January 3 and 4, a second during the last days of May, and 
the last September 5 to 8. It is of course impossible to get an accu- 
rate idea of the effects of the treatment by comparing the ratio of 
living to dead at the time of the treatment with the corresponding 
ratio at a considerable interval thereafter, unless the scales remain 
attached to the tree after death ; and that they gradually disappear, 
especially after a lime and sulphur treatment, was shown by observa- 
tions reported by me in 1902.* 
System of Inspection. 
The choice of a system of inspection by which the condition of 
trees treated with various insecticides could be compared with their 
condition before treatment, and with that of other trees not treated 
but reserved as checks, was a difficult and perplexing matter. At 
first I was disposed to depend mainly on counts of dead and living 
scales from various parts of carefully selected trees, from which 
average ratios of dead to living, before and after treatment, might 
be computed, but such comparisons rapidly become misleading as 
the scales killed by the spray disappear under exposure to the 
weather. Comparisons of the results of treatment with different in- 
secticides can not be reliable if based on counts of living and dead, 
♦Bull. 111. Agr. Exper. Station, No. 71, p. 243. 
