17 
the number of hills of corn infested by them, we find an average of 
66 ants and 23 root-lice to the hill in the checks, 9 ants and 2 root- 
lice to a hill in the oil of lemon plot, 16 ants and 3 root-lice in the 
kerosene plot, 30 ants and 6 root-lice in the formalin plot, and 25 ants 
and 12 root-lice to the hill in the plot treated with carbolic acid. The 
ratios of benefit in respect to the numbers of root-lice were conse- 
quently as follows. With respect to ants : oil of lemon, 86 percent 
benefit; kerosene, 76 percent; formalin, 55 percent; and' carbolic acid, 
62 percent. With respect to root-lice : oil of lemon, 83 percent benefit ; 
kerosene, 87 percent ; formalin, 74 percent ; and carbolic acid, 52 per- 
cent. The use of the three secondary applications, on the other hand, 
"Rex Dip," chlorid of lime, and sulphur, had evidently accomplished 
comparatively little. In the plot treated with chlorid of lime there 
were 31 ants and 25 aphids to the hill; in that treated with sulphur, 
45 ants and 31 aphids; and in the plot treated with "Rex Dip," 72 
ants and 53 aphids — the last two of these numbers being actually 
greater than those for the checks. 
The total normal infestation of this field at this time, as shown by 
the condition of the checks, was 240,000 ants and 124,000 root-lice to 
the acre. The aphids were all in hills of corn, while only so many 
of the ants were there' established as was necessary to the purposes 
of the colony, the central home of which was commonly outside the 
hill between the_ rows of corn. The data of the foregoing discussion 
are briefly presented in the tables on page 20. 
RESULTS OF THE EXPERIMENT. SECOND INSPECTION 
June 21, 43 days after planting, a second count was made of both 
ants and aphids in 250 hills^ — 50 hills each for the check and for the 
four principal experimental plots. At this second counting, IZ percent 
of the hills in the check were infested with ants, 22 percent of those 
in the oil of lemon plot, 44 percent in the kerosene plot, 46 percent 
in the formalin plot, and 76 percent in the plot treated with carbolic 
acid. A few hills contained ants which did not contain root-lice also. 
The corresponding ratios of infestation by root-aphids for these five 
plots were as follows : check, JZ percent ; oil of lemon, 20 percent ; 
kerosene, 40 percent ; formalin, 46 percent ; carbolic acid, 74 percent. 
In respect to number of insects in each lot of 50 hills dug from the 
various plots, we find the difference quite as marked. The check plot 
contained 39 ants and 166 aphids to the hill ; the oil of lemon plot, 8 
ants and 18 aphids ; the kerosene plot, 16 ants and 52 aphids ; the 
formalin plot, 20 ants and 64 aphids ; and the carbolic acid, 32 ants 
and 139 aphids. 
Converting these data into percentages of improvement, by com- 
paring the ratios for each experiment with those of the check, we 
find that the benefit to the plot treated with oil of lemon was 70 per- 
cent in number of hills infested by ants, JZ percent in hills infested 
by aphids, 79 percent in the number of ants in the plot, and 89 percent 
in the number of aphids. The corresponding ratios of benefit for the 
