20 
counted in six rows of the checks and in three rows each of the four 
experimental plots. 
Number of hills of corn to the rozv. — The following were the aver- 
age number of hills of corn to the row 372 hills long. In the check, 
3293^ hills ; in the carbolic acid plot, 328 ; oil of lemon, 326 ; formalin, 
317; kerosene, 282. In other words, 11.4 percent of the places for 
hills were without corn in the check, 11.8 percent in the carbolic acid 
plot, 12.4 in the oil of lemon plot, 14.7 percent in the formalin plot, 
and 24.2 percent in the plot treated with kerosene. The differences 
between the check on the one hand and the plots treated with carbolic 
acid and oil of lemon on the other are doubtless insignificant, and per- 
haps the formal'n plot should be included in this statement. Consid- 
erable importance must be attached, however, to the fact that more 
than twice as many hills were missing in the plot treated with kero- 
sene as in the check, three rows of the kerosene plot containing 271 
missing hills as against 255 missing from six rows of the check. In 
view of the injurious effect of kerosene upon the seed, as shown by 
our insectary work and by our field experiment of 1905, we can only 
suppose that some injury was done even by the minimum amount of 
kerosene used in this experiment of 1906. 
Number of stalks of corn to the rote/.— -The number of stalks to 
the row in these various plots was as follows : check, 621 ; oil of 
lemon, 641; formalin. 545; carbolic acid, 543; kerosene, 511. The 
plot treated with the oil of lemon, it will be noticed, contained 20 
stalks more to the row than the check — an increase of 3.2 percent, 
apparently due to a complete destruction of the plants by the insects 
in the check, against which the experimental strip had been protected 
by its treatment. In the remaining plots, on the other hand, the num- 
ber of stalks was diminished in the experimental plots from 12.2 per- 
cent in the formalin strip to 17.5 percent in the kerosene — a diminution 
which can only be accounted for on the supposition that a certain 
amount of the seed was injured by these substances. Indeed, taking 
as a basis of comparison the number of stalks in the plot planted 
with seed which had been treated with oil of lemon, we find the loss 
of plants in the kerosene strip to have been 20 percerft and that in 
the other plots to have been approximately 15 percent — all due, so far 
as one can see, to injury to the seed by the substances employed. 
Ear-bearing stalks to the rozv. — The most significant comparison 
of these plots was made by a count of the stalks bearing ears, the 
number of barren stalks varying greatly according to the treatment 
applied. In the check plot were 412 ear-bearing stalks to the row; 
in the oil of lemon plot, 526 ; in the carbolic acid plot, 505 ; in the 
formalin plot, 485 ; and in the kerosene plot, 439, — all four numbers 
considerably larger than those of the check. The greatest improve- 
ment to be seen in this final test of treatment is in the oil of lemon 
plot, where the yield in ear-bearing stalks was 27.7 percent greater 
than that of the check the corresponding ratios of increase for the 
remaining plots being as follows: carbolic acid, 22.6; formalin, 17.7; 
