26 
of furrow. If they are fewer than this it may be that, in a year favor- 
able for corn, no sufficient injury to any of the plants will be done by 
the root-lice to make any noticeable difference between corn growing 
from treated and from untreated seed. 
As a safeguard against possible injury to the kind of seed one 
is using by the kind of materials one may chance to purchase, and in 
view especially of the well-known differences in seed and the differ- 
ences also in materials of chemical manufacture, each experiment 
should be preceded by a seed-corn test, made in the usual form but 
in a way to bring into comparison treated and untreated seed with 
respect to the percentage of kernels to grow. In this way the possible 
failure of an experiment and some loss of stand may be prevented. 
