95 
borne by three trees which had not been sprayed, but which stood 
in the row next the sprayed half of the plot, were counted and ex- 
amined in comparison with the product of ten other trees taken at 
various points in this unsprayed part of the orchard: (See dia- 
gram, p. 83 and table on p. 97). Eighty-one and five tenths per 
cent, of the apples on these 3 trees next the experimental plot had 
been injured by the curculios, and 90^ per cent, of the apples on 
the 10 trees farther back, — a difference of 8.8 per cent, of all the 
apples on these trees, which can only be accounted for on the sup- 
position that more curculios went from the three trees nearest the 
sprayed half of the orchard into that half than came into it from 
that half. 
Next, three trees which had been sprayed and which stood in 
the row nearest the unsprayed part of the orchard were similarly 
examined, with the result that 43^ per cent, of the apples on 
these trees were found to have been injured by curculios, while 
26^ per cent, of the apples on 10 other trees taken elsewhere in 
this sprayed plot were so injured. That is, 16.8 per cent, more of 
the fruit on the trees standing- nearest the unsprayed portion was 
injured than on the trees farther back. 
It will be seen that the increase of average injury to apples on 
unsprayed trees near the check plot was nearly twice as great as 
the decrease of injury on the unsprayed trees standing next the 
experimental plot. This can only mean that the excess of curculios 
on the unsprayed fruit to which their greater injury was due did 
not all come from the unsprayed trees in the adjoining row, but 
that approximately half of them must have come in from remoter 
parts of the experimental plot. In order to see how far this mutual 
effect of one plot on the other actually extended inward in either 
direction from the dividing line, apples were picked and counted 
separately for each tree from a series of 14 trees running length- 
wise of the orchard at right angles to the dividing line between the 
two plots. It was found in this way that notable dimunition of cur- 
culio injury on the one side and increase on the other did not extend 
so far as the third row from the dividing line in either direction; 
in other words, that it seemed limited to the first two rows on either 
side of that line. 
To test this conclusion in a more general way, the percentages 
of curculio injury to apples were figured separately for all trees in 
each cross-row from which any fruit was picked, beginning at the 
outer end of the sprayed half of the orchard and ending at the op- 
posite end of the unsprayed half. The details are shown on the ac- 
companying table (p. 97,) from which it will be seen that 12 trees 
in six different transverse rows of the sprayed half of the orchard, 
ranging from the first to the ninth row, bore from 26 to 30 per cent. 
