July i, 1921 
Further Experiments in Field Technic in Plot Tests 485 
aged 91 and the middle row 83. In another trial, four rows of turnips 
each 18 inches apart w T ere sown July 16 in plots 13 feet by 9 feet with 9 
feet 6 inches as the distance between the outside rows of any two plots. 
These were harvested September 28. Here on the basis of 100 for the 
heaviest row of each plot, the roots from the outside rows yielded at the 
rate of 95 and the inside rows at the rate of 70. Discarding the plants in 
the outside rows before estimating yield is recommended. 
In a later paper the same writer (6) reports that potatoes in 3-foot 
rows, with alleys several feet wide on one side and from 4 to 5 feet wide 
on the ends, showed marked border effect. The rows extended north 
and south. Inner row No. 1, next to the outer row, yielded 72, and the 
third row in the plot yielded 72, as compared with 100 for the outside 
rows. Yields of the end plants of the rows was as 100 to 82, as com¬ 
pared with the average of the other plants in the same rows. I11 other 
trials end plants yielded 100 to 87, 100 to 92, and 100 to 88.4, respec¬ 
tively, as compared with the average for the other plants in the same 
rows. Corner plants—those having additional space in two directions— 
yielded 100 to 79, as compared with the other plants in the outside 
rows, and 100 to 57, as compared with the inner plants. The statement 
is made that in yield trials at Wisley outer rows are planted which do 
not essentially belong to the plots. They are to protect the rows of 
the plants from border effect. 
Wheeler ( 21) notes that in Germany at some locations provision had 
been made to prevent border effect, and that in some experiments at 
the Rhode Island Station border effect was eliminated by removing a 
strip 3 feet wide on the sides and 6 feet wide on the ends. This width 
was decided on because the area remaining was then i f 10 acre. 
Bedford and Pickering (3) in field tests found the weights of the entire 
plants in outside rows to be heavier than the weights of the produce 
from the inner rows. The relation in percentage was as follows: Mustard, 
outside 297, 201, and 200, as compared with 100 per cent for the inner 
rows; wheat at Ridgemont, outside 131, inside 100; wheat at Woburn, 
outside 204 and 161, inside 100; barley at Harpenden, outside 126, 
inside 100. On plots manured from 100 to 300 tons per acre the outside 
rows of mustard were 190, as compared with the inner 100; on land 
less heavily manured the outside were 228 and the inside 100. From 
these results the authors conclude that, under ordinary circumstances, 
approximately one-fourth of the border effect is due to increased food 
supply and three-fourths to decrease in toxic effect. 
The observation is also made that with the mustard plant the border 
effect did not extend beyond 6 inches from the edge of the plot, no 
effect in the second rows being noticed when the rows were 9 inches 
apart. The width of the alleys or fallow spaces between plots is not 
mentioned. 
