Relation of Size of Seed and Sprout Value to Yield 27 
single plant indicated in many cases a mixture of varieties. 
The degree of such error in the separation of plants averaged 
7.6 per cent as shown in table 10. 
EFFECT OF COMPETITION BETWEEN PLANTS GROWN FROM SEEDS DIFFERING 
MARKEDLY IN SIZE AND SPROUT VALUE 
The plan of this experiment was to alternate in the row, at 
the ordinary field rate of planting, both large and small wheat 
seeds, and to determine at harvest the relative productiveness 
of the two grades as compared with the relative yields when 
planted alone. The investigation was made with both winter 
and spring wheat. In order to enable a separation of mature 
plants, grown in competition, it seemed necessary to use two 
varieties of wheat with some definite distinguishing feature. 
Consequently, a bearded and a beardless variety were chosen for 
both the winter and the spring wheat. These were respectively 
Turkey Red and Big Frame for the winter wheat, and Scotch 
Fife and Marquis for the spring wheat. Reciprocal tests were 
made in which, for example, large seed of Turkey Red was 
alternated with small seed of Big Frame in the one case and small 
seed of the Turkey Red and large seed of the Big Frame in the 
other. By this arrangement, the effects of variety competition 
were largely counterbalanced. 
The plats used were 7 J feet long and contained 5 rows spaced 
8 inches apart. No additional space was left between plats, but 
the two outside rows were discarded in all cases to prevent the 
effect of plat competition as a source of error. The seeds were 
spaced one-half inch apart in the row which corresponds to the 
normal rate used in farm practice. 
The initial number of plants was counted for each plat and 
the per cent stand or field germination calculated. One hundred 
and eighty seeds were planted per row. In these rows where 
alternating seeds of the two varieties were planted in competition, 
only 90 seeds of each variety were used per row. The yields, 
however, have been calculated on the basis of 180 seeds planted 
in all cases in order that the data be comparable. 
The number of culms and the yield of grain and straw per 
180 seeds planted were determined for all tests. 
The results are contained in tables 11 to 14 and are sum- 
marized in table 15. In these tables, the ratio of results from 
small seed to the results from large seed has been calculated. 
At the bottom of each table is given the average of ratios between 
grades for reciprocal varieties. Thus, the effect of variety differ- 
ences and variety competition has been eliminated in the sum- 
mary data, because both the small and the large seed are repre- 
sented by both varieties. 
