Nov. 26, 1917 
Field Technic in Rod Row Tests 
403 
The work in 1915 was of a preliminary nature and does not necessi¬ 
tate an extensive discussion. It is only of interest as indicating the reason 
for the methods used in extensive tests made in 1916. In 1915 oat and 
wheat rows replicated nine times were grown for comparison with yields 
obtained in duplicated one-one hundred and tenth-acre plots. Corre¬ 
lation coefficients were calculated, giving the comparison between yields 
of 18 varieties of wheat and 21 varieties of oats. The coefficients are as 
follows: 
Wheat.f=+o. 593±o. 123 
Oats.f = 4*o. 200 ±0. 141 
The same plot and series was used for these tests so that environ¬ 
mental conditions are relatively uniform. The fact that no closer re¬ 
lation in yield was obtained may be partially explained by the unre¬ 
liability of the results as obtained in the duplicated one one-hundred- 
and-tenth-acre plot tests. (See following paragraph.) There is also 
the possibility of a greater competitive effect between adjacent rows of 
oats than for wheat. 
A comparison of the relative reliability of single one one-hundred-and- 
tenth-acre plots and single rows of Bluestem wheat and I 4 gowa oats 
is given as a reason for further studies of the row method. The coeffi¬ 
cient of variability is used as a measure of comparative reliability of 
the results of these tests (Table I). 
Table I .—Coefficients of variability for yield of wheat and oats as tested in one one- 
hundred-and-tenth-acre plots and rod rows , igi$ 
Variety. 
Type of plot. 
Coefficient of 
variability. 
Haynes Bluestem wheat. 
Single rod row. 
13. 73 ±°. 76 
14* 43 ±2. 16 
18. 87 ± I. 02 
10. 74 ±I- 29 
Do... 
Single one one-hundred-and- 
tenth-acre plot. 
Single rod row. 
Ligowa oats. 
Do. 
Single one one-hundred-and- 
tenth-acre plot. 
These tests indicate that a single row is as reliable as a single one- 
one hundred and tenth-acre plot for Haynes Bluestem wheat, although 
for oats the plot test appears to be more reliable. The result of this 
test, together with experiments of other workers, has led to the tentative 
adoption of one-fortieth-acre plots at University Farm. Further 
studies will determine the number of replications necessary for the plot 
work. The rod row test appears very promising, although further 
studies seem necessary, particularly as to competition between adjacent 
rows and number of necessary replications. 
The main part of the experimental results here reported consist of 
studies of competition between adjacent rows of different varieties and 
