Aug. 26, 1918 
Variation and Correlation in Wheat 
389 
Considering the interrelation of plant characters as a whole, there is a 
range from practically none to a high correlation. Correlation is modified 
by environment, the degree of modification due to this cause varying 
with the characters considered. ♦ 
An increased yield of kernels is very closely accompanied by an increase 
in number of kernels, number of culms, and total length of spikes; and 
somewhat less closely accompanied by increase in average weight of 
kernels per plant, average height of culms, and average length of spikes. 
A larger number of culms per plant is accompanied by a greater total 
length of spikes but not by a greater average length of spikes. 
Average weight of kernels is substantially and fairly consistently 
correlated with yield of kernels; and, subject to radical change due to 
environment, moderately correlated with average length of spikes. 
With number of kernels, the correlation is rather low but always con¬ 
sistent. Average weight of kernels is practically independent of average 
length of spikes. 
There is a distinct tendency for greater average height of culms to 
be accompanied by a greater average length of spikes, number of kernels, 
and a higher yield of kernels. Average height of culms is substantially 
correlated with average weight of kernels and moderately correlated 
with total length of spikes in three years out of four. The correlation 
between average height of culms and number of culms is always low. 
There is a distinct tendency for plants of varying heights at second 
leaf to maintain the same relative heights at six weeks; but there is a 
lesser tendency for this relation to be maintained at maturity. 
GENERAL DISCUSSION 
During early growth in 1914 and 1915, the means for height of the 
plants are greater than those in 1916 and 1917, owing to the somewhat 
more productive soil on which the plants were grown, to the more favor¬ 
able weather conditions, and to a higher average weight of seed planted. 
In 1915 the favorable growing conditions continued throughout the 
season, and the mean for each plant character at maturity, except 
average weight of seed, is the highest in the 4-year period. In 1914, 
during July, drouth followed by an epidemic of black-stemrust lowered 
materially the means for all plant characters at maturity. 
For each of the characters studied, except yield of kernels, the vari¬ 
ability as indicated by the standard deviations, is as high as or higher 
in 1914 and 1915 than in 1916 and 1917. The generally higher variabil¬ 
ity in the former two as compared with that in the latter two years is 
accompanied by generally lower correlation coefficients (1) when weight 
of seed sown is correlated with resultant plant characters and (2) when 
yield of kernels is correlated with other plant characters. 
