Aug. 26, 1918 
Variation and Correlation m Wheat 
387 
The coefficients for average weight of kernels correlated with average 
length of spikes are o.i53±o.o38, 0.120^0.027, 0.552^=0.017, and 
0.411 ±0.025, respectively, for the four years. The relatively low corre¬ 
lation in the first two and the substantial correlation in the last two 
years indicates that under the conditions of environment which pre¬ 
vailed in 1916 and 1917 there is a strong tendency for the two characters 
to move together; and that under extreme environmental conditions 
such as prevailed in 1914 and 1915 the relation is considerably reduced. 
For average weight of kernels correlated with total length of spikes, 
the coefficients range from 0.001 ±0.038 in 1914 to o.i59±o.o27 in 1915. 
There is no correlation in 1914, and idt the three other years the rela¬ 
tion is low. Therefore the conclusion must be that the two characters 
move practically independent of each other. 
When average weight of kernels is correlated with number of kernels, 
number of culms, average length of spikes, and total length of spikes, 
no consistently high relationship is found. Subject to radical change 
by environment, there is a moderate relation with average length of 
spikes. With number of kernels the correlation is rather low but con¬ 
sistent. Average weight of kernels is practically independent of total 
length of spikes. 
AVERAGE HEIGHT OE CULMS CORRELATED WITH OTHER PLANT CHARACTERS 
The coefficients for average length of culms correlated with number 
of kernels per plant are 0.257 ±0.036, 0.339±0.025, 0.364±0.022, and 
0.431 ±0.025, respectively, for the four years. This is a substantial and 
fairly consistent relation very similar to that found between yield of 
kernels and average height of culms. There is a tendency for an increase 
or a decrease in average height of culms to result in the production of a 
larger or smaller number of kernels per plant. 
When average height of culms is correlated with average weight of 
kernels, the coefficients are o.458±o.o30 in 1914, 0.071 ±0.028 in 1915, 
o.648±o.oi4 in 1916, and o.426±o.o25 in 1917. With the exception of 
the very low correlation in 1915 due to extremely favorable environ¬ 
mental conditions, the relation is substantial. The indications are that 
under ordinary conditions there is a tendency for increase or decrease in 
average height of culms to be accompanied by a raising or lowering of 
average kernel weight. 
For average height of culms correlated with number of culms per plant 
the coefficients are —0.195±0.037 in 1914, —0.092±0.028 in 1915, 
o.o46±o.o 25 in 1916, and 0.205±0.029 in 1917- The correlation varies 
considerably from year to year and is low in each instance. Therefore 
the conclusion may be drawn that the slight tendency for the two char¬ 
acters to vary together is highly modified by the influences of environ¬ 
ment. 
