June i4 ,1924 Correlation among Quantitative Characters in Maize 1099 
from the fact that in maize the occurrence of more than 20 factors almost cer¬ 
tainly would necessitate linkages. If the linkage were perfect the effect would be 
of a single factor and as the degree of linkage decreased the effect of two inde¬ 
pendent factors would be approached. 
With the exception of the number of days to silk, the F 2 distributions are 
fairly regular and very similar to those for the plants of the Fj. The distribu¬ 
tion for number of days to silk suggests trimodality with a large group of early 
plants and a fairly well-defined population of late plants. However, the parental 
ranges are not reached and no plants were obtained as early as the latest of the 
early parent nor as late as the earliest plants of the late parent and the factorial 
difference between the parents is estimated at 26. 
EJEN&TH OB 
//V CBNT/MBTEB^S 
Fig. 5.—Frequency distributions for length of ear. Estimated factorial difference between parents 17-18 
In the other characters the parental types were represented in the extreme 
plants of the F 2 distributions, but the parental ranges were not exceeded except 
in the previously-mentioned case of number of rows on the ear. No frequency 
polygon is given for this latter character, since the parental difference was not 
large. 
CORRELATION BETWEEN THE CHARACTERS 
The coefficients of correlation among the characters of the Fi plants may be 
taken as an indication of the physiological as distinct from the genetic relation¬ 
ships. An examination of the Fi coefficients given in Table II shows that there 
is only one significant correlation, namely, that between length of ear and num¬ 
ber of rows, indicating that the conditions which favored the production of long 
ears also resulted in the development of a large number of rows. 
96462—241’-4 
