152 



Clyde E. Leighty 



The correlations obtained between the several characters considered for 

 plants of the same pure lines grown in hills and in drills do not in the main 

 differ radically from each other for the two methods of planting. There 

 are, however, some rather large differences in a few cases. The differences, 

 as given in Table 9, range from .002 ± .004 to .391 ± .046. The latter 

 is of considerable value and is doubtless significant. Five out of fourteen 

 correlations have differences amomiting to more than .10, and of these four 

 are found in the Sixty Day variety. Only three have differences amounting 



TABLE 9. Summary of Differences ix Correlatiox Coefficients for the Same 

 Characters in the Same Pure Lines Grown in Hills and in Drills 



Characters correlated 



Extreme difference 



Difference 

 P. E. difference 





1219 and 1221 



1257 and 1259 



1219 

 and 

 1221 



1257 

 and 

 1259 



Yield and height 



Yield and number of kernels 



.006 ± .015 

 .018 ± .006 

 .015 ± .008 

 .077 ± .011 

 .046 ± .040 

 .228 ± .044 

 .090 ± .050 



.087 ± .017 

 .005 ± .004 

 .002 ± .004 

 .138± .016 

 .156± .042 

 .391 ± .046 

 .229 ± .050 



0.4 

 3.0 

 1.9 

 7.0 

 1.1 

 5.2 

 1.8 



5.1 

 1.2 



Yield and number of spikelets 



0.5 



Yield and weight of straw 



8.6 



Yield and weight of kernels 



3.7 



Weight of kernels and height 



Weight of kernels and number of kernels . . 



8.5 

 4.6 



to more than about five times their probable errors, and two of these are 

 in Sixty Day, these two being when yield is correlated with weight of 

 straw and when weight of kernels is correlated with average height; 

 the third case is when average yield is correlated with weight of straw 

 in Early Champion. 



Whenever great differences in the coefficients occm', those for the plants 

 grown in hills are always the smaller in amount. It seems, then, that 

 correlation is reduced in amount by allowing more room for the growth 

 and development of plants. Correlation may be considerably modified 

 by the method of growmg plants. 



