Variation and Correlation of Oats — Part I 13 



TABLE 2. Correlation Coefficients for the Different Characters of Wheat 



Studied 



Characters 



Correlation 

 coefficients 



Height of plant and yield 



Number of kernels per plant and yield 



•Height of plant and average weight of grains 



Total number of grains per plant and average weight of grains. 

 Yield in grains and average weight of grains in milligrams 



0.294 ±0.032 

 0.985 ±0.001 

 0.278 ±0.033 

 0.251 ±0.033 

 0.327 ±0.031 



culm and yield, of .925 ±.004 on ordinary soil and .890 ±.006 on manured 

 soil. All correlations were greater on the poorer soil. Increased fertility 

 tended to decrease variability. 



Roberts (1912) made biometrical studies of three pure strains of wheat 

 for two years. The characters considered were culm length, head length, 

 number of culms per plant, number of grains per head, and yield. The 

 means of the different characters are considerably modified by conditions 

 of growth, increasing with improvement in these conditions, but a better 

 growing season reduces their variability. A high positive correlation 

 was found between yield and number of culms, and culm length and num- 

 ber of grains, and a fairly high positive correlation between culm length 

 and head length, also between number of culms per plant and both culm 

 length and number of grains. The correlation between culm length and 

 head length varied considerably from one year to another, but there was 

 little variation observed in the correlation between culm length and number 

 of kernels. 



The results reported by Leighty (1912) on a pure line of oats are incor- 

 porated in this paper and need not be discussed here. 



Whitcomb (1913) reports biometrical results on two varieties of barley. 

 The coefficients of correlation for average of plant, of the two varieties, 

 are ^iven in the table on the following page. 



DEVELOPMENT AND VARIATION 



The means, standard deviations, and coefficients of variability for the 

 different characters under observation are given in Table 3. The mean 



