14 



H. H. Love and C. E. Leighty 



Yield per culm and height of culm 



Yield per culm and number of kernels per 



culm 



Yield per culm and number of spikelets per 



culm 



Yield per culm and average weight per kernel 

 Average weight per kernel and height of culm 

 Average weight per kernel and number of 



kernels per culm 



New Zealand 



.511 ±.029 



.836±.012 



.965 ±.020 

 .790±.015 

 .221 ±.037 



.365 ±.034 



Berkeley 



.654 ±.022 



.864 ±.010 



.521 ±.028 

 .753 ±.017 

 .384 ±.033 



.378 ±.033 



for average height of plant in centimeters is 52. 990 ±.181 for 1908. The 

 plants were grown in drill rows, and the conditions were not so favorable 

 for the development of tall culms in that year as they were in later years. 

 The means for average height in the other years are, in order of value, 

 70.840±.234 for 1910, 75.200±.182 for 1912, and 76.110±.277 for 1909. 

 The differences in height are not great for these years. The variabihty 

 for this character as indicated by the standard deviation* is greatest 

 for 1909 (the year for which the mean is largest), being then 9. 188 ± . 196; 

 but the variability for 1908, 7.715±.128, is next in order although the 

 mean is smallest for that year. This doubtless results from the fact 

 that culms were then the units. For 1912 the variability is least, being 

 then 5. 405 ±.129; for 1910 it is 6.950±.166. 



The means for the total yield of plant vary from 3. 458 ±.045 grams 

 for 1910 to 4. 032 ±.068 for 1909 and 7. 962 ±.113 for 1912. The 

 variation as indicated by the standard deviations is in the same order, 

 being 1.323 ±.032 for 1910, 2. 249 ±.048 for 1909, and 3. 353 ±.080 

 for 1912. The average yield of culm per plant has a mean for 1908 of 

 .409 ±.006 gram, this being lower than the mean for any other year. 

 For 1910 the mean is .962±.008, for 1909 it is .978±.010, and for 1912 

 it is 1.266 ±.008. The standard deviations for the average yield are 



* Variability is expressed by means of both the standard deviation and the coefficient of variability. 

 The standard deviation expresses the variation as measured from the mean, and is expressed in the units 

 of measurement. It is thus an absolute expression of variability. The coefficient of variability is a relative 

 expression by means of which, for instance, a variable character whose data have been recorded in pounds 

 is directly comparable with another whose data have been measured in inches. For the points under 

 consideration the standard deviation very well expresses conditions as they actually are; at the same time 

 the coefficient of variability is used also in these discussions. 



