34 H. H. Love and C. E. Leighty 



.042±.034 for 1910 to .523 ±.024 for 1912. This correlation, like the 

 one just considered, is fluctuating and ranges from no correlation to a high 

 correlation. And in the .same way it must be concluded that the cor- 

 relation which exists between the height of plant and the number of 

 culms produced by that plant is strong under some conditions, but that 

 it is responsive to, and is modified by, environment to such an extent 

 that on occasion the two characters may be practically independent of 

 each other. 



Considering the relation of average height of plant to the other plant 

 characters as a whole, it has been shown that there is generally a very 

 high, positive, and stable correlation between the average height of 

 plant and (1) total and average yield, (2) total and average number of 

 kernels produced, (3) average number of spikelets per culm. These 

 relations are disturbed to some extent by enviroimiental conditions 

 such as exist in different years. It is conceivable that they might vary 

 to a greater extent, even to such a degree that there would be no cor- 

 relation between these characters. Under very abnormal weather con- 

 ditions which would allow vegetative development to some extent but 

 under which grain development would be prohibited, correlations would 

 be impossible; but correlations such as those mentioned above are hkely 

 to obtain in any case when kernels are produced in sufficient quantity 

 to render possible the determination of averages. Whenever kernels are 

 developed, the taller plants would tend to develop the greater numbers 

 of spikelets and kernels, and consequently to produce the higher yields. 



The correlation of average height with the averages per culm are higher 

 than those with the plant totals. This is undoubtedly due to the fact 

 that height itself is an average, and when averages are used there has 

 been effected a similar and simultaneous change in both the height and 

 the other character under consideration. 



The correlations between height of plant and (1) average w^eight of 

 kernels, (2) number of culms, are fluctuating in amount, occasionally 

 being high, but being influenced by environmental conditions from year 

 to year so that the relation of one to the other may become such that they 

 vary independently. A short plant, then, on occasion may develop as 

 large kernels as a tall plant, or may develop as many culms. 



