ii8 PLANT-BREEDING 



In selection, uniformity is one of the main purposes, 

 but the shape and color of the ears, their butt and tip ends, 

 the number and direction of the rows, the furrows between 

 the rows and many other points have to be considered. It 

 is only by an actual study of these variations that a farmer 

 may become familiar with all the different types. It would 

 be quite superfluous to try to describe them here, the more 

 so as we shall have to cjuote a number of instances, when 

 dealing with the work of selection. 



Perhaps the most important discovery which has been 

 made concerning these minor variations, is that of their 

 constancy. All the kernels of a selected ear have the same 

 qualities, provided, of course, that cross-pollination has been 

 sufficiently excluded. This is easily seen in their visible or 

 physical quahties, but the experiments of Hopkins have 

 shown that the same rule prevails for the chemical consti- 

 tution, including the relative development of the main indus- 

 trial constituents. Moreover, it is true for the hereditary 

 qualities. 



In the first place, direct experiments have shown that 

 neither the yield nor the quaUty of the grain is essentially 

 affected by choosing the seed-grains from the butt end, the 

 middle, or the tip of an ear. Furthermore, it is now cus- 

 tomary, as we shall soon see, to sow the kernels of selected ears 

 in single rows, each ear to a row, and by this method the 

 fact of the individuality of the rows has become quite con- 

 spicuous, A whole row, grown from the kernels of a single 

 ear, may produce numerous barren stalks, or weak plants, 

 or small ears with imperfect yield, or be excellent in strength, 

 productivity, and uniform in other peculiar characteristics. 

 This fact is now the acknowledged basis of the main prin- 

 ciple of com selection. 



From these and many other concurring observations we 

 may conclude that the variability of corn within the varieties 



