The Inheritance of Quantitative Characters in Maize 111 
for crossing are alike in all but one or two characters. A breeder 
does not usually want merely to combine one desirable character 
of a certain variety with one other equally desirable character of 
another variety. He perhaps has a variety that for certain uses 
is almost ideal but it lacks conspicuously some one or two very 
desirable characters that he sees in some other variety. He 
wants to keep his variety as it is but transfer bodily to it the 
coveted character from the other variety. Very likely this second 
variety is only mediocre in numerous respects As a matter of 
fact, the two varieties, instead of differing by merely this one 
prominent character, actually differ in very many characters. 
In transferring, therefore, the much desired character to the 
otherwise ideal variety, one is apt to lose some of the highly- 
prized characters of the latter, unless very large numbers are 
grown in F 2 or selection practiced for two or three more genera- 
tions. The only difference in manipulation between quantitative 
and qualitative characters is, then, one of degree. And in 
practical breeding even this distinction is of little importance, 
since practically every cross between commercial varieties must 
necessarily include both qualitative and quantitative characters. 
The large number of factors by which distinct commercial 
sorts of crop plants are differentiated is almost staggering. 
Take as an illustration an extremely simple case. We will sup- 
pose that it is desired to produce a better silage corn than that 
afforded by the tall few-stalked dent varieties of the West and 
South or the short many-stalked flint varieties of the Northeast — 
in short a type of corn combining the tallness of one variety with 
the many-stalked condition of the other. For sake of simplicity, 
we will disregard differences in season of ripening and various 
other characters, which, as a matter of fact, could not ordinarily 
be disregarded in actual practice. We will assume that the 
varieties in question differ by five Mendelian factors influencing 
height of plant and by five factors affecting number of stalks per 
plant and that each of these factors is independent of the others 
in inheritance — assumptions which, tho they may not represent 
accurately the actual facts, are at least not unreasonable and 
will at any rate serve for purpose of illustration. From a cross 
of these varieties, an F 2 of only about 1,000 plants, or say one- 
tenth acre, would be required to give an even chance of recovering 
a many-stalked type like the one parent or a tall-stalked type 
like the other parent. Perhaps three or four times that many 
plants should be grown to be at all sure of recovering the parent 
types. But with this number of individuals there could be no 
hope that any of the very tall plants would also have very many 
stalks. With the parents differing by ten factors, five for height 
