The Inheritance of Quantitative characters in Maize 115 
contained individuals very near the minus extreme of the small 
parent. 
In actual practice, of course, one must use larger numbers 
than those declared, a few paragraphs back, to be sufficient. 
Following one assumption it was shown that 1,000 F 2 and 1.000 
F 3 plants — 2,000 in all — would be as likely to give the desired 
results as about 250.000 F 2 plants, while following another as- 
sumption, it was seen that 1,000 F 2 , 100 F 3 , and 50 F 4 plants — 
only 1.150 in all — would be required to secure the same result. 
And the assumptions in these two cases were that the F 2 plants 
selected differed not in size but merely in zygotic formulae. 
Since in practice, however, there is no way of distinguishing 
between the several F 2 plants all belonging to the same size 
class but having different zygotic formulae, as many F 3 indi- 
viduals will of course have to be grown from the F 2 plant that is 
heterozygous in only two factors as from the one that is heter- 
ozygous in four factors. Furthermore — and even more impor- 
tant than this — it cannot be known whether two plants of the 
same size actually belong in the same size class. Not only may 
plants that belong in the same size class have different zygotic 
formulae, but those that belong in very different classes may be 
forced together and made to appear to belong in the same class 
by the specially favorable conditions surrounding one and the 
unfavorable conditions surrounding another, or those not only 
belonging in the same class but having identical zygotic form- 
ulae may similarly be made to appear in different classes. 
Numerous examples of confusion of this sort arising thru 
our inability to distinguish at sight between genetic variations 
and mere fluctuations have been encountered in our experiments. 
One will be sufficient for illustration. Of all the P 3 families 
in one cross (Table 36) . Xo. 1146 had the largest number of stalks 
per plant, its mean being 4.24 stalks, but its F 2 parent had only 
three stalks — less than the mean of the F 2 family (510) to which 
it belonged. F 3 family 1149, on the other hand, had a mean 
number of stalks of only 2.04 notwithstanding that its F 2 parent 
had four stalks — above the F 2 mean. One of these F 2 plants 
had been forced below and the other above its true class. In 
height of plant (Table 29), family 1149 exceeded family 1146 by 
only 2.37 decimeters, their respective mean heights being 25.14 
and 22.77 decimeters, tho the F 2 parents of 1149 belonged to the 
highest F 2 class and was about four decimeters taller than the 
F 2 parent of 1146. If one had been selecting a single F 2 plant 
for both height and number of stalks, the 23-decimeter. 4-stalked 
parent of 1149 would doubtless have been chosen in preference 
to the 19-decimeter. 3-stalked parent of 1146. but 1149 actually 
