56 Nebraska Agricultural Exp. Station, Research Bui. 7. 
The F2 bush plants were on the whole considerably shorter than 
the bush parent, none being so tall as the taller plants of Tall- 
bush. It was among the F2 pole-bean plants, however, that the 
most remarkable variation was exhibited. The smallest of these 
plants were only about 35 centimeters high — shorter than any 
of the few plants of Snowflake grown with them but not much 
shorter than are sometimes found among weak plants of that race. 
At the other extreme were plants nearly 200 centimeters tall — a 
height not materially less than that attained by the taller races 
of pole beans when grown in as small pots as were these F2 plants. 
The great variation in height of the F2 plants of this cross is 
perhaps made even more obvious by the photographs reproduced 
in Figure 16. It seems a clear inference from the data here pre- 
sented that a tall race of bush beans, which inherited its relatively 
great height from a tall pole-bean parent of an earlier cross, has 
transmitted tallness to its pole-bean progeny when crossed with a 
very short pole bean. Other factors for height of plant are, then, 
inherited independently of habit of growth. 
FACTORS FOR PLANT HEIGHT— THEIR SEGREGATION AND 
POSSIBLE MODIFICATION IN CROSSES. 
It was noted in the introduction to this paper that quantita- 
tive characters, including height of plants in maize at least, are 
usually intermediate between the parents in Fi and show a wide 
range of variation in F2, thus furnishing a basis for the isolation 
of numerous quantitatively distinct types in F3. It was also 
noted that some quantitative characters, particularly height of 
plants of several very diverse groups including beans, exhibit 
perfect dominance in Fi, followed by simple 3:1 segregation in F2 
and typical Mendelian behavior in later generations. That 
height of plants should in some cases fall into the one and in 
other cases into the other of these two classes of behavior presents 
an important problem. If the inheritance of quantitative charac- 
ters in general is to be interpreted on the basis of the multiple- 
factor hypothesis, it is to be expected that some quantitative 
characters will be found to differ by only a single genetic factor 
just as other quantitative differences may be due to three or four 
or ten factors. It seems probable, then, that, within any one 
species, and with respect to the same plant part, great quantitative 
differences are in general due to many factors and small quantita- 
tive differences to one or to a few factors, tho of course it is not 
maintained that all size factors, even in this restricted case, 
necessarily have equal value. (In this connection see Shull 1914.) 
But the 3-1 segregation, denoting a single-factor difference, is 
