A Genetic Study of PlantlH eight in Phaseolus Vulgaris. 51 
21.20 ±1.90 for Fi of the cross between them, and 36.90 ±2.37 for 
F2 of the same cross. The indication here of an F2 segregation of 
internode-length factors in a cross between two pole beans of 
very unequal internode length is fairly definite. 
In every one of the four crosses between pole and bush beans, 
the F2 pole-bean segregates exhibit somewhat greater variation 
than the pole-bean parent, as measured by the coefficient of varia- 
tion, but not always as measured by the standard deviation. 
The coefficients of variation for July and Snowflake and for the 
F2 families of their crosses with bush beans are: 
July 26.11 ±1.52 per cent 
July-Triumph F2 32.97 ±2.43 percent 
July-Red Marrow F2 29.62±2.03 percent 
Snowflake 27.68 ± 1.86 per cent 
Snowflake-Triumph F2 28.20 ±2.02 percent 
Snowflake-Red Marrow F2 31.74±2.02 percent 
While the indications are not so clear in some of these crosses as 
in others, the evidence as a whole favors the conclusion that there 
is segregation of internode-length factors in crosses between bush 
and pole beans, just as in crosses between two bush or two pole 
races that differ in internode length. 
Whether there has been segregation of genetic factors for 
internode length in Fo can be determined much more positively 
by the constant types that are isolated in F 3 or later generations 
than by statistical constants calculated from the F2 generation 
and from the parent races. ^ No F3 generation of any of the crosses 
considered here has been grown, but partial records of another 
cross give positive evidence of the production of bush beans of 
diverse types with respect to internode length. The cross in 
question is that between Longfellow, a bush bean with rather 
short internodes, and Fillbasket, a pole bean with fairly long 
internodes. The data for internode length were obtained from 
the same cultures as the data for number of internodes, which were 
presented earlier in this paper. 
From a small F2 family of the Longfellow-Fillbasket cross, 
grown in the greenhouse in 1909, a very tall bush segregate was 
chosen and a few F3 plants were grown from it in the garden in 
1910. One of the tallest plants of this F3 family was the parent of 
^ Shull (1914) has pointed out the fact that increased variability of quanti- 
tative characters in F2 might be due to the unequal stimulation of various 
degrees of heterozygosis (unequal heterosis) of factors other than those directly 
concerned in the development of the quantitative characters in question 
and that somewhat diverse F3 types might also be due to different average 
degrees of heterozygosis. No constant difference between extracted types 
could, however, be accounted for in this way. 
