A Genetic Study of Plant Height in Phaseolus Vulgaris. 37 
Red Marrow were found in the Fo plants, the relatively large 
number of plants with the extreme numbers of internodes, 4 and 7, 
suggests that it would be possible to isolate distinct types of 
bush beans in F3 from this F2 lot. Only by further breeding tests 
can it be told whether this suggestion is correct. The F3 genera- 
tion of this cross has not been grown, but fortunately I have 
partial records of another cross which afford direct evidence of 
the production of diverse types of bush beans from a cross between 
bush and pole races. The cross is one between Longfellow and 
Fillbasket. 
Longfellow is a rather short bush bean and Fillbasket is a 
medium tall pole bean. Counts of 10 to 20 plants each when 
grown under similar conditions in the garden in 1909 (except F2, 
which was grown in the greenhouse) gave ranges of variation and 
means as follows: 
Internode numbers 
Mean Range 
Fillbasket 17.3 14-22= 8 
Fillbasket-Longfellow, Fr 21.4 18-24= 6 
Fillbasket-Longfellow, Fo pole 20.8 13-26 = 13 
Fillbasket-Longfellow, F2 bush 6.2 4- 8 = 4 
Longfellow 5.3 4-6= 2 
From two Fo bush plants with 6 and 7 internodes respectively, 
two small F3 families were grown in 1910 each showing about the 
same range of variation as that of Fo. In 1911 near Boston, 
Massachusetts, were grown several F4 families of this cross. One 
of these (Family 3,254) was from a bush plant with 6 internodes 
which was of the F3 lot descended from the 6-internode F2 plant. 
The other lot (3,251) was from a bush plant with 7 internodes of 
the F3 progeny of the 7-internode F2 plant. In the same garden 
and very near these F4 families, were grown several families of 
Longfellow. All of these were descended directly from the indi- 
vidual plant of Longfellow used in making the cross with Fill- 
basket now under discussion, but were four generations removed 
from that plant and therefore were its great-great-grandchildren. 
The several families of Longfellow showed only slight differences 
in height of plants. One of them (3,247) was examined for number 
of internodes. The results are given in Table 3. 
One of the F4 families of the Longfellow-Fillbasket cross was 
very similar to the family of Longfellow of which internode 
counts were made, the means for number of internodes being 5.17 =fc= 
0.06 and 4.94 ±0.08 respectively, a difference of only 0.23 ±0.10. 
The other F4 family, however, had a mean number of inter- 
nodes of 6.44 ±0.08, which is 1.27 ±0.10 more than the first F4 
family had. The variation of the three lots, as indicated by the 
coefficient of variation, was so small and so nearly the same for 
