52 Nebraska Agricultural Exp. Station, Research Bui. 7. 
an F4 family (3,251) grown near Boston, Massachusetts, in 1911. 
A medium small bush plant was also chosen from the same F2 
family, a small F.3 family grown from it, and a medium small 
F3 plant chosen as the parent of an F4 family (3,254), also grown 
near Boston. In the same garden with these F4 families, were 
several lots of the parent race, Longfellow, all descended directly 
from the individual plant used in making the Longfellow-Fill- 
basket cross. All these lots of the bush parent were much alike 
in height. The plants of one of them (3,247) were measured for 
comparison with the two very unlike F4 families. In making these 
records the total length of the main axis and the total number of 
internodes were determined instead of measuring each internode 
separately, as has been done in most of the more recent work. 
The internode lengths reported here were calculated from the 
data recorded. As has been pointed out before, this does not 
give as accurate a measure of mean internode length as is obtained 
from a definite number of internodes, but it does nevertheless 
give some notion of differences in internode lengths between the 
three lots. Statistical constants for the three lots of plants are: 
Number of Standard Coefficient 
individuals Mean deviation of variation 
Longfellow (3,247) 18 27.7±0.7 4.6±0.5 16.5±1.9 
Longfellow-FillbasketF4'(3,254). .36 22.3±0.6 5.3±0.4 23.7±2.0 
Longfellow-FillbasketF4 (3,251). .41 54.2 ±1.5 14.7 ±1.1 27.1 ±2.2 
One of the F4 families had, it is seen, somewhat shorter inter- 
nodes than the family chosen to represent the bush parent of the 
cross, while the other F4 family had internodes almost twice as 
long as the parent race. It is probable that the several families 
of the bush-bean parent would have been found to differ some- 
what in internode length, had they been measured, but certainly 
no such difference existed between them as between the two F4 
families. 
Since these F4 families not only differed much in internode 
length but also showed some difference in number of internodes 
(Table 3), it follows that they differed also in height of plants. 
In fact the difference in this respect was remarkable. F4 family 
No. 3,251 was one of the tallest lots of bush beans that I have 
ever seen. The mean length of the main axis in Longfellow 
(3,247) was about 137 mm., in the short F4 family (3,254) about 
121 mm., and in the tall F4 fami^ (3,251) about 349 mm. The 
extreme difference in height of plants of these families is shown 
in Figure 15. It is plain, then, that, while the segregation into 
pole and bush beans in Fo of a pole-bush cross is as definite as 
segregation in any simple Mendelian character, some of the bush 
segregates at least are very unlike the bush parent in height of 
plant. 
