36 Nebraska Agricultural Exp. Station, Research Bui. 7. 
pole and bush beans be thrown into one array, the variation is 
markedly increased. The array for the Snowflake-Triumph 
cross arranged in three-internode classes extends from 5 to 32 
internodes and the frequency distribution is: 
11—2—0—0—3—8—8—4—2—3. 
The statistical constants are : 
Mean— 17.66 ±0.96 internodes. 
Standard deviation — 9.07 ±0.68 internodes. 
Coefficient of variation — 51.34 ±4.72 per cent. 
Similarly the frequency distribution for the whole F2 of the 
Snowflake-Red Marrow cross, in three-internode classes from 
5 to 44 internodes is: 
16—4—0—0—3—6—12—19—9—2—4—2—0—1. 
The statistical constants are: 
Mean 21.00 ±0.78 internodes. 
Standard deviation 10.16 ±0.55 internodes. 
Coefficient of variation 48.38 ±3.19 per cent. 
The very large coefficients of variation shown by these F2 families, 
about 50 per cent, as compared with the much smaller coefficients 
for the parents and Fi, from about 10 to 17 per cent, are plainly 
an expression of the Fo segregation in habit of growth. Indeter- 
minate habit necessarily carries with it the ability to produce 
many internodes, while determinate growth makes impossible 
the development of more than a few internodes. But it has been 
shown that distinct races of pole beans, both equally indetermin- 
ate in growth, differ noticeably in the number of internodes they 
ultimately produce. Is it not possible then that a tendency to 
produce a large number of internodes, say 30 to 40, might be 
inherited from a bush bean, which, owing to its determinate 
habit of growth, is itself unable to develop more than a few 
internodes? Likewise is it not possible that a tendency to pro- 
duce a few internodes may also be inherited independently of 
habit of growth? 
These questions, if I have correctly interpreted my data, are 
given an affirmative answer by the results of the cross between 
Snowflake and Red Marrow. Snowflake showed a range of 12 
internodes (14-26) and a mean of 20.48 ±0.44. The Fi plants 
had a range of 12 internodes (20-32) with a mean of 26.56 ±0.48. 
The mean for the F2 pole-bean segregates was 26.31 ±0.47, very 
nearly the same as in Fi, but the range was 27 internodes (17-44). 
Red Marrow had a range of 3 internodes (4-7) and a mean of 
5.78 ±0.05. The F2 bush segregates of the cross between it and 
Snowflake had a smaller mean, 5.45 ±0.16. The range of varia- 
tion was the same as for Red Marrow, but the distribution was 
not that of normal fluctuation. While no new values not seen in 
