50 Nebraska Agricultural Exp. Station, Research Bui. 7. 
Triumph and Snowflake are found to have nearly equal internode 
lengths while Red Marrow and Snowflake are found to have 
quite unHke internode lengths, the fact that the variation in F2 
of the Triumph-Snowflake cross was no greater than the variation 
in the parent races would be readily understood, as already 
pointed out, but greater variation would then be expected in F2 
of the Red Marrow-Snowflake cross than in the parent races. 
That this was just what occurred is seen from the data given in 
Table 4. For mean lengths of the first five internodes, the coeffi- 
cients of variation are for Snowflake 15.81^1.01 per cent, for 
Red Marrow 15.81 ±0.91 per cent, for the F2 pole-bean segregates 
33.33 ±2.16 per cent, and for the F2 bush-bean segregates 22.86 ± 
2.86 per cent. The range of variation in F2 was 40 mm. for the 
pole-bean segregates and 25 mm. for the bush-bean segregates, 
as against 20 mm. for Fi, 20 mm. for Snowflake, and 15 mm. for 
Red Marrow. Increased variation of F2 over the parent races 
was also shown for each individual internode of the first five. 
By way of summary it can be said that there is distinct 
evidence of segregation in F2 of factors for length of the first five 
internodes of two crosses between pole and bush beans. No 
evidence of such segregation has been found in case of two other 
such crosses. Where there is distinct segregation in F2 it should be 
possible to isolate types of both bush and pole beans of different 
internode lengths from a single cross of pole and bush races. 
No F3 progenies of the crosses here under consideration have 
been grown, but very distinct types of bush beans with respect 
to internode length have been isolated from a cross of other 
varieties. The evidence of this will be given later. 
Table 5 exhibits the variations in length of the first 15 inter- 
nodes of some of the races and crosses discussed above. It is 
obvious that only plants of indeterminate growth habit can be 
included in the table. This eliminates from the comparison not 
only the bush races and crosses between bush races but also 
the F2 bush segregates of pole-bush crosses. For the races and crosses 
that can be included here, the data are more reliable than those 
for the first five internodes. The reasons for this were discussed 
in the consideration of methods of calculating internode lengths. 
The mean length of 15 internodes is a better measure of internode 
length than the mean length of only five internodes. 
The mean length of the first 15 internodes in F2 of the July- 
Snowflake cross, 69.29 ±1.99, is almost exactly intermediate 
between the means of the parent races, 89.69 ±1.84 for July and 
46.28 ±1.13 for Snowflake. In range of variation, the F2 plants 
extend to the outer extremes of the two parents. The coefficients 
of variation are 27.68 ±1.86 for Snowflake, 26.11 ±1.52 for July, 
