A COMPARISON OF MAIZE-BREEDING METHODS 3 
between the degree of inbreeding and vigor. This is true especially 
where the combination of a number of lines approximates an average 
condition, as in the yield tests of 1923 and 1924. 
Table 5 gives the average coefficient of inbreeding of the seed 
stocks compared in 1923 and 1924 and affords a striking illustration 
of the extent to which the effects of inbreeding may be overcome by 
selection. The F, progenies of 1924 exceeded all other seed stocks 
in yield, yet the average coefficient of inbreeding of these progenies 
was 78.91, the highest coefficient of the series. A coefficient of 78 
indicates a degree of inbreeding slightly in excess of that resulting 
from two generations of selfing. If uninfluenced by the selection of 
heterozygous individuals, more than three-fourths of the factors 
heterozygous in the foundation stock would have become homozygous. 
TaBLe 5.—Coefficient cf inbreeding of the seed stocks of maize compared in 1923 
and 1924 
Coefficient of 
inbreeding 
Year and seed stock 
Range of 
| Average | individual 
| progenies 
1923: 
CeURSIIG DT 2 NN 2 a SE ee Se Penns 2 Se ee ee eee ee 1159) ve Sirol 
eee ee re ees rene eee A ee ee A ee eS 4252} *3i: 3 10 1aok 
1924: 
amen ake Ry od ey ee Fe 13.6| 8.7t018.3 
ee rea eee eg SEW se ee Se ie ee ee et 78.9 0 to 96.9 
Pap soee SS Se Se a ee ee ee Se ees 51.6 | 25.0 to 78.3 
ere ee er ae 2 eee ee ea ee 27.1 | 23.3 to 28.5 
Some explanation is needed of the diversity in the behavior of the 
various seed stocks representing combinations of selfed lines. The 
1923 F, and 1924 F, seed stocks proved to be decidedly inferior to 
the crossbred seed, while the 1924 F, and F, equaled or exceeded the 
crossbred stock. It is of course expected that individual combina- 
tions will differ widely in yield and vigor, but it was thought that in 
each seed stock a sufficient number of combinations were lumped to 
represent an average condition, and the difficulty of making accurate 
field comparisons makes it impracticable to test a large number of 
individual combinations. 
The disparity is not to be explained by differences in heterosis, for 
the coefficients of inbreeding show the 1923 F, and the 1924 F, to be 
less inbred than the other seed stocks from the selfed experiment. 
The selfed lines from which the 1923 F, and 1924 F, were derived 
were combined in 1920 and 1921. The lines comprising the F, and 
F, of 1924 were not combined until 1922 and 1923. In the latter 
stocks there had been, therefore, more opportunity for selection. An 
examination of the pedigrees shows that the descendants of a single 
row in 1921, No. 1, contributed 45 per cent of the blood of the F , prog- 
enies and 35 per cent of the blood of the F, progenies tested in 1924. 
This particular 1921 row was not represented in the ancestry of 
either the 1923 F, or the 1924 F,. The original crosses from which 
the 1924 F, descended were made in 1920. 
The selfed descendants of row 1, 1921, stood higher than the 
descendants of any other 1921 row in 1922, 1923, and:1924, and it 
