A COMPARISON OF MAIZE-BREEDING METHODS 7 
The comparison was made by planting the seed stocks in alternate 
rows. There were twelve 400-foot rows, and in harvesting each row 
’ was subdivided into four sections, making 24 comparisons. 
There was a slight positive correlation between the number of 
plants per section and the yield per plant, indicating that conditions 
favorable to germination were likewise favorable to growth. Yield 
per plant therefore seems a more suitable measure of production 
than yield per section. 
The 12 rows occupied one irrigation border and part of another. 
There were thus three outside rows. Two of these were planted 
with F, seed; the other with seed from the crossed block. It is fur- 
ther necessary to consider the possibility of a general trend in the 
fertility of the field that would favor one or the other kind when taken 
in pairs. The uncorrected mean difference (CB-F,) of the entire 24 
pairs was 0.22 +0.02 pound per plant. Excluding all pairs involv- 
ing outside rows, the | 
mean difference was ‘2% 1922 1921 1220 I919 1918 
0.205+0.029. Drop- td I.———=}/. 2/ 
ping the first row and 
subtracting the yield 
per plant of the F, e 
from the _ preceding 
instead of the follow- 
ing adjacent section, <6 
to disclose the effect ON 
of a possible trend in 4 
the field, the mean dif- eg 
ference was 0.24+ _ 
0.025. All of these 
Measurements are in 2/ Fs 
substantial agree- ee Ys 
ment, and the differ- 
ence is from 7 to 10 
times the probable [sy i? 
error. Fic. oo g Reprcsentative pedigree of progenies in the F»> seed stock of 
, maize tested in 1923. The numbers indicate individual plants. 
The mixed sample Selfed generations are shown by double lines. F=40.6 
of F, seed that was 
compared with crossed seed in 1923 comprised 10 ears. These were 
derived from five F, hybrids grown in 1922. These five F, hybrids 
came from six selfed lines grown in 1921; five of these selfed lines 
were sister progenies descending from selfed plants in a single line in 
1920. The sixth selfed line was unrelated, having descended from a 
different original ear. 
Sample pedigrees of the crossbred and F, seed stocks grown in the 
1923 yield test are shown in Figures 1 and 2. 
YIELD COMPARISONS OF THE TWO BREEDING METHODS IN 1924 
In the season of 1924 a more thorough trial was conducted. The 
yields of five seed stocks were compared. These were: (1) Seed 
from the crossbred experiment designated CB, (2) composite sample 
of F, crosses between selfed lines designated F,, (3) composite 
sample of crosses between F, lines designated F,, (4) composite 
sample of crosses between F; lines designated Fy, and (5) crib-selected 
