10 . BULLETIN 13896, U. S. DEPARTMENT OF AGRICULTURE 
rows of each border were excluded from the comparisons and were 
planted with crib-selected seed. The same stock of crib-selected 
seed was used as one of the members compared in blocks 3, 4, and 8, | 
thus ie ineiama an opportunity to compare the effect of outside rows 
on yield. 
The yield of the bulk seed of inside rows exceeded that of outside 
rows 1.02+0.28 pounds per section—D+K=8.6. The difference 
in yield per plant is slightly less significant, being 0.028+0.01 
pound—D+HE=2.8. Although not certainly significant, the results 
justify the rejection of the outside rows. 
The possible influence on yield of position in the border, other 
than outside rows, was tested by averaging the differences between 
the mean yield per plant of the second and sixth rows and that 
of row 4, the center row. Rows 2, 4, and 6 were always planted 
with the same kind of seed. The eight sections in each of the eight 
borders afforded 64 comparisons. The result showed the mean 
difference to be 0.0176 +0.0067 pound per plant, the mean of rows 
2 and 6 being the higher. 
The difference is 2.6 times the error; but, since it is in the oppo- 
site direction from that of the comparison of the outside rows with 
rows 3 and 5, this difference is probably the result of chance. 
The results of the 1924 yield test are set forth in Table 4. The 
comparisons are arranged in the order in which they were placed in 
the field. In each comparison the two seed stocks compared are 
designated as first and second members. Thus in column 2 crossbred 
seed (first member).is compared with F, seed (second member). 
TABLE 4.—Comparison of yields of crossbred and recombined selfed strains of maize 
= 7 i 
Crossbred | Crossbred Kote Crossbred | », anq F, |F1and crib 
Items of comparison = andecrib | 
and Fe | and F,4 | ‘gelected. | and F, selected 
| 
| 
Yield per acre (bushels): | 
First member____--..---- | 42.1 41.4 | 38.1 33.3 32.7 34.5 
Second member-.--------- 42.6 | Ay dal 33. 2 35. 7 31.5 23. 6 
First member minus 
yield of second mem- | 
Dera. Se a eee \—. 701.68 |13.0041.02 | 4.70+.56 |—2.82+. 88 -17+.8 /10.86+.86 
Difference+probable | 
error difference_--~--.-- 42 | 12.4 8.4 3.2 | .2 12.6 
Yield per plant (pounds): 
First member minus | | 
second member.------- 006.021 | .094+.012 | 0224. 003 |—. 039+. 009 | . 013+. 009 . 134+. 012 
Difference+probable | 
etrordifferenco.._.)z._- | 28 7.8 7.3 | 4.3 | 1.4 1 
It will be seen that seed from the crossed experiment was compared 
with, each of the four other kinds, and in addition there were two 
borders in which F, seed was compared with crib-selected and F, 
seed. ) 
The results show that the five seed types fall into two groups with 
respect to yield, a high-yielding group comprising the Fy, F,, and 
the crossbred. All the members of this group show significantly — 
higher yields than either the F, or the crib selected, the difference 
ranging from 15 to 45 per cent and each of them over eight times 
the probable error of the difference. ‘4 
In the higher yielding group the F, seed stock resulting from the 
selfed experiment outyielded the crossbred block by 8 per cent in ~ 
a — a ee ee ee ee ae ee 
—_—#_ Ste OO eC OO 
oo 
