14 
BULLETIN 1354, U. S. DEPARTMENT OF AGRICULTURE 
That competition may have an important effect in comparisons 
under the hill-checking method has been shown U)- This effect 
unquestionably is a disturbing factor in experiments in which an 
attempt is being made to establish absolute differences in productive- 
ness. In experiments like the present, however, competition may be 
helpful in accentuating small differences. It was partly with this 
in mind that the hill-checking method was used. Under the condi- 
tions of the experiment there was little apparent effect except toward 
the lower limits of productiveness of the crosses. Omitting the data 
on the four comparisons for which no probable errors are shown and 
the three in which the yields of the crosses were so low as to be 
clearly beyond the normal range of the crosses as a whole, 70 com- 
parisons remain. The average yield of the checks in these compari- 
sons is 5.39 pounds, and the correlation between the yield of a check 
and the excess yield of a cross is — 0.34 ± 0.07. If the 14 comparisons 
in which the crosses yielded from 1.30 to 2.95 pounds less than the 
checks also are omitted, the correlation between the yield of a check 
and the excess yield of a cross disappears, being only 0.001 ± 0.09. 
The chief interest of this study lies in the. value of the individual 
lines and strains for making productive combinations. This is 
shown more clearly in Table 7, in which the yields of the crosses above 
or below the checks are arranged to show the strain tendencies. The 
individuality of certain lines in combination is very evident. Thus, 
crosses of three of the lines of 5-1- with 2-2-2-4-2-1 are slightly 
superior to the check, whereas the same lines crossed with 2-2-2-4-4-1 
give decidedly inferior yields. Similarly. 7 of the 10 crosses involving 
7-4-1-6-6-2 yielded more than the check, whereas all of the 7 crosses 
involving 7-4-1-7-4-1 yielded less than the check. The lines of the 
10-3-1- strain are the outstanding ones of the experiment. The 
crosses involving these lines have a higher mode and higher mean 
productiveness than any other group. The 10-3-1- lines also have a 
wider range of compatibility, producing one or more high-yielding 
crosses when combined with each line of the other strains. 
Table 7. — Average yield of crosses between self-fertilized lines of corn above or 
below the yield of the corresponding check, arranged to shoic the tendencies within 
lines and strains 
[Data in pounds, showing the difference in yields from 10 hills] 
Parent lines 
5-1-2-2-1-3. 
5-1-2-3-2-2. 
5-1-3-1-1-2. 
5-1-3-3-1-2. 
7-4-1-6-6-2. 
7-4-1-7-4-1. 
7-4-2-1-1-2. 
7-4-2-6-2-2. 
10-3-1-1-1-4 
10-3-1-1-2-2 
10-3-1-2-1-4, or 5. 
10-3-1-2-2-3 
10-3-2-1-2-1 
4 o 
4-0.10+0.72 -2.3S 
1.64| +.27 -1.75+0.32 
1. 49! +. 03 -1. 09 -1. 67 
-2.95! -1.63 
+.17 
-.55 
-.35 
■1.66 
-1.05 
-.45 
-.66 
■1.63 
+0.21 -1.21-0.49 
1.30-1.36; -.90 -.53 
08 1-I.I8 
-1.02 
-.14 
4.451 -.12 -.12 -.32 
-.49+1.651+2.40 +.17+1.26 
! +.51; -.69 
+.94 ' +.83 +.43 
-.17 +.01 
10-3-2-1-3-3 i -5. 67 -1. 50 -5. 52 
44 
+.51 -.55+1.23 ._.. 
-.66+1.20 
+.54 +.31+1.77 
-.54-1.13 -.54 +2. 
-1.17 -.64 -2.38.... 
-.63-1.38 +.44 2 +.37! _. 
.. -0. S3 
.- -.51 
.. -20 
31. 
1 Estimated dry weight. 
10-3-2-1-3-3 X 7-4-1-Arlington. 
