416 
Journal of Agricultural Research 
Vol. XI, No. 9 
For the tests of replications with barley there were 15 plots available 
in the Farm Crops test and 9 in the Plant Breeding test. 
The Farm Crops work showed a lower variability in all cases than for 
the Plant Breeding results. For the Farm Crops test two replications 
proved to be only slightly more valuable than a single test. In the cen¬ 
tral row tests three and four replications seem to be about equally reliable, 
while six replications reduce the error about 50 per cent as compared with 
three replications. The use of 3-row plots gave a lower error as deter¬ 
mined by the coefficient of variability and the standard deviation than 
the use of central rows alone. 
For the Plant Breeding tests central rows or 3-row plots show about 
the same standard deviation. Two or three replications reduce the error 
nearly 40 per cent. 
The results for replication studies in winter wheat are given in Table 
XIV. In considering these results it should be noted that both the Plant 
Breeding and Farm Crops tests were made on the same soil series. 
Table XIV .—Value of replications of central rows , 3-row plots , and border rows of 
Turkey winter wheat (Minnesota 52Q) used as a check in the Farm Crops and Plant - 
Breeding tests 
Designation of test. 
Method of test. 
Num¬ 
ber of 
tests. 
Farm Crops variety test.. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Plant Breeding Nursery.. 
Do.. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do... 
Single central rows 
2 replications. 
3 replications. 
4 replications. 
6 replications. 
9 replications...... 
3-row plots.. 
2 replications. 
3 replications. 
4 replications. 
6 replications. 
9 replications. 
Single central rows. 
2 replications. 
3 replications. 
4 replications. 
6 replications. 
12 replications. 
3-row plots. 
a replications. 
3 replications. 
4 replications. 
6 replications. 
12 replications. 
41 
20 
13 
10 
6 
4 
4 i 
20 
13 
10 
6 
4 
72 
36 
24 
18 
12 
6 
72 
36 
24 
18 
12 
6 
Mean. 
29.1 ±0. 504 
29.6± .593 
29. id: . 406 
29.4± *469 
28. 5± -666 
2S. 8± . 762 
28. 4± .403 
28. 4± .338 
28. s± .393 
28. 4± .397 
28.1 ± . 774 
28. o± .661 
27* S± *370 
27 -.335 
27. 4± .346 
27. 6± .238 
27 * 3 ± *391 
27- 3 ± *333 
27. o± .327 
27. idb .310 
27. 2dr .304 
27. idh .234 
27- idh .323 
27. 2± .333 
Standard 
deviation. 
4.78d:0.356 
3 * 93 ± * 4 i 9 
2.17± . 287 
2. 2o± . 332 
2.42 i • 471 
2. 26dh .539 
3* 83 dr . 285 
2. 24dr . 239 
2. lodr . 278 
1. 86± . 281 
2 . i8dr * S 47 
1.96± .467 
4 - 65 ± • 261 
2. 98dr .237 
2-si± .244 
1.49± • 169 
a.oidr . 277 
1. 21 dr . 236 
4* I2dr .232 
2. 76dr .219 
2. 2i± . 215 
l. 47 ± *165 
I. 66dr .229 
I. 21 dr . 236 
Coefficient of 
variability. 
16.43 ± 1. 256 
13. 28dri. 441 
7 * 4 6 ± -987 
7. 48dri. 128 
8. 49± I. 653 
7. 93 ±1.891 
I3.49il.023 
7. 89dr . 841 
7 * 37 ± -975 
6. ss± .988 
7. 76d:i. 511 
7. oodri. 669 
l6.9l± .979 
10.8odr .868 
9. i6± . 892 
5* 43 dr .610 
7*36d:i. 013 
4- 43± *863 
15. 26 dr . 875 
IO. iSd: .817 
8. i6dr . 794 
5- 42± *609 
6. 13 dr . 844 
4 * 45 dr .866 
There is very little difference for the Farm Crop tests between the 
variability as determined by central rows or 3-row plots with the excep¬ 
tion of one or two replications, which are lower in the 3-row plot than 
for the central row tests. Three replications appear to be about as 
reliable as any larger number. 
The Plant Breeding test shows no significant difference in the varia¬ 
bility as determined by the use of 3-row plots or central rows for any 
number of replications studied. Four replications for both central rows 
