FACTORS FOR YIELD AND QUALITY OF SPRING WHEAT 5 
casin is located under semiarid or average dry-land conditions in the 
Judith Basin and has a normal annual precipitation of 16.51 inches. 
Havre is under marginal dry-farming conditions in the so-called 
"Triangle" section and has a normal annual precipitation of but 
13.71 inches. Thus, three distinct environmental conditions are 
represented. 
The Marquis X Hard Federation and reciprocal crosses here studied 
were made, at the request of the senior writer, by V. H. Florell (8) at 
Chico, Calif., in May, 1921. The F t material was grown at Davis, 
Calif., in 1922. The Marquis 9 X Hard Federation <? crosses 
were given the hybrid number 21202 and the Hard Federation 9 X 
Marquis $ crosses number 21203. The first two numerals, 21, 
stand for the year 1921, in which the crosses were made. The 
numbers 202 and 203 were assigned to these particular crosses. 
The F x material was divided. Families 21202 Bl, B2, and B3 and 
21203 Al, A2, and A3 were sent to Bozeman, Mont., and families 
21202 Al, A2, and A3 and 21203 Cl, C2, and C3 were sent to Moc- 
casin, Mont., where the 12 F 2 families were grown in 1923. 
In 1923 the hybrid plants at both Bozeman and Moccasin were 
definitely spaced at 3-inch intervals in rod rows 1 foot apart. Unfor- 
tunately, the parent varieties were not sown at Moccasin. At 
Bozeman the Hard Federation parent was sown identically with 
the hybrids, but the Marquis parent was sown two days later and 
under slightly different conditions. 
In 1924 the F 3 material was grown at Bozeman, Moccasin, and 
Havre in nursery rows 1 foot apart. At Bozeman 70 kernels were 
spaced 3 inches apart in each row. At Havre the seeding was done 
in a similar manner, but the number of kernels varied somewhat 
because of the limited quantity of seed available from some of the 
F 2 selections. At Moccasin the 1924 sowings were not definitely 
spaced, but 2.3 grams of seed was used in each of the rows, which 
were 16 feet long and 1 foot apart. Checks of the parent varieties, 
sown in the same manner and at the same time as the F 3 hybrids, 
alternated in every tenth row at the three stations. The F 3 strains 
grown at Bozeman and Havre in 1924 were from F 2 plant selections 
from Bozeman in 1923. The F 3 strains grown at Moccasin in 1924 
were from F 2 selections made at that station the previous year. 
The material thus grown constitutes the basis for the following 
study. Probable errors for numbers of individuals where only two 
classes are concerned were obtained from tables of probable errors 
of Mendelian ratios prepared in the department of plant breeding 
o f Cor nell University, Ithaca, N. Y., from the formula 0.6744898 
-yjn p q, in which n is the total number of individuals and p and q 
the percentages corresponding to the ratios involved. Deviations 
from the expected which are less than three times their probable 
errors are not here considered significant. 
When more than two classes are concerned, the goodness of fit to 
the expected ratio is determined by the X 2 method in which 
io — c) 2 
X 2 = 2 From X 2 the value of P is determined by interpolation 
from Elderton's table given by Pearson (14) for determining this 
value. The values of P usually are less than 1, which value repre- 
sents a perfect fit. Ratios having deviations with a value of P lower 
than 0.043 are not here considered significant. 
