INHERITANCE IN THE BARLEY SPIKE. 9 



INHERITANCE OF LENGTH OF INTERNODES IN CROSSES BETWEEN 



PURE LINES. 



Each cross studied has been considered as a separate family. 

 For convenience, the data from each such family will be discussed 

 separately. In considering crosses, statements will be made as to 

 the number of homozygous and heterozygous forms. Such state- 

 ments can be only relative. Using the variability of the pure lines 

 as a standard, it is assumed that progeny lines of low variability are 

 homozygous, while those of high variability are heterozygous. There 

 is no reasonable doubt of the classification of the extremes, but there 

 is a borderland where the most varying homozygotes may be in doubt. 



FAMILY MANCHURIA (360) X SVANHALS (458). 



The actual F t generation of the cross between Manchuria and 

 Svanhals, which was the basis of later generations discussed in this 

 bulletin, was grown in 1915. A considerable number of crosses 

 between these same pure lines of Manchuria and Svanhals were made 

 in 1917 in the greenhouse at Washington, D. 0. The data for the F x 

 reported in Table II (sec. A) are from this greenhouse seed. On the 

 basis of the coefficient of variability, this F t generation proved 

 no more variable than the parents. 



In 1917 the mean average density in millimeters of the Svanhals 

 parent was 2.53 ±0.01 mm.; of the Manchuria, 3.34±0.01 mm.; and 

 of the F 1; 2.70 ±0.01 mm. There is almost a complete dominance of 

 the dense over the lax form. 



An F 2 generation was grown both in 1916 and in 1918. The means 

 for these two F 2 generations were 2.94 ±0.01 and 2.96 ±0.02 mm., 

 respectively. The variation as determined by the frequency distri- 

 bution and the coefficient of variability was much greater in F 2 than 

 in ¥ 1 or in the parental forms, the coefficient of variability of the Fj 

 generation being 6.30 ±0.30 mm. and of the F 2 generations of 1916 

 and 1918, 10.20 ±0.27 and 11.82 ±0.48 mm., respectively. 



Thirty-two F 3 lines, representing all F 2 types of density, were 

 grown. Thirteen of these F 2 plants appeared to give homozygous 

 progeny in the F 3 generation. The writers recognize that too few 

 plants were grown in F 3 to determine with certainty which forms 

 were homozygous. Eight of these 13 lines were continued in F 4 , and 

 five of these appeared to be homozygous. These results show that a 

 considerable number of the F 2 plants selected bred true in F 3 , although 

 no conclusion as to the actual percentage can be made. 



The five types which proved to be homozygous in F 4 gave mean 

 densities as follows: 378-1, mean 2.57±0.01 mm.; 378-11, mean 

 2.64 ± 0.01 mm. ; 378-14, mean 3.37 ± 0.02 mm. ; 378-23, mean 2.55 ± 

 0.01 mm.; 378-31, mean 2.58 ±0.01 mm. Selection 378-88 gave the 

 highest coefficient of any third-generation line. Two heads were 

 selected which bred true in F 4 for densities near the Manchuria parent. 

 182694°— 20— Bull. 869 2 



