224 



The Journal of Heredity 



While one may see in this table a 

 slight tendency of the yellow berry seed 

 to produce a higher percentage of yel- 

 low berry in the offspring, the ditifer- 

 ences are so slight and the regression 

 so strong that one may easily (piestion 

 its signiticance. A comparison of 

 Tables 111 and iX is sufhcient to con- 

 vince one immediately that the heredity 

 of vellow berry is at least something 

 wholly ditYerent from that of true hard- 

 ness and softness as exhibited in 'i'able 

 III. The study of the diffuse type of 

 opaqueness (softness) has clearly 

 shown that this character (though 

 quantitative and, hence, difficult of ex- 

 act separation into genetic classes) is 

 strictly hereditary. It may, moreover, 

 be pointed out that pure lines of true 

 soft wheats (diffuse type) such as So- 

 nora. Algerian Red Bread, and Early 

 Baart are 100/r opaque'- every year, 

 and never exhibit environic variations 

 in hardness such as is shown by pure 

 lines of hard wheats, examples of which 

 are given in Table VI. 



i)i:i'k\i)i:nce ox kxvikoxmkxt 



The great dependence of yellow 

 berry on environic conditions and its 

 almost complete regression from the 

 selection of extreme variants, suggest 

 strongly that this is quite largely a 

 l)hysiological response. True, a tend- 

 ency toward greater or less sensitivity 

 of this response may be inherited and 

 it may be possible that those grains 

 in a head carrying genetic factors for 

 greater sensitivity would be the first 

 to respond to environic conditions 

 tending to favor the yellow berry 

 character whereas those same grains 

 under conditions favoring greater hard- 

 ness would have remainefl perfectlv 

 translucent throughout. This would 

 account for the slight but (|uite uniform 

 differences exhibited in tables VHI and 

 IX. Convincing ])roof that the tend- 

 dency toward the jjroduction of yellow 

 berry in hard wheats is hereditary is 

 shown in an exi)erimcnt with 145 



pure races of turkey wheat extending 

 through the years 1914. 1915 and 1916. 

 Here the correlation of the percentage 

 of vellow berry in 1914 with the same 

 in 1915 was +^57% ±4%, whereas that 

 between 1915 and 1916 was +33% 

 ±5%, and between 1914 and 1916 was 

 +41% ±5%. These correlations being 

 more than six times their probable 

 errors are certainly significant. 



Among the F, plants originating 

 from the ]\Iacaroni-Sonora cross, there 

 were 65 which had grams recorded as 

 all hard ( or with a greater or less per- 

 centage of yellow berry), and which 

 gave all hard or "yellow berry" off- 

 spring, /. c, produced no plants having 

 any intermediate or soft grains (dif- 

 fuse type).-* Many of these, however, 

 contained a greater or less percentage 

 of "yellow berries." The correlation 

 calculated between the percentage of 

 yellow berries in these parents and the 

 average percentage of yellow berries in 

 their offspring was +51% ±6%. 



KECRESSIOX IN PARENTAL STRAINS 



The strong regression from extreme 

 selections, but still quite definite in- 

 heritance of the tendency to produce 

 yellow berry, is perhaps better shown 

 by arranging the parental strains or 

 mother plants in accordance with the 

 |)ercentage of yellow berry they contain 

 and then showing the distribution and 

 average of their respective offspring 

 as is (lone in the liable X. 



The arrangement here shown was 

 made so as to divide the parental 

 strains or plants, as nearly as possible, 

 into three equal groups. While the dif- 

 ferences between the groups are not 

 large, the numbers involved and the 

 uniformity of results certainly render 

 them significant. 



COXCLUSIOXS 



1. The hardness of a wheat is deter- 

 mined by the solidity of the grain, and 

 this, in turn, by the nature and relative 

 proportions f)f gluten and starch in the 

 endosperm. 



2. When the ratio of gluten to starch 



* Like Plate I, FiR. b. 



» Like Plate III, Figs, b and r. 



