As we have seen, it is from the genotypical not the phenotypical constitution that an offspring is derived and what a given form will bring forth depends then on its genotype.
Crosses With More Than Two Pairs of Characters.—In crosses in which more than two pairs of contrasted characters are involved the underlying principles are in no way different, only with each pair of additional characters there is, of course, a greater number of possible combinations. Thus with three pairs of characters there will be eight different classes of gametes in each sex and consequently sixty-four possible combinations in mating, giving eight different phenotypes in the proportion of 27:9:9:9:3:3:3:1. The largest class manifests the three dominant characters; the smallest class, the three recessives; the three classes in the proportion of 9 each exhibit two dominant and one recessive characters; and those in the proportion of 3 each display two recessive and one dominant characters.
THE QUESTION OF BLENDED INHERITANCE
We come now to certain types of inheritance in which there seems to be a true fusion or blend of the contributions from the two parents, the intermediate condition apparently persisting in subsequent generations without segregation. Numerous cases of blended inheritance have been cited in earlier literature of heredity, but as our knowledge of genetics has progressed many experimental breeders have come to believe that the blends in such cases are apparent rather than real and that the phenomena can be best explained on a non-blending unit-character basis, just as we would explain ordinary Mendelian phenomena.
Nilsson-Ehle’s Discoveries.—To get their point of view we may review certain experiments on wheat made by Nilsson-Ehle, together with their Mendelian interpretation. Nilsson-Ehle found that a certain brown-chaffed wheat when crossed with a white-chaffed strain yielded a brown-chaffed hybrid, apparently in accordance with the simple principle of Mendelian dominance. But these heterozygous brown-chaffed individuals did not in turn give the expected ratio of 3:1 in the F2 generation but a ratio of 15 brown to 1 white, and furthermore the browns were not all of the same degree of brownness. To be exact, from fifteen different crosses of the strains he obtained 1,410 brown-chaffed and 94 white-chaffed plants.
This apparent anomaly in segregation was easily explained, however, when it was finally figured out that there were really two independent determiners for brown color, either of which alone could produce a brown individual, but when combined produced individuals of correspondingly deeper shades of brown. In such a case then Nilsson-Ehle discovered that he was dealing merely with a Mendelian dihybrid where two different determiners B and B′ and their respective absences b and b′ are involved. The original brown wheat had both B and B′ and the original white b and b′. The formula for the F1 heterozygote was therefore BbB′b′. The four possible types of gametes for male and female are BB′, Bb′, bB′, bb′, and the tabulation would be as follows:
| BB′ | Bb′ | bB′ | bb′ | |
| BB′ | BBB′B′ | BBB′b′ | BbB′B′ | BbB′b′ |
| Bb′ | BBB′b′ | Bb′Bb′ | BbB′b′ | Bbb′b′ |
| bB′ | BbB′B′ | BbB′b′ | bbB′B′ | bbB′b′ |
| bb′ | BbB′b′ | Bbb′b′ | bbB′b′ | bbb′b′ |
It will be observed that there are more brown determiners in some combinations than others. For instance one of the sixteen contains four such determiners, viz., B, B′, B, B′, four contain three determiners, six contain two, four contain only one, and one contains none. Thus all but one of the sixteen contain at least one determiner and will therefore be brown in color but the depth of color will depend on the number of brown determiners in a given individual. This is more graphically represented in Fig. 19, [p. 90]. The largest number of similar individuals, six in all, contain two determiners each and represent an intermediate “blend” between the original brown-chaffed and white-chaffed strains. The deeper and the lighter browns due to more or fewer determinants in an individual would if one did not know the units in this case look like the fluctuations around this average which we might expect in a blend.
