Mendel's Dihybrid Cross
Mendel crossed pure breeding plants with round seeds and yellow albumen to pure breeding plants with wrinkled seeds and green albumen. The F1 plants all had found seeds and yellow albumen and Mendel predicted that they would be heterozygous for both traits (RrYy).
He then did a self-cross of the F1 plants to produce the results shown as the F2. He predicted that the plants with round seeds and yellow albumen would be of four genotypes in the specific ratio of 1 RRYY : 2 RRYy : 2 RrYY : 4 RrYy. And he made similar predictions for the plants with round seeds and green albumen and the ones with wrinkled seeds and yellow albumen.
In order to test his prediction of more than one genotype within a phenotypic category in the F2, he allowed the F2 plants to self-cross and he recorded how many of the plants bred true or showed segregation for one or both of the characters. This is shown in the F3 data. The observed numbers of phentoypic types matches the predicted ratios.