Selfing the F2s to produce F3s

The easiest experiment to perform was to let the plants self-pollinate and then keep good records. After scoring his 556 F2 seeds he took the 315 that were round and yellow and planted them in one part of his garden. The plants that grew were allowed to self-pollinate. Of the 315 round and yellow seeds planted, 301 plants matured and produced seed. The seed produced was the F3 generation. At harvest, Mendel needed to exercise the utmost care. Each F2 plant was handled separately. The seeds from the plant were harvested and Mendel then scored the F3 seeds that came from the same F2 plant. This can be referred to as F2.3 data and the table below summarizes his complete experiment using all of the F2 phenotypes.

Mendel's F2 data supported his principle of independent assortment. There were four different types of round yellow F2s based on the kinds of progeny they could produce or their breeding behaviors. Based on the F3 progeny produced, the F2 genotype was deduced. For example if a round, yellow seed gave all round progeny it must have the genotype RR. If it gave both round and wrinkled it was Rr. Furthermore, the numbers of F2 plants with each breeding behavior were in agreement with what was expected with independent assortment. There were four times as many round and yellow F2s that gave all four phenotypes of F3 seeds (138) compared to the round and yellow F2s that were true breeding (38). Overall, there were nine types of breeding behaviors demonstrated in the F2s demonstrating that there were nine F2 genotypes. In all cases, the fractions observed in the F2s were in agreement to the principle of independent assortment. Mendel's well planned experiment provided a convincing demonstration that genes behaved in this predictable manner.

The only thing better than performing an experiment that shows you were right about a new hypothesis is performing two experiments that show that you were right. That is what Gregor Mendel did! In his second experiment he crossed dihybrid F1 plants with homozygous recessive plants in a test cross. This type of cross is named because the geneticist wants to perform a cross that will test or reveal the genotype of an organism. Therefore a test cross is usually made between an organism with a dominant trait and a partner with a recessive version of this trait. Mendel performed the RrYy X rryy testcross and the expected progeny are shown in the Punnett square below:

The observed result closely matched the expected. The testcross experiment provides additional support for the principle of independent assortment.

Mendel established a rigorous precedent for using carefully planned multi-generation experiments to reveal the principles that governed trait inheritance. The beauty of Mendel's accomplishments is that both the principles and his experimental approach can be applied to understanding the genetic control and inheritance of traits in many kinds of organisms still today.