STUDIES ON SELF-STERILITY. IX. THE BEHAVIOR OF CROSSES BETWEEN SELF-STERILE AND SELF-FERTILE PLANTS

Abstract

1. N. Langsdor$ii contains a fertility factor-a factor evoking rapid pollen-tube growth in all combinations-which is allelomorphic to the sterility factors found in N. Sandere. It is designated S,. 2. When plants of N. Sandera, homozygous for known self-sterility factors, are crossed reciprocally with plants of the self-fertile species N. Langsdorfiii, the F1 plants SjSn are self-fertile. 3. SfS, plants, unless they contain modifying factors, are sterile to the pollen of SnSn plants and to the S, pollen of S,S, plants. In the absence of other variables, therefore, S,Sn plants selfed yield Ft populations consisting of equal numbers of plants belonging to classes SfSf and SI&, both of which are self-fertile. There is but one known exception to this rule of behavior for S factors. A factor discovered by ANDERSONan d DE WINTONin N. alata and designated SF by them has the property of inhibiting the development of Sj pollen tubes. Under optimum conditions it may be expected that some seed would be produced by selfing SfS, plants and that both self-fertile and self-sterile offspring would be produced. 4. Plants S,S, crossed with plants S$, will produce self-fertile and selfsterile plants in the ratio of 3 to 1. The self-sterile plants will be incompatible with each other in reciprocal crosses. 5. Genes other than S factors may exist which will so modify the effects of the primary S factors that a plant S,SnXlX2 selfed could produce both self-fertile and self-sterile offspring (see KAKIZAK1I9 30). It is not improbable, therefore, that the results which I reported in 1919 for crosses between N. Langsdorfii and N. Sandere-where selfed F1 plants gave both self-fertile and self-sterile plants in the Fz generation-were correct. The constitutions of the pIants used in these earlier experiments were unknown. Many modifying factors may have been present. 6. The effects of three modifying factors independent of the S locus- A 1, A 2, and A 3-are described. These three factors have no effect on the action of SI, and but a slight tendency to nullify the action of S3 in incompatible matings. A tends slightly to nullify the action of Sz in incompatible matings. A 2 tends markedly to nullify the action of S2 in incompatible matings and appears to have alethal action on the zygote which contains Ss3. AI, the existence of which is more questionable, appears to have no lethal effect on zygotes which contain SS,, though otherwise it behaves like A 2. 7. N. Sandere, in so far as it was tested, seems to be homozygous for a factor aa at the A locus. The N. Sandere plants used have been closely bred for about ten generations and should be homozygous for most factors. 8. Plants carrying two S factors other than Sf were male sterile in the presence of cytoplasm derived largely from N. Langsdorfii. They produced normal pollen and were self-sterile (apart from the effect of the A factors) when in the presence of cytoplasm derived largely from N. Sandera 9. The inheritance of certain factors C and P affecting corolla color is discussed. C lies in the same chromosome as S. The crossover values in crosses between N . Langsdorfii and N. Sandera appear to be of the same order of magnitude as in N. Sandereat least in megasporogenesis. 10. Anther color is due to the reaction of factors C,P, andD. CP, CD and CDP give dark anthers. Plants of these constitutions have dark anthers with Langsdorfii cytoplasm, and light anthers with Sandere cytoplasm. 11. Blue pollen is due to three supplementary factors acting in combination, CDB. Sandere cytoplasm, when in combination with Sf and with certain A factors (especially As), is unfavorable to the development of blue pollen. 12. Green in the corolla is due, primarily, to a dominant factor G in N. Langsdorfii. There is evidence that certain N. Sandere plants contain a factor 2 which is favorable to the production of chlorophyll in the corolla when Sf is present with N. Langsdorfii cytoplasm and unfavorable to the production of chlorophyll when two S factors other than Sf are present and the cytoplasm derives largely from N. Sandera. A third factor X , coming from N. Sandere, appears to inhibit the action of G, but is effective only with Sf and in N. Langsdorfii cytoplasm. A fourth factor for green, G,, may exist. G, appears to act only when two N. Sandere sterility factors other than Sf are present with N . Sandere cytoplasm. 13. Certain combinations of factors, therefore, produce different phenotypic effects depending on the derivation of the cytoplasm. 14. Certain combinations of factors are viable or non-viable depending on the derivation of the cytoplasm. 15. N. Sandere is treated as a species. It is a horticultural combination of N. Forgetiana and N. alata, and behaves as do the parental species. The distributions of the wild species N. Forgetiana, N. alata, and N. Langsdorsi are not well known, and one can not say whether they overlap or not. But N. LangsdorBi is certainly distinct from the other two on any basis except sterility in hybrid combinations. Nevertheless, though the distribution of genes is normal in the hybrid combination N. Langsdorsi X N. Sandere, the two species have come to differ in such a way that many combinations of factors which are successful in the pure species are unsuccessful in the cytoplasm of the other species.