抄録
N. longiflora (n=10), N. plumbaginifolia (n=10), N. repanda (n=24) and N. suaveolens (n=16) were crossed with N. Sanderae as the pollinator. The crosses N. longiflora×N. Sanderae, N. plumbagi- nifolia×N. Sanderae and N. suaveolens×N. Sanderae gave some seeds.
In external characters, the F1 longiflora-Sanderae and F1 plumbaginifolia-Sanderae are intermediate between the respective parents but the flower colors are red, resembling the fathers. F1 longiflora-Sanderae was examined by Christoff (1928) and Isakovich (unpublished, after Kostoff 1941-43), but F1 plumbaginifolia-Sanderae has never been reported, so far as I know. The morphology of F1 suaveolens-Sanderae agrees with that of Kostoff's description of the same hybrid.
At diakinesis in the PMC's of F1 longiflora-Sanderae, the chromosome conjugation 9II+1I, 1III+8II and 1III+7II+2I were most frequent. At the first metaphase the chromosome conjugation 9II+1I was mostly observed, followed by the conjugations 1III+7II+1I, 1III+8II and 8II+3I, in the cited order. The chromosome conjugations agree with Isakovich's findings in the same hybrid (after Kostoff 1941-43). The meiotic chromosome behaviour of F1 plumbaginifolia-Sanderae is the same as in F1 longiflora-Sanderae. According to the observations on the meiosis in the F1 hybrids longiflora-Sanderae and plumbaginifolia-Sanderae, it is assumed that there are some homologous or semihomologous chromosomes between the Sanderae genome on one hand and the longiflora or plumbaginifolia genome on the other.
At the first metaphase of F1 suaveolens-Sanderae, 1-7 bivalents, mostly 3-4, were found. Kostoff (1941-43) observed 0-4 bivalents in the same hybrid. The cause of the difference between the author's and Kostoff's observations has not been studied. And also it was undetermined whether the chromosome conjugations in this hybrid were caused by intergenomic or intragenomic affinity.
