Cytological Studies in the Allium Paniculatum Group

N the genus Alliurn there occasionally occur in nature forms with I different chromosome numbers within the same species or within a group of closely related species.In the majority of cases these then form a polyploid series, the most common case being the occurrence of diploids and tetraploids.However, triploids also occur in nature and can by means of vegetative propagation acquire a wide geographical distribution.In the Alliuzn group, which is the subject of treatment in the present paper, the paniculatum group, I have up to now found only euploid forms in nature, viz. 3 diploid species, A. paniculatum, A. pulchelluni and A. flavum, 1 triploid species, A. carinatum, and 1 tetraploid, A. oleraceum.The somatic chromosome numbers of these species are multiples of 8, the different forms thus having 2n == 16, 24 and 32 respectively.Morphologically, these species are closely related and clearly distinguished from other species of Allium, and therefore it is practical to combine them into one unit, the panicnlatum group.The cytology of A. carinatum and A .oleraceum has been described in an earlier paper (LEVAN, 1933).Mainly on cytological grounds, chromosomal morphology, meiotic pairing, etc., 1 put forward the suggestion that these two polyploid species were links in one and the same polyploid series, the diploid members of which were to be found among the closely allied diploid species A. paniculatum, A. pulchelluni and A. flavuni.I emphasized, however, that experimental results would be necessary before this hypothesis could be anything but a conjecture.During the following years I have also been carrying on experiments for the purpose of testing this hypothesis.I have now succeeded in obtaining definitive proofs that the hypothesis, if not entirely wrong, is nevertheless insufficient.Jn a short preliminary paper (LEVAN, 1936c) I have already published the gist of these investigations: A. paniculatum has by