The Cytological Aspects of the Determination of Sex in the Diœcious Forms of Lychnis

The present research is part of a more general inquiry into the conditions of sex determination in the genera Lychnis and Silene. This paper consists of an account of the cytology of the diœcious forms of Lychnis with special reference to sex chromosomes. Lychnis Flos-cuculi is referred to at times for the sake of comparison. The somatic number of chromosomes in L. dioica (agg.), as well as in L. Flos-cuculi L. Flos-Jovis, and Silene pendula, is twenty-four. The hybrid L. alba x L. dioica behaves in a perfectly regular manner and is practically indistinguishable cytologically from its parents. In Lychnis dioica (agg.) the pollen mother cells show a particularly close synizetic knot from which the spireme never opens out fully. Just before the second contraction the spireme becomes differentiated into thicker and thinner portions of which the thicker clearly represent the chromosomes. The thread becomes much shorter and stouter and, as the contraction figure, opens out, the chromosomes separate off in pairs forming typical bivalents lying either crossed or parallel. The reduction is thus typically telosynaptic and there is no trace of a split thread at any stage. As the bivalents contract, one pair is seen to be larger than the rest and to consist of two unequal portions, both of which are larger than the other chromosomes. At diakinesis, the bivalents become completely dissociated, once more giving twenty-four separate chromosomes. L. Flos-cuculi differs in haying ring-shaped chromosomes, both at diakinesis and on the equator of the heterotype spindle. In L. dioica (agg.), the large pair of chromosomes lies at the periphery of the metaphase plate and is seen to consist of one very large hooked member and one smaller, more pear-shaped, chromosome. Anaphase figures show the larger chromosome in the form of a cross, as seen from above, suggesting a double structure. There is no difficulty in distinguishing which member of the unequal pair is present in any given daughter plate. In the interkinesis there are two points worthy of note, (1) Several nucleoli are present, in place of the usual single one typical of all other stages. (2) The twelve chromosomes separate completely into halves, thus giving twenty-four bodies lying within the membrane, just as in diakinesis. At the ensuing metaphase the X and Y chromosomes are again most distinct, especially when the two spindles lie parallel. Megaspore development is briefly described. Meiosis is similar to that found in the male plant, except in relation to the large chromosomes. The large chromosomes are equal in size, and by careful comparison of size and shape, are shown to correspond to the small one of the unequal pair in the male. The Y chromosome is thus larger than the X, quite contrary to expectations based on the condition in animals. It is also probably double in nature though it does not separate into two parts, as far as has been observed. A brief review of the literature shows that in Rutnex there is a pair of Y chromosomes ; in Humulus and Vallisneria a double structure is presumed to be the X element, but on the basis of Lychnis may possibly turn out to be the Y instead. The cytological results above are shown to corroborate the results of the experimental work of Shull and Correns on the same forms.