第225章

Primula officinalis (P. veris) (Cowslip) ... 69Primula elatior (Oxlip) .................... 27Primula acaulis (P. vulgaris) (Primrose) ... 60Further, the plants produced by the illegitimate method of fertilisation showed, e.g. in P. officinalis, a decrease in fertility in later generations, sterile pollen and in the open a feebler growth. (Under very favourable conditions (in a greenhouse) the fertility of the plants of the fourth generation increases--a point, which in view of various theoretical questions, deserves further investigation.) They behave in fact precisely in the same way as hybrids between species of different genera. This result is important, "for we thus learn that the difficulty in sexually uniting two organic forms and the sterility of their offspring, afford no sure criterion of so-called specific distinctness" ("Forms of Flowers", page 242): the relative or absolute sterility of the illegitimate unions and that of their illegitimate descendants depend exclusively on the nature of the sexual elements and on their inability to combine in a particular manner. This functional difference of sexual cells is characteristic of the behaviour of hybrids as of the illegitimate unions of heterostyled plants. The agreement becomes even closer if we regard the Primula plants bearing different forms of flowers not as belonging to a systematic entity or "species," but as including several elementary species. The legitimately produced plants are thus true hybrids (When Darwin wrote in reference to the different forms of heterostyled plants, "which all belong to the same species as certainly as do the two sexes of the same species"("Cross and Self fertilisation", page 466), he adopted the term species in a comprehensive sense. The recent researches of Bateson and Gregory ("On the inheritance of Heterostylism in Primula"; "Proc. Roy. Soc." Ser. B, Vol. LXXVI. 1905, page 581) appear to me also to support the view that the results of illegitimate crossing of heterostyled Primulas correspond with those of hybridisation. The fact that legitimate pollen effects fertilisation, even if illegitimate pollen reaches the stigma a short time previously, also points to this conclusion. Self-pollination in the case of the short-styled form, for example, is not excluded. In spite of this, the numerical proportion of the two forms obtained in the open remains approximately the same as when the pollination was exclusively legitimate, presumably because legitimate pollen is prepotent.), with which their behaviour in other respects, as Darwin showed, presents so close an agreement. This view receives support also from the fact that descendants of a flower fertilised illegitimately by pollen from another plant with the same form of flower belong, with few exceptions, to the same type as that of their parents. The two forms of flower, however, behave differently in this respect. Among 162 seedlings of the long-styled illegitimately pollinated plants of Primula officinalis, including five generations, there were 156 long-styled and only six short-styled forms, while as the result of legitimate fertilisation nearly half of the offspring were long-styled and half short-styled. The short-styled illegitimately pollinated form gave five long-styled and nine short-styled; the cause of this difference requires further explanation. The significance of heterostyly, whether or not we now regard it as an arrangement for the normal production of hybrids, is comprehensively expressed by Darwin: "We may feel sure that plants have been rendered heterostyled to ensure cross-fertilisation, for we now know that a cross between the distinct individuals of the same species is highly important for the vigour and fertility of the offspring."("Forms of Flowers", page 258.) If we remember how important the interpretation of heterostyly has become in all general problems as, for example, those connected with the conditions of the formation of hybrids, a fact which was formerly overlooked, we can appreciate how Darwin was able to say in his autobiography: "I do not think anything in my scientific life has given me so much satisfaction as making out the meaning of the structure of these plants." ("Life and Letters", Vol. I. page 91.)The remarkable conditions represented in plants with three kinds of flowers, such as Lythrum and Oxalis, agree in essentials with those in Primula. These cannot be considered in detail here; it need only be noted that the investigation of these cases was still more laborious. In order to establish the relative fertility of the different unions in Lythrum salicaria 223 different fertilisations were made, each flower being deprived of its male organs and then dusted with the appropriate pollen.

In the book containing the account of heterostyled plants other species are dealt with which, in addition to flowers opening normally (chasmogamous), also possess flowers which remain closed but are capable of producing fruit. These cleistogamous flowers afford a striking example of habitual self-pollination, and H. von Mohl drew special attention to them as such shortly after the appearance of Darwin's Orchid book. If it were only a question of producing seed in the simplest way, cleistogamous flowers would be the most conveniently constructed. The corolla and frequently other parts of the flower are reduced; the development of the seed may, therefore, be accomplished with a smaller expenditure of building material than in chasmogamous flowers; there is also no loss of pollen, and thus a smaller amount suffices for fertilisation.