It is remarkable that this book, the result of eleven years of experimental work, owed its origin to a chance observation. My father had raised two beds of Linaria vulgaris—one set being the offspring of cross and the other of self-fertilisation. The plants were grown for the sake of some observations on inheritance, and not with any view to cross-breeding, and he was astonished to observe that the offspring of self-fertilisation were clearly less vigorous than the others. It seemed incredible to him that this result could be due to a single act of self-fertilisation, and it was only in the following year, when precisely the same result occurred in the case of a similar experiment on inheritance in carnations, that his attention was "thoroughly aroused," and that he determined to make a series of experiments specially directed to the question.
The volume on Forms of Flowers was published in 1877, and was dedicated by the author to Professor Asa Gray, "as a small tribute of respect and affection." It consists of certain earlier papers re-edited, with the addition of a quantity of new matter. The subjects treated in the book are:—
(i.) Heterostyled Plants.
(ii.) Polygamous, Diœcious, and Gynodiœcious Plants.
(iii.) Cleistogamic Flowers.
The nature of heterostyled plants may be illustrated in the primrose, one of the best known examples of the class. If a number of primroses be gathered, it will be found that some plants yield nothing but "pin-eyed" flowers, in which the style (or organ for the transmission of the pollen to the ovule) is long, while the others yield only "thrum-eyed" flowers with short styles. Thus primroses are divided into two sets or castes differing structurally from each other. My father showed that they also differ sexually, and that in fact the bond between the two castes more nearly resembles that between separate sexes than any other known relationship. Thus for example a long-styled primrose, though it can be fertilised by its own pollen, is not fully fertile unless it is impregnated by the pollen of a short-styled flower. Heterostyled plants are comparable to hermaphrodite animals, such as snails, which require the concourse of two individuals, although each possesses both the sexual elements. The difference is that in the case of the primrose it is perfect fertility, and not simply fertility, that depends on the mutual action of the two sets of individuals.
The work on heterostyled plants has a special bearing, to which the author attached much importance, on the problem of the origin of species.[288]
He found that a wonderfully close parallelism exists between hybridisation (i.e. crosses between distinct species), and certain forms of fertilisation among heterostyled plants. So that it is hardly an exaggeration to say that the "illegitimately" reared seedlings are hybrids, although both their parents belong to identically the same species. In a letter to Professor Huxley, given in the second volume of the Life and Letters (p. 384), my father writes as if his researches on heterostyled plants tended to make him believe that sterility is a selected or acquired quality. But in his later publications, e.g. in the sixth edition of the Origin, he adheres to the belief that sterility is an incidental[289] rather than a selected quality. The result of his work on heterostyled plants is of importance as showing that sterility is no test of specific distinctness, and that it depends on differentiation of the sexual elements which is independent of any racial difference. I imagine that it was his instinctive love of making out a difficulty which to a great extent kept him at work so patiently on the heterostyled plants. But it was the fact that general conclusions of the above character could be drawn from his results which made him think his results worthy of publication.
FOOTNOTES:
[269] The "Genealogy of Animals" (The Academy, 1869), reprinted in Critiques and Addresses.
[270] An English edition is published by the Clarendon Press, 1890.
[271] Sachs, Geschichte d. Botanik, p. 419.