6 THE FERTILISATION OF FLOWERS. [PART I. 
the case of the relatively few hermaphrodites which generally 
fertilise themselves, yet for the majority the common notion that 
the flowers are fertilised with their own pollen, either spontaneously 
or by the aid of insects or the wind, was as yet not disproved (cf. 
Treviranus, 742). They said justly that the observations quoted to 
prove the disadvantages of breeding in-and-in or of fertilising a 
flower with its own pollen were quite insufficient, and they called 
for more extended experiments. Finally, they pointed to the not 
rare occurrence of flowers which inevitably fertilise themselves, 
which even remain closed, and yet which are fully productive, as a 
difficulty in the way of Darwin’s hypothesis not yet removed. From 
the nature of the case, complete proof seems impossible, either for or 
against this law; for neither, if it is true, can the necessity of 
occasional crossing be shown for all bi-sexual plants and animals; 
nor, if it is false, can any hermaphrodite which as a rule fertilises 
itself be kept under observation for an unlimited number of 
generations. But, since the facts which come within the scope of 
this law and by which its validity may be subjected to detailed 
proof are inexhaustibly numerous, continued research will either 
bring the probability of the law to the verge of certainty or make 
its improbability continually more conspicuous. And so the 
Knight-Darwin law was admirably fitted to lead to numerous 
investigations of phenomena hitherto left unobserved, and so to be 
highly valuable in furthering our knowledge, even if in the end its 
truth could not be universally and absolutely affirmed. 
Darwin opened these new lines of investigation with his own 
incomparable researches. A few years after the publication of his 
Origin of Species, he showed by his wonderful book on orchids 
that he had by no means affirmed the general truth of Knight’s 
law without having engaged in special researches himself. For he 
showed in this work that in almost all British orchids, and in all 
the foreign species within his reach, the flowers were adapted down 
to the most minute details for insect-visits, in such a way that 
insect-visitors could not fail to carry the pollen to the stigmas 
of other flowers. Only a few species in which self-fertilisation 
regularly took place formed an exception, as yet unexplained, 
to this general law; but since even in these cases the possibility 
of occasional crossing was not excluded, they formed no valid 
argument against the Knight-Darwin law. 
This work, freed from the fundamental flaw of Sprengel’s 
theory and permeated by Darwin’s acute reasoning and observation, 
was a model for the study of the forms of flowers, and it gave — 
