PART I.] HISTORICAL INTRODUCTION. 23 
It is obvious that this proposition. is much narrower than the 
Knight-Darwin law. For it is easily conceivable, and it is com- 
patible with all known facts, that, in all plants without exception, 
the offspring of self-fertilisation, as soon as they come into com- 
petition with other individuals of their species which are the 
offspring of cross-fertilisation, finally succumb, and that therefore 
the above-mentioned proposition is thoroughly correct; but that 
nevertheless many species which are regularly self-fertilised, and in 
which the struggle for existence between the offspring of self- 
fertilisation and of cross-fertilisation never takes place, reproduce 
by self-fertilisation for an unlimited number of generations, and 
that therefore the Knight-Darwin law is false. 
Whether the offspring of self-fertilisation finally succumb in 
the struggle for existence to the offspring of cross-fertilisation in 
the same species may probably be decided in the course of a few 
generations for certain plants by the above-described experiments 
contrived by Darwin. Whether, on the other hand, plants in which 
such competition is avoided, owing to exclusive and continuous 
self-fertilisation, finally become extinct for want of crossing, can 
probably in many cases not be decided. At least Darwin declares 
expressly (Variation, etc., chap. xvii.) that, in order to recognise 
the difference between plants produced by self-fertilisation and 
those produced by cross-fertilisation, it is often absolutely necessary 
to place both together in competition. And, moreover, important 
facts stand opposed to the Knight-Darwin law, which have mostly 
been collected by Darwin in his Variation of Animals and Plants 
with scrupulous care. I may refer to the examples of plants 
(mentioned in chap. xviii. of that work) which spread over 
unlimited areas by asexual reproduction; to the Brazilian culti- 
vated plants, described by Fritz Miiller and discussed in the third 
section of this book, which after they have been multiplied for 
many generations exclusively by asexual means, have lost even the 
power of sexual reproduction ; to the well-known facts that numerous 
mosses reproduce over great part of their area of distribution by 
exclusively asexual means, and that many are only known in the 
sterile form: all facts which are not easily brought into accordance 
with the supposition that occasional crossing is essential for lasting 
conservation. 
There is good foundation, therefore, for the demand that the 
explanation of floral mechanisms shall rest only on the sufficient 
and demonstrable assumption that cross-fertilisation Beige more 
vigorous offspring than self-fertilisation. 
