8 THE FERTILISATION OF FLOWERS. [PART I. 
earlier, and they produced more capsules. In other cases such 
superiority was less marked or was imperceptible, but in no mp 
case was the advantage shown upon the other side. 
So these researches also gave results on the whole favourable to 
Knight’s law; but probability bordering on certainty could only be 
attained when researches of this kind were undertaken on the 
largest scale and carried on through many generations. As far as 
lay within one man’s power Darwin had fulfilled this condition, 
for he had prosecuted the above-mentioned research for eleven 
years. The number of individual plants produced by crossing and 
also of those produced by self-fertilisation which he watched from 
germination to maturity reached more than a thousand ; they be- 
longed to fifty-seven species, fifty-two genera, thirty large families, 
and included natives of the most various countries. The result of 
this whole research may be summed up in the single sentence: 
“Whenever plants which are the offspring of self-fertilisation are 
opposed in the struggle for existence to the offspring of cross- 
fertilisation, the latter have the advantage ;”’ whence it is clear 
that all peculiarities which aid cross-fertilisation will be retained 
and perfected by natural selection. But if the direct rivalry does 
not occur, self-fertilisation may suffice for the propagation of the 
organism for an unlimited number of generations, producing healthy 
and fertile offspring. 
A third line of research prosecuted by Darwin with the same 
success, namely the experimental study of dimorphic and trimorphie 
plants, took away the last shadow of foundation from the old 
belief that species differ radically from varieties, and threw some 
light upon the obscure question of hybridisation. Sprengel had 
remarked, in discussing Hottonia palustris (p. 103): “Some 
plants bear only flowers whose anthers are included within the tube 
but whose style is exserted, and others bear only flowers whose 
style is shorter and whose stamens are longer than the tube of the 
corolla. I believe that this is not accidental but that it is a 
disposition of nature, though I am not able to suggest its purpose.” 
Subsequently this peculiar phenomenon of long-styled and short- 
styled plants was recognised in several other species, and in 
Lythrum salicaria three forms were found, long-styled, mid-styled, 
and short-styled, without any further use being made of the facts. 
Darwin subjected the peculiar sexual relations of these plants, 
which he called dimorphic and trimorphic, to a closer considera- 
tion ; he also performed numerous experiments, crossing the plants 
in various ways, and cultivating the produce of the various unions. 
