Cross and Self-fertilisation 413 
experiments, as we have pointed out’, could determine whether cross- 
fertilisation is in itself beneficial, and self-fertilisation on the other 
hand injurious; a conclusion which a merely comparative examination 
of pollination-mechanisms renders in the highest degree probable. 
Later floral biologists have unfortunately almost entirely confined 
themselves to observations on floral mechanisms. But there is little 
more to be gained by this kind of work than an assumption long ago 
made by C. K. Sprengel that “very many flowers have the sexes 
separate and probably at least as many hermaphrodite flowers are 
dichogamous ; it would thus appear that Nature was unwilling that 
any flower should be fertilised by its own pollen.” 
It was an accidental observation which inspired Darwin’s experi- 
ments on the effect of cross and self-fertilisation. Plants of Linaria 
vulgaris were grown in two adjacent beds; in the one were plants 
produced by cross-fertilisation, that is, from seeds obtained after 
fertilisation by pollen of another plant of the same species; in the 
other grew plants produced by self-fertilisation, that is from seed 
produced as the result of pollination of the same flower. The first 
were obviously superior to the latter. 
Darwin was surprised by this observation, as he had expected 
a prejudicial influence of self-fertilisation to manifest itself after a 
series of generations: “I always supposed until lately that no evil 
effects would be visible until after several generations of self-ferti- 
lisation, but now I see that one generation sometimes suffices and 
the existence of dimorphic plants and all the wonderful contrivances 
of orchids are quite intelligible to me®.” 
The observations on Linaria and the investigations of the results 
of legitimate and illegitimate fertilisation in heterostyled plants were 
apparently the beginning of a long series of experiments. These 
were concerned with plants of different families and led to results 
which are of fundamental importance for a true explanation of sexual 
reproduction. 
The experiments were so arranged that plants were shielded from 
insect-visits by a net. Some flowers were then pollinated with their 
own pollen, others with pollen from another plant of the same species. 
The seeds were germinated on moist sand; two seedlings of the same 
age, one from a cross and the other from a self-fertilised flower, were 
selected and planted on opposite sides of the same pot. They grew 
therefore under identical external conditions; it was thus possible to 
compare their peculiarities such as height, weight, fruiting capacity, 
etc. In other cases the seedlings were placed near to one another in 
the open and in this way their capacity of resisting unfavourable 
external conditions was tested. The experiments were in some cases 
1 Ante, p. 408. 2 More Letters, Vol. u. p. 373. 
