PART I. | HISTORICAL INTRODUCTION, 7 
a powerful impetus to further research based upon Sprengel’s 
work. Even Treviranus’s well-founded objection that in most 
orchids, even our native species, the operation of inséct-visitors was 
only concluded indirectly from the structure of the flowers and 
had not been directly observed, could not hinder this result of 
Darwin’s book; on the contrary, it could only direct more general 
attention to the insects that actually performed the work of 
fertilisation, A paper published seven years later by Darwin (159), 
enumerating the insects observed to visit a large number of native 
orchids, shows to how great an extent attention had been attracted 
to this point. 
Another line of investigation which Darwin initiated in the 
same masterly way, was the direct observation of differences 
between the action of pollen from the same and from another 
flower. It has been already mentioned that Sprengel instituted 
experiments on self-fertilisation in the case of Hemerocallis 
Julva, and that he recognised that the flowers of this plant were 
not fertile to their own pollen, and other instances of the same 
kind, or at least of diminished fertility upon self-fertilisation, 
had been accumulated by other botanists as has been already 
explained. Darwin collected these scattered facts and brought 
them under Knight’s law which for the first time revealed their 
full significance, and at the same time he initiated a new method of 
research which placed the produce resulting from self-fertilisation 
alongside of the produce of cross-fertilisation in the struggle for 
existence, and so permitted the result of persistent self-fertilisation 
to be ascertained under natural conditions. In numerous plants 
which were found to be fertile with their own pollen, he fertilised 
some flowers with their own pollen and other flowers on the same 
plant with pollen from a neighbouring plant growing under the 
same conditions, excluding insects with great precautions; he 
allowed the resulting seeds to germinate on damp sand in the same 
vessel, and then planted them in pairs on opposite sides of the 
same pot; then, while all conceivable precautions were taken to 
keep both sides under the same conditions, he watched the growth 
of the plants to maturity. The same experiment was repeated 
with the seeds produced by these, and was continued from 
generation to generation. 
In several cases (Jpomea purpurea, Mimulus luteus), the plants 
resulting from cross-fertilisation showed even in the first generation 
a marked superiority over the others; they were larger in the 
proportion of four to three or even three to two, they flowered 
