PART I.] HISTORICAL INTRODUCTION. 11 
at first and in the bastard offspring is not due to differences in the 
general structure, but exclusively to a difference in the action 
of the reproductive elements. 
Attracted by Darwin’s brilliant researches on heterostyled plants 
(1861-1868), many other investigators have since worked at the 
same subject. Darwin has collected their results, controlling them 
by his own observations and experiments and incorporating them 
with the results of his later investigations. His book contains all 
that we certainly know as yet concerning heterostyly, and treats also 
as comprehensively of polygamous, dicecious, and _ cleistogamic 
plants. It restricts the name polygamous to plants which: possess 
male, female, and hermaphrodite flowers, and introduces the fol- 
lowing new terms: gynodiecious, for plants which have hermaphro- 
dite and also purely female individuals; androdiecious, for plants 
which consist of hermaphrodite and purely male individuals ; gyno- 
monecious, for plants which have hermaphrodite and purely female 
flowers upon the same individual ; andromonecious, for plants in 
which one individual bears hermaphrodite and male flowers. 
The most complete collection of all the known facts which 
contribute to prove Knight’s law is given by Darwin in his work 
on the Variation of Animals and Plants under Domestication, in 
which he suggests many new and fruitful lines of research ; but 
the three methods of investigation which Darwin originally used 
have been the chief aids in investigating the determining con- 
ditions of the forms of flowers. Numerous observers, among whom 
Friedrich Hildebrand, Federico Delpino, my brother Fritz Miiller,and 
Severin Axell deserve special mention, have pushed forward along 
these new paths that Darwin opened ; they have not only brought to 
light a mass of new facts, all tending to elucidate floral mechanisms 
on the basis of the Knight-Darwin law, but they have also disclosed 
many new general principles. These we may briefly review. 
Hildebrand, in several laborious works, demonstrated that many 
floral contrivances which Sprengel had investigated, but had 
explained on the theory of self-fertilisation, were really adapted 
for regular cross-fertilisation; and he explained on the same 
principle, and in most cases figured, many forms that Sprengel had 
not examined. He extended the list of dimorphie and trimorphie 
plants, and applied Darwin’s experiments to Primula sinensis, 
Pulmonaria officinalis, and several species of Ozalis; he also 
performed artificial self-fertilisation on these forms, and found that 
it led in general to even greater sterility than the illegitimate 
crossings. By his experiments on Corydalis cava, he showed that 
