466 TROPICAL NATURE Ix 
made showing that crosses between flowers with stamens and 
styles of unequal length were always nearly barren. During 
these experiments 20,000 seeds of Lythrum salicaria were 
counted under the microscope. For several years a further 
supplementary series of experiments were carried out, show- 
ing that the seeds produced by the illegitimate crosses (as he 
terms them) were not only very few, but, when sown, always 
produced comparatively weak, small, or unhealthy plants, not 
likely to exist in competition with the stronger offspring of 
legitimate crosses. There is thus the clearest proof that these 
complex arrangements have the important end of securing 
both a more abundant and more vigorous offspring. 
Perhaps no researches in the whole course of the study of 
nature have been so fertile in results as these. No sooner 
were they made known than observers set to work in every 
part of the world to examine familiar plants under this new 
aspect. With very few exceptions it is now found that every 
flower presents arrangements for securing cross-fertilisation, 
either constantly or occasionally, sometimes by the agency of 
the wind, but more frequently through the mediation of 
insects or birds. Almost all the irregularity and want of 
symmetry in the forms of flowers, which add so much to their 
variety and beauty, are found to be due to this cause; the 
production of nectar and the various nectar-secreting organs 
is directly due to it, as are the various odours and the various 
colours and markings of flowers. In many cases flowers which 
seem so simply constructed that the pollen must fall on the 
stigma and thus produce self-fertilisation, are yet surely cross-> 
fertilised, owing to the circumstance of the stigma and the 
anthers arriving at maturity at slightly different periods, so 
that, though the pollen may fall on the stigma of its own 
flower, fertilisation does not result; but when insects carry 
the pollen to another plant the flowers of which are a little 
more advanced, cross-fertilisation is effected. There is liter- 
ally no end to the subjects of inquiry thus opened up, since 
every single species, and even many varieties of flowering 
plants, present slight peculiarities which modify to some 
extent their mode of fertilisation. This is well shown by the 
remarkable observations of the German botanist Kerner, who 
points out that a vast number of details in the structure of 
