PART III. | THE MECHANISMS OF FLOWERS. 559 
unsymmetrical curvature of the reproductive organs especially of 
the style. 
The colour of the perianth in Liliaceze must originally have 
been greenish, as it still is in Paris, while the flowers at first 
made themselves conspicuous to insects by the colour of: the 
reproductive organs. By the selective agency of short-lipped 
insects, those flowers which had freely-visible honey, and then 
such as had their honey partially concealed, developed greenish- 
yellow colours (Veratrum), yellow (Tofieldia, Gagea), greenish- 
white and white (Lloydia, Anthericum), and their perianth- 
segments became devoted to purposes of display. It was only 
when sand-wasps, bees, long-tongued flies and Lepidoptera began 
to play a leading part as the fertilising agents, and led to the 
development of flowers that were no longer accessible to the 
great host of short-lipped visitors, that red, violet, and blue 
colours began to be developed through the higher colour-sense of 
these insects. In the genus Allium, for instance, the species with more 
accessible honey (A. wrsinum, A. victoriale) have a white perianth ; 
those with less accessible honey, which are visited chiefly by 
sand-wasps and bees, and also by-Lepidoptera and long-proboscised 
flies (eg. A. rotunduwm), have the perianth red. In Liliacex, 
as in other cases, the flowers which are fertilised by bees have by 
far the greatest variety of colour, as we see by a glance at Tulipa, 
Fritillaria, Scilla, Muscari, Hyacinthus, Asparagus, and Conval- 
laria. The colours of those which are fertilised by Lepidoptera 
are far less various. JLiliwm bulbiferwm, which is fertilised by 
diurnal Lepidoptera, is clothed in fiery red; Paradisia, which is 
adapted for nocturnal species, is white; LZ. Martagon, which 
only became modified for Sphingidze supplementarily, has ex- 
changed its bright colours for dull ones since it ceased to be 
fertilised by diurnal Lepidoptera (609, pp. 55, 56). 
Orv, AMARYLLIDEZ. 
396. GALANTHUS NIVALIS, All.—Sprengel has given a full, 
and on the whole accurate, account of this flower; he was, however, 
inaccurate in regard to the structures which he supposed to 
shelter the honey, and his account is incomplete in regard to the 
circuinstances which favour cross-fertilisation. 
The green parts of the grooves on the inner surface of the 
inner petals secrete and lodge the honey, which is sufficiently 
sheltered from rain by the pendulous position of the flower. As 
