558 THE FERTILISATION OF FLOWERS. [PART III. 
bear male flowers only. That is to say, the species exhibits a 
passage through andromonescism to androdiccism. The flowers 
are dirty-yellow, and the visitors are principally flies (609). 
REVIEW OF LILIACEZ. 
In regard to the genetic relations of Liliacez, the following 
conclusions may be drawn from the floral characters that have 
just been described. 
The Liliaceze must have once possessed open, regular flowers 
devoid of honey, visited and cross-fertilised by pollen-seeking 
insects only. They attained, after dividing into several groups, 
to the secretion of open, freely accessible, honey, partly secreted 
by the segments of the perianth, partly by the carpels; and 
in part they still remain honeyless, and are still crossed only by 
insects which collect or feed on pollen (Tulipa), or else have 
become developed into deceptive flowers which attract stupid carrion- 
loving flies (Paris). Those Liliacez in which honey is secreted by 
the carpels have in part open flowers with generally accessible 
honey (Tofieldia, Anthericum); but in part they have become 
adapted, by approximation of the perianth-segments, for a limited 
but still very miscellaneous lot of visitors (Allium), or even to a 
special long-proboscised form (Paradisia), or by cohesion of the 
perianth-segments into a longer or shorter pendulous bell, to bees 
in general (Convallaria verticillata) or to humble-bees and other 
long-proboscised bees only (C. Polygonatwm). Similarly those 
Liliaceze in which the honey is secreted by the segments of the 
perianth have in part remained with fully or moderately open 
flowers, destined for cross-fertilisation chiefly by short-lipped in- 
sects (Diptera), (Veratrum, Gagea, Lloydia) ; in others the perianth- 
segments have become approximated without cohesion to form a 
pendulous bell, fertilised by bees (Fritillaria); in others the 
nectaries have become modified into narrow covered grooves, 
which are only accessible to Lepidoptera (Lilium); and, finally, 
in this last genus adaptation has passed from diurnal Lepidoptera 
to Sphingidz in the case of Lilium Martagon. | 
These various adaptations have all taken place with complete 
or almost complete retention of the regular symmetry of the flower ; 
only the adaptation to Lepidoptera (in Paradisia and Lilium) and 
the oblique position of the flower in Anthericum have caused an 
