36o THE BIOLOGY OF FLOWERING PLANTS 



associated with a variety of organs, may occupy a number of 

 positions. In the orchids the nectar is secreted by the walls 

 of the perianth spur. In the violets it is contained in a 

 corollar spur but is secreted by two staminal spurs. In the 

 UmbeUiferae and many others, the nectaries form discs or 

 protrusions on the upper surface of the ovary, while in 

 Allium they lie on the outer ovarial wall. In many Rosacea^ 

 the nectaries are on the inner surface of the cup-shaped re- 

 ceptacle ; in the Cruciferae they form discs at the apex of the 

 receptacle. In Ranunculus they lie at the base of the petals, 

 sometimes covered by a scale ; in Helleborus the cup-hke 

 nectaries are modified stamens. These few examples show 

 that the secretion of nectar may be performed by almost 

 any organ of the flower. 



The nectar is a watery fluid containing about 25 per cent. 

 of glucose. It should not be confused with honey, the 

 product manufactured by the bee, of which it is only the 

 raw material. Many of the features which may protect 

 pollen from rain also serve to prevent the nectar from being 

 washed out. Whether the danger of this is great may be 

 doubted, for in a multitude of flowers the nectar is freely 

 exposed. It is of interest to know that Sprengel's first 

 observation on floral biology was the presence of hairs in 

 the throat of the corolla of Geranium sylvaticum, which he 

 decided had the function of preventing the nectar from being 

 washed out. From this beginning sprang his great work on 

 pollination. 



Colour. — Colour and scent are the two chief means by 

 which the insect is guided to the flower, and a good deal of 

 discussion has taken place as regards their relative efficiency. 

 For us the flower's most striking character is, in most cases, 

 its colour ; it does not follow that the same is true for the 

 bee, butterfly, and fly. The question turns on the colour 

 sense of insects, and can only be answered experimentally, 

 though there are suggestive facts open to observation. 



It is significant that in the diff^erent classes of flowers 

 detailed above difi^erent colours prevail. T'here are many 

 exceptions, but, speaking generally, in class A the colour 



