ii4 THE COLOURS OF FLOWERS. 



the bracts. This case may be looked upon as 

 analogous to those of the ripe cones in the juniper 

 and yew, which have similarly assumed the guise of 

 attractive fruits, eaten by birds, who disperse their 

 seeds. Such gymnosperms may be said metaphori- 

 cally to have taken a hint from the angiosperms 

 about them, and have acted upon it for their own 

 advantage. 



It has .been assumed throughout that the progres- 

 sive modification of the colours of petals is due in 

 the main to increased oxidation of their contents. It 

 may be added here that the thin edges of the petals 

 where the newer colours usually first make their ap- 

 pearance, are the parts where oxidation would most 

 naturally take place. Hence, probably, the distinct 

 analogy between fading colours and progressive 

 colours. In most cases, colours appear most vividly 

 on the outside of the petals, where they were ex- 

 posed in the bud, and where oxidation would most 

 readily occur. The red tinge on the outside of 

 daisies, apple-blossom, &c., is here once more very 

 significant. In Convolvulus arvensis the mass of the 

 corolla is white ; but the lines exposed in the bud 

 are deep pink, evidently from oxidation ; and at the 

 same time they form excellent honey-guides of the 

 ordinary simpler sort. In many others of the same 

 genus a similar result may be observed. The under 

 side of the petals in St. John's worts, and the back 

 of the standard in Lotus, are frequently ruddy red. 

 The outer glumes of grasses are often purplish ; the 

 fruiting perianths of Rumex grow red as they ripen. 

 Put beside the rosy cheeks of peaches, apples, and 

 many other fruits, and the obvious oxidation colours 



