54 THE COLOURS OF FLOWERS. 



I think, fairly establishes the fact that white is a 

 higher development of colour than yellow ; for the 

 change must surely have been made in order to 

 attract special insects. Certainly, such a differentia- 

 tion of the flowers in a single head cannot be without 

 a good purpose. In the true daisy, again (Bcllis 

 perennis], the white rays become tipped with pink (Fig. 

 1 8) which sometimes rises almost to rose-colour ; and 

 this stage is exactly analogous to that of apple- 

 blossom, which similarly halts on the way from white 

 petals to red. In our own asters (A. tripolium, &c.) 

 and the Michaelmas daisies of America, we get a 

 further advance to purple, lilac, and mauve, while 

 both in these and in the chrysanthemums true shades 

 of blue not infrequently appear. The Cinerarias of 

 our gardeners are similar forms of highly-developed 

 groundsels from the Canary Islands. 



Tubular flowers with an irregular corolla are ob- 

 viously higher in their mode of adaptation to insect 

 visits than tubular flowers of the ordinary symmetrical 

 type. Amongst them, the first place must be as- 

 signed to the Labiates perhaps the most specialised 

 of any so far as regards insect fertilisation. Not 

 only are they deeply tubular, but they are very 

 bilateral and irregular indeed, displaying more modi- 

 fication of form than almost any other flowers except 

 the orchids. They mostly secrete abundant honey, 

 and often possess highly aromatic perfumes. More- 

 over, they form geologically one of the latest families 

 of flowering plants, specially developed in adaptation 

 to bees and other highly-evolved honey-suckers. 

 Almost all of them are purple or blue. Among the 

 best known English species are thyme, mint, marjoram, 



