88 THE COLOURS OF FLOWERS. 



the petals have generally become quite obsolete ; in 

 some species of Lepidium, Arabis, Draba, &c., they are 

 inconspicuous and often wanting. So in the smaller 

 AlsinecB there are many signs of Degeneration. The 

 normal forms of CaryopJiyllaccce have two whorls of 

 five stamens each ; but these little creeping or weedy 

 forms have often only one whorl, as in Holostewn, 

 some Cerastiums, the smaller Stellarias, Spergula, 

 Polycarpoit, &c. In Sagina, Cherieria, and other very 

 small types, the petals are often or always wanting. 

 Indeed, most botanists will probably allow that nearly 

 all our minute- flowered species, such as Montia 

 fontana^ Claytonia perfoliata, Elatine hexandray Radiola 

 Millegrana, Circcea bttetiana, Liidzuigia pains trisy 

 Peplis Porttda, Tillcsa inuscosa, MyriopJiylliun spicattim, 

 Hipptiris vidgaris, Centuncidiis ininivius^ and Cicendia 

 pusillay are distinctly degenerate forms. Though ob- 

 viously descended from petaliferous ancestors, and 

 closely allied with petaliferous genera or species, 

 many of them have lost their petals altogether, while 

 others have them extremely reduced in size. In 

 several cases, too, the number of sepals, petnls, or 

 stamens has been lessened, and the plant as a whole 

 has suffered structural degradations. Most of these 

 dwarfed and degenerate flowers, if they have petals 

 at all, have them white or very pale pink. 



Readers of Sir John Lubbock's admirable little 

 book on British Wild-Flowers in Relation to Insects 

 will readily understand the reason for this change. 

 They will remember that white flowers, as a rule, 

 appeal to an exceptionally large circle of insect 

 visitors, mostl}'- of small and low grades. Hence, 

 some among these very small flowers may often 



