20 Flowers and their Pedigrees. 



lakes and rivers.' From some such simple form as 

 this we may be pretty sure that all existing flowering 

 plants are ultimately descended. 



In most modern flowers, however, each blossom 

 contains several stamens and several carpels (or pistil- 

 divisions), and the way in which such a change as 

 this might come about can be .easily imagined ; for 

 even in many existing plants, where the separate 

 flowers have only a single stamen or a single pistil 

 each, they are nevertheless so closely packed together 

 that they almost form a single compound flower, as 

 in the case of the bur-reed and the various catkins, 

 not to mention the arum and the spurge, where only 

 a trained eye can make out the organic separateness. 

 I shall not trouble you much, however, with these 

 earlier stages in the development of the daisy, both 

 because I shall describe them elsewhere in part, 

 d propos of other subjects, and because the later stages 

 are at once moPe interesting and more really instruc- 

 tive. It must suffice to say that at some very 

 ancient period the ancestors of the daisy, and of 

 one half the other modem flowers, had acquired an 

 arrangement of stamens and pistils in groups of five, 



' In all probability, the duckweed is not itself a'rea'ly primitive 

 type, but a degraded descendant of higher ancestors. This, however, 

 does not prevent it from standing as an excellent representative of the 

 real original unspecialised flowering p'ant, which must have been quite 

 as simple in structure. 



