THE ORIGIN OF PETALS. ir 



the Wolfian theory is absolutely irrefutable — but with 

 the light shed upon the subject by the modern 

 doctrine of evolution, we can no longer regard petals 

 and sepals as intermediate stages between the two. 

 The earliest flowering plants had true leaves on the 

 one hand, and specialised pollen-bearing or ovule- 

 bearing leaves on the other hand, which latter are 

 what in their developed forms we call stamens and 

 carpels ; but they certainly had no petals at all, and 

 the petals of modern flowers have been produced at 

 some later period. It is probable, too, that they have 

 been produced by a modification of certain external 

 stamens, not by a modification of true leaves. Instead 

 of being leaves arrested on their way towards 

 becoming stamens, they are really stamens which 

 have partially reverted towards the condition of 

 leaves. They differ from true leaves, however, in 

 their thin, spongy texture, and usually in the bright 

 pigments with which they are adorned. 



All stamens show a great tendency easily to become 

 petaloid, as it is technically called ; that is to say, 

 to flatten out their filament or stalk, and finally to 

 lose their pollen-bearing sacs or anthers. In the 

 waterlilies — which are in certain ways one of the 

 oldest and simplest types of flowers we now possess 

 still preserving many antique points of structure 

 unchanged — we can trace a regular gradation from 

 the perfect stamen to the perfect petal. Take for 

 example our common English white waterlily, 

 Nymph(Ea alba (Fig. 3). In the centre of the flower 

 we find stamens of the ordinary sort, with rounded 

 stalks or filaments, and long yellow anthers full of 

 pollen at the end of each ; then, as we move outward, 



