208 PLANT-LIFE 



leaf, the petal is a leaf; stamens and carpels are also 

 leaves, and because they bear spores they are correctly 

 described as spore-bearing leaves — sporophylls. We 

 naturally conclude that plants possessing such highly 

 modified leaves are much farther advanced than plants 

 which do not possess them. And these modifications 

 involve specialization in function. The cone-scales of 

 the seed-bearing Pine are modified leaves, and the 

 Pine is an undoubted advance upon a Club-Moss. A 

 gaily coloured perianth may usually be regarded as an 

 advance upon the monotonous green. Cohesion of 

 parts is certainly an indication of specialization. If we 

 detect progress from a lower to a higher form in the 

 modification of green leaves into highly coloured petals, 

 we must note further progress when petals are united, 

 or coherent, or, in other words, when the corolla is a 

 unit composed of organically coherent petals. This 

 being so, we have no hesitation in concluding that the 

 Sympetalse are as a whole more advanced than the 

 Choripetalse, because the latter have free perianth leaves, 

 while the former have coherent petals. The modifica- 

 tion of a leaf into a carpel is evidence of advance, and 

 specialization in the carpels marks a higher development. 

 A Buttercup bears flowers in which all the parts are 

 separate or free; its carpels are clustered together, but 

 not fused. It is, therefore, less highly specialized than 

 the Primrose, which is sympetalous, and has its carpels 

 fused into a compound pistil. Indeed, we may take it 

 for granted that in floral evolution the passage from 

 lower to higher types has involved reduction in the 

 number of free parts — a reduction due to their cohesion. 

 Bilabiate flowers, such as the Toadflax, Snapdragon, 



