THE LOWER DICOTYLEDONS. 315 



from 5 to 10 carpels. Note the resemblance of each carpel 

 to the single carpel (" pistil ") of Bean or Gorse. Open up 

 a carpel and note the arrangement of the seeds : how many 

 rows are there, and on what side of the carpel, inner (towards 

 centre of flower) or outer, are they carried ? When ripe it 

 becomes a dry fruit (follicle) which splits open along the 

 seed-bearing inner edge ; the seeds are gradually shaken out 

 as the wind rocks the fruit to and fro. 



325. The Buttercup Family ( Rammculaceae) is dis- 

 tinguished by the complete absence of cohesion or adhesion 

 between the parts of the flower ; the sepals, petals, stamens, 

 and carpels are always free [except in Nigella, where the 

 carpels are joined together and the fruit is a capsule]. The 

 stamens are indefinite in number, i.e. they are numerous and 

 bear no definite numerical relation to the petals or sepals. 



The only really common genera in Britain are Ranunculus, 

 Anemone, Clematis, and Oaltha, but many are cultivated in 

 gardens for their showy flowers. 



326. Flower Mechanisms in Raiiun enlace ae. This 

 family shows a wide and interesting range in flower struc- 

 ture, and includes some of the most beautiful examples of 

 adaptation to visits of particular insects. The perianth is 

 generally petaloid and rarely shows a " typical " calyx and 

 corolla, though this does occurinthe largest genus (the one with 

 most species) Ranunculus. In nearly all cases the " petals " 

 either bear nectaries or are represented by more or less elaborate 

 honey-organs. These nectaries, which so often occur between 

 the perianth leaves and the stamens, are commonly regarded 

 as modified petals, but more probably they have been derived 

 from the outer stamens and have never been petals at all. 



Which genera have flowers whose structure seems to sug- 

 gest (1) that the flower has not yet developed petals, whose 

 attractive function is assumed by the coloured (petaloid) 

 sepals, or (2) that petals may have been derived from the 

 outermost stamens, or (3) that sepals may have evolved from 

 foliage-leaves or bracts growing just below the flower? 

 Perhaps the flowers of this family are still in course of 

 evolution ; at any rate, very few show the distinct sepals and 

 petals of a " typical flower." " Sepals " are always present 



