CH. xxv] RANALES AND CRUCIFERS 309 



Dealing first with those more primitive members of the 

 Archichlamydeae, which are known as the Polypetalae, we find 

 that, in the Ranalean plexus, the Nymphaeaceae offer a striking 

 example of a family rich in genera and species, and consisting 

 entirely of water and marsh plants. There is great variation in 

 the structure of the flower, and the carpels range from superior 

 to inferior. The variety of form occurring in the family suggests 

 that it is an old one which has had a long time to evolve, since 

 it adopted aqua tic life. It should be noted that various observers 1 , 

 have regarded the Nymphaeaceae as Monocotyledons, but it 

 seems more reasonable to suppose that they are truly Dicoty- 

 ledonous, though descended from a stock closely related to that 

 which gave rise to the Monocotyledons. 



The curious genus Ceratophyllum, on whose affinities the 

 most divergent claims have been made, seems best regarded 

 as a reduced form, closely related to the Nymphaeaceae 2 and es- 

 pecially the Cabomboideae 3 . Thus this genus, which on account 

 of its extreme specialisation is reasonably relegated to a distinct 

 family, may be regarded as the ultimate term in the Nymphaea- 

 ceous series ; its rootlessness, reduced anatomy and submerged 

 pollination indicate how completely it has identified itself with 

 aquatic life. 



The Ranunculaceae are typically terrestrial, but the genus 

 Ranunculus contains, besides purely terrestrial species, the sub- 

 genus Batrachium which is definitely aquatic, and also a number 

 of species such as R. Flammula, which are amphibious. 



The Cruciferae include certain types, e.g. Nasturtium amphi- 

 bium and Cardamine pratensis, which are capable of living either 

 in damp places or actually submerged. These form a link 

 between the terrestrial Crucifers and such definitely aquatic 

 forms as Subularia aquatica. This plant, which superficially 

 resembles a tiny Juncus, lives entirely submerged and has been 

 described as cleistogamic. 



1 Trecul, A. (1845) and (1854), Henfrey, A. (1852), Seidel, C. F. 

 (1869), Schaffner, J. H. (1904), Cook, M. T. (1906). 



2 Brongniart, A. (1827), Strasburger, E. (1902). 3 Gray, A. (1848), 



