Mar., 1'.) 11.] The Classification of Plants, VI. 295 



greater numbers involved, and the complexity of structure. This 

 is especially true of the choripetalous families, where it is almost 

 impossible for the mind to grasp the enonnous number of types 

 to be considered. 



The Thalamiflorae are the lowest dicotyls. The first order, 

 the Ranales, constitute a parallel group to the hypogynous 

 Helobiae. The Ranales are closely followed by the Sarraceniales, 

 Brassicales, and the lower Gerianales and Malvales. 



The Centrospermae are a small branch having its origin 

 in the Thalamiflorae. Its lowest family, the Caryophyllaceae, 

 indicates the relationship while the higher forms are greatly 

 specialized, passing over into reduced apetalous and naked flowers 

 with high development of the inflorescence. 



The Calyciflorae represent another great, fundamental branch 

 of choripetalous dicotyls arising from near the Ranales, but some- 

 what more specialized. The lower genera of Rosaceae show 

 marked resemblances to some Ranunculaceae. The other fam- 

 ilies placed in this subclass are fairly certain and there may be 

 families and genera at present associated with other subclasses 

 that properly belong here. 



The Amentiferae are a small but important group whose real- 

 tionship may be traced from the lower Calyciflorae through the 

 Hamamelidaceae, Platanaceae, Ulmaceae, Moraceae, etc., up to 

 the Salicaccac. A ntunber of families now included arc uncertain. 

 One of the structures of the group is the highly specialized flower 

 cluster, the catkin, and there are other significant features as 

 chalazogamy besides various peculiarities of buds, twigs and leaves. 



The Myrtiflorae are an epigynous branch, mostly of chori- 

 petalous dicotyls, probably derived from the Calyciflorae, from 

 the vicinity of the Saxifragales. Some of the families may be 

 excluded in the future but the main mass represents a distinct 

 type of floral development and appears to be phylogenetically 

 related, excepting, perhaps the cactales which may even belong to 

 the Centrospemiae. 



The Heteromerae appear to have come from near the same 

 region as the Centrospermae. The strong resemblance of certain 

 Caryophyllaceae to Primulaceae is very suggestive and gives 

 support to this view. The Heteromerae also show their primitive 

 character by frequent choripetaly and other peculiarities. 



From the lower Heterom.erae it is but a step to the Tubiflorae, 

 the lowest forms of which are represented by the Convolvulaceae 

 and Polemoniaceae. The families of the Tubiflorae are for the 

 most part quite certain, except perhaps the Plantaginaceae. 



The last and highest subclass of Dicotylae, the Inferae, appears 

 to be an offshoot from the lower Calvciflorae, the line leading 

 almost directly from the Saxifragales to the Umbellales, Rubiales, 

 Campanulales, and Compositales. 



