CONSTANCE: SYSTEMATICS OF THE ANGIOSPERMS 411 



ledons (via Alismatales) and for a hypogynous line (Strobiloideae) and a peri- 

 gynous-epigynous line (Cotyloideae) of dicotyledons, ten)iiti;i1iiig in Lahiatae 

 and Compositae, respectively. Apetalons dicots were all referred to ])rcdomi- 

 nantly choripetalons orders, and "reduction" was heavily stressed. The three- 

 dimensional depiction of Bessey's system by Rodriguez (1950) is particularly 

 instructive. The actual sequence of groups closely followed that of Bentham 

 and Hooker, and it may be questioned whether the dicta were not drawn from 

 observations on this arrangement, rather than the arrangement of orders deter- 

 mined by application of the dicta. Bessey specifically cautioned (1897, p. 169) : 



And let it be fully understood that this is not presented as final, or as entirely satis- 

 factory; it is merely a working hypothesis which claims no other merit than that of an 

 attempt at conformity to the suggestions sometimes faint, sometimes doubtful, from 

 palaeontology, fi-om embryology (ontogeny), and from morphology. 



There were recognized in all some 300 families under 32 orders, with the brief- 

 est kind of skeletal descriptions. The classification is readily presentable in 

 graphic form and the postulated evolutionary trends may be easily grasped. 

 Among the chief objections to the system are the gross oversimplification which 

 results in ponderous and polymorphous orders, the lack of substantiation for 

 some of the basic evolutionary assumptions, the great weight accorded hypogyny 

 versus perigyny-epigyny, and the scant attention given to tropical groups. More 

 than any system proposed before it, however, the Besseyan served to emphasize 

 the objective of basing classification squarely on presumed phylogeny. Hallier 

 (1905) asserted, similarly, the monophyletic character of the angiosperms, the 

 primitiveness of the Magnolian floral type (which he believed to have been de- 

 rived directly from cycadophyte ancestors), the origin of monocotyledons from 

 dicotyledons, and the unnatural character of such divisions as Apetalae and 

 S^Tupetalae. "The Amentiferae" he regarded as the "highest and most reduced 

 types of one of the lines of Dicotyledons" (1905. p. 154). Later, he decided 

 (1912) that Berberidaceae, rather than ilagnoliaceae, represented the key 

 primitive group of angiosperms. 



Hutchinson, like Bessey and Hallier, stressed the theme that "a phylogenetic 

 system of classification should be the ultimate aim of taxonomy. In fact the 

 description of every new genus, every new species or form of plant, may be re- 

 garded as a contribution towards this end" (1926, pp. 1-2). From his extensive 

 experience with African plants, Hutchinson pointed out tlie difficulty of charac- 

 terizing large groups by a general tendency founded on a single structure, and 

 believed that a more natural system might be obtained by the recognition of 

 smaller groups bound together by a combination of characters. Thus, his clas- 

 sification admits 332 families in 105 orders. The major peculiarity of this sys- 

 tem is the recognition within the dicotyledons of two major parallel lines of 

 development: a woody one stemming from arborescent Magnoliales, and a her- 

 baceous one arising from herbaceous Ranales. In his volume on monocotyledons, 

 Hutchinson (1934) proposed a quite original classification. Although monocots 

 were regarded as a monophyletic offshoot of herbaceous Ranalian dicots, it was 

 suggested that the group was early differentiated into three principal (and a 

 number of minor) evolutionary lines: (1) Calyciferae, with a biseriate perianth 

 and a rhizotomous habit; (2) Corolliferae, with a uniseriate perianth and a 



