CONSTANCE: SYSTEMATICS OF THE ANGIOSPERMS 441 



factors to serve a complex end. The morphological factors are the province of compara- 

 tive anatomy; the chemical factors deserve treatment by a similar discipline. Before 

 they can aid us in understanding the evolutionary problem, we must develop criteria for 

 judging their true homologies. This will come as we learn more of the origins of the 

 substances we find in particular organisms and understand something of the mechanisms 

 underlying their multiplication and variation. When it comes our studies will be on a 

 new footing. 



Primitive Habit of Angiosperms 



Tlie groups of dicotyledonous angiosperms postulated as most primitive in 

 the current phylogenetic classifications, with one exception, are indicated as 

 being perennial, woody, and arborescent. The system of Hutchinson visualizes 

 separate woody and herbaceous lines of dicots. The large measure of agreement 

 on the priority of woody habit is based on: (1) the prevalence of woody habit 

 in g^Tnnosperms; (2) its correlation with anatomical and floral characters re- 

 garded as primitive and, conversely, the association of advanced anatomical and 

 floral characters with herbaceous habit; (3) the predominance of woody plants 

 in moist tropical and subtropical areas, habitats which are widely believed to be 

 the modern equivalents of the climatic conditions widespread at the period of 

 angiosperm inception; (4) the revelation of the peristence of cambium in seed- 

 ling and herbaceous stems; (5) the supposed correlation between fleshy fruits 

 and woody habit; (6) the evolution in vascular erytogams from protostele to 

 siphonostele to eustele; and (7) the lack of undoubted fossil herbaceous angio- 

 sperms in older geological strata. Bews stated his conviction that "the earlier 

 fossil Angiosperms were closely similar to the types now occurring in moist and 

 subtropical areas" (1927, p. 20), and visualized this ecological type as having 

 given rise both to the trees and shrulis of arid and temperate regions and the 

 lianes, epiphytes, and herbs developing within forest understories, openings, and 

 margins, and spreading outward. Andreanszky (1950, 1952) entertains similar 

 views, but believes that such aquatics as Nymphaeaeeae were formed very early. 

 Davy suggested that the suffruticose habit forms "an intermediate stage in the 

 evolution of an herbaceous from an arborescent type" (1922, p. 219). Sinnott 

 and Bailey stipulated that "the herbaceous vegetation of today should be re- 

 garded as of comparatively recent development" (1914, p. 595; 1915a, 1915b; 

 Eames, 1911; Sinnott, 1916; Jeffrey and Torrey, 1921). 



The chief recent opponent of the primitiveness of arborescent habit has been 

 Arber. She suggested that the wide distribution of her])s argues for their an- 

 tiquity, that the separate vascular bundles of dicot herbs could not have been 

 attained by the progressive dissection of a continuous woody cylinder, and that 

 any correlation of primitive features with arborescent habit was due to the "evo- 

 lutionary lag" imposed by the longer life-span of woody as opposed to herba- 

 ceous plants. 



The frequency of the tree habit in Angiosperms is held to point to the extreme 

 antiquity of the flowering plant stock, which has allowed time for many lineages to reach 

 a phase of senility; for trees show two characters which are indicative of old age in 

 animal races — growth to a relatively large size, and the accumulation of non-living 

 material in the body (Arber, 1928, p. 83). 



She pointed to Agavaceae, Clematis, and Berheris as examples of the origin 



